Positioning For A Liquidity Drain: U.S. Interest Rates Outlook 2010

INTEREST RATES RESEARCH
January 2010
U.S. INTEREST RATES OUTLOOK 2010
POSITIONING FOR
A LIQUIDITY DRAIN
PLEASE REFER TO THE LAST PAGE FOR ANALYST CERTIFICATION(S) AND IMPORTANT DISCLOSURES
Barclays Capital | US Interest Rates Outlook 2010
FOREWORD
2009 was the year that the world worried about deflation and a continued spiral down in
the worldwide economy, driven by problems in financial markets. It is easy to forget the
remarkable strength of negative feedback loops between markets and the real economy at
the start of 2009. Financial asset prices would fall, hurting the real economy, which would
cause asset prices to fall further, and the cycle would feed on itself. Unprecedented
monetary and fiscal stimuli worldwide succeeded in breaking these loops. And, fittingly,
financial markets have led the way up. Asset prices bounced back everywhere, and most
major economies started growing in the second half of the year.
All actions have reactions. And given the extreme policy measures taken in 2008-09, it is
inevitable that there be some side effects as those policies started to work. So while 2009
was spent figuring out if stimuli would work, 2010 should be the year of forecasting how
and when these will be withdrawn and what the effects will be. Indeed, it is easy to draw
contrasts between the year to come and the one that passed. Fears of deflation should give
way to concerns over inflation, US securitized and housing markets will likely stop hogging
the headlines, policy focus should shift from easing to tightening, and there could be a rise
in expectations about economies overheating (specifically in Asia). But we expect 2010 to
follow 2009 in one important way. US Treasuries underperformed in 2009, while spread
products did very well – 2010 should be more of the same, though the magnitude of spread
outperformance could be smaller.
Investors will also need to factor in macroeconomic and regulatory changes. Who will buy
US Treasuries once the Fed walks away? What is the end game for the GSEs? Do new
accounting rules spell the end of securitization? What are the effects of increased capital
and liquidity requirements for the banking system? And will inflation expectations finally
come unhinged? In other words, while 2010 should be less exciting than 2009, it will bring
its own set of challenges.
The goal of this publication is to help you, our clients, in navigating these challenges. We
discuss our views on various US rates products against the backdrop of an improving
economy and the start of policy tightening. We look at supply-demand imbalances in US
fixed income, our forecasts for the timing and magnitude of Fed exit strategies, and the
effect thereof. On the securitized front, we update our expectations of losses still to come,
our estimates for further home price declines, and on how we expect the mortgage markets
to manage in the absence of Fed buying. And we wrap up with the best trades to execute
our views, across a range of rates asset classes.
As always, our focus remains unchanged – to help you make informed investment decisions
in the US rates markets in 2010.
Ajay Rajadhyaksha
Head of US Fixed Income and Securitized Research
Barclays Capital
8 January 2010 1
CONTENTS
OVERVIEW
US rates markets: A new, new world 4
US rates markets face a challenging 2010. As the economy recovers and the Fed stops asset
purchases, yields should rise, especially from Q2. We also expect the yield curve to stay
steep in the first half of the year. The housing/MBS markets should fade as a source of
systemic risk, but regulatory changes and sovereign ratings could pose new challenges.
US RATES
Supply/demand: Supply deluge 11
We expect a sharp upward swing in term fixed income supply from 2009 to 2010. This
should overwhelm demand, given the lack of a Fed bid. Moreover, the demand shortfall
should be concentrated in Treasuries. We maintain our curve steepener view and expect
long-end swap spreads to drift tighter through 2010.
Money markets: Short rates – Choppy seas ahead 23

Markets needed relief after the sturm und drang of 2008. A long period of low, stable, and
predictable interest rates in 2009 was exactly what the short rates markets required. But this
sweet spot is about to end in 2010, in our view.
Treasuries: Pick your poison 32

We discuss the implications of Fed policy and the Treasury’s terming out of debt in the
context of total fixed income supply on the level of rates, the curve, and the curvature. We
also highlight our relative value recommendations for the year.
TIPS: Still cheap insurance 42

Although outright TIPS returns may be weak as real rates rise, relative performance should
be positive and breakevens still offer cheap inflation insurance.
Agencies: The long and winding road 48

We examine the scope of potential changes to the GSEs but do not believe anything
substantive will be decided in 2010. A more pressing concern is finding sources of demand
to replace the exiting Fed, as net term supply should remain brisk.
INTEREST RATE DERIVATIVES

Treasury futures: A new bond contract may change dynamics 59
In early 2010, there will be a new contract at the long end of the curve. We expect the
contract to find favor with hedgers, to cheapen the 15y sector and to richen the 25y sector.
Treasury supply-driven trades should continue to offer value.
Swap spreads: Weighed down by supply 65

The swap spread curve should flatten and long-end spreads tighten through 2010. Treasury
supply expectations embedded in long-end spreads are optimistic, and related sovereign
risk is underpriced. Spread widening related to convexity paying or the introduction of the
ULB contract should provide good entry points for these trades.
Vol: Entering a quiet period 74

The first half of 2010 should be characterized by a Fed on hold, an economic recovery that
puts paid to fears of a double dip, and increased volatility supply driven by higher yields. In
this environment, shorter options should drop across tails. However, longer expiries could
stay well bid, given demand from the MBS universe and very little supply.
8 January 2010 2
SPECIAL TOPICS
Sovereigns: How risk-free are US Treasuries? 84
The rise in sovereign risk in other developed economies is a near-term positive for US rates.
However, over the medium term, if the CBO’s deficit projections are realized in conjunction
with even an innocuous rise in rates, the US AAA rating could suffer. Against this backdrop,
continued diversification away from the USD by global central banks could precipitate
negative ratings action, though not in 2010, in our view.
Mortgage basis outlook: Life after Fed 96

Agency MBS will probably adjust well to life after the Fed. The basis should widen around
30bp, but banks, money managers and foreign investors should then provide a backstop.
8 January 2010 3
OVERVIEW
US rates markets: A new, new world

Ajay Rajadhyaksha This year should be one of changes and challenges for US rates investors. The Fed is set
+1 212 412 7669 to stop buying assets, the economy should recover, and supply should continue
ajay.rajadhyaksha@barcap.com unabated. Housing and mortgage markets will probably fade as sources of systemic risk,
but regulatory changes and sovereign debt fears could become new areas of concern.
Rates should trend higher across 2010, with the bulk of the rise occurring in Q2 and Q3.
We expect rates to take a bit of a breather in Q1, given the December sell-off.
Both funds to 10s and the 2s/10s curve should stay around current steep levels in the
first half of the year. We expect curve flattening to start in the second half as Fed
tightening seems imminent.
Volatility should remain low, and swap spreads tight, at least in the next few months. In
agency debt, investors expecting a quick resolution to the status of the GSEs are likely to
be disappointed.
Home prices could drop a little further, before stabilizing by midyear. Meanwhile, there
are considerable losses still to come in non-agency and commercial mortgages, and
hence in the banking system. However, we believe that neither the losses nor the home
price declines are big enough to pose systemic risk.
Regulatory changes and sovereign debt risks could be wild cards for 2010. We remain
confident about US sovereign risk in 2010, and expect any rise in fears about other G8
countries to help Treasuries.
Après, le deluge?
Rates should rise across 2010, Life should change for US rates investors in 2010, and not for the better. For more than a
driven by an end to Fed buying year, rates have been supported by two factors. One was uncertainty about the magnitude
and a stronger economy of an economic recovery and fears of a double dip recession. The other is the Fed’s asset
purchase programs, which have provided $1.7trn in net demand.
Both these pillars of support should weaken in 2010. We expect the strength of the recovery
to put to rest any lingering doubts of a double dip. Meanwhile, the Fed’s asset purchase
programs will stop in March, leaving a big hole on the demand side. Increased demand from
banks, households and foreign investors could help plug some of that hole, but not all of it.
Meanwhile, low realized inflation across 2010 should prevent inflation expectations from
coming unhinged and support nominal yields. Overall, rates should still trend higher across
the year, but less than many bond bears might expect – we are calling for the 10y to finish
2010 at 4.5%. The curve will likely stay steep for much of the year, before starting to flatten
in H2 as rate hikes appear imminent.
Housing and MBS-related On the plus side, the housing and securitized markets should fade as a source of systemic
systemic risk should fade, to be risk, though they may still command attention in 2010. But rates investors will have to start
replaced by regulatory and spending lots of time on other issues, such as regulatory and accounting changes, as well as
sovereign risks sovereign risk in some of the G8 economies. The end result – 2010 will likely not be a repeat
of 2009 by any means, but for rates investors, it could be just as challenging.
8 January 2010 4
US fixed income markets: Life without the Fed

An end to the Fed purchases Figure 1 highlights the role that the Fed played in bond markets in 2009. We track net fixed
leaves a big demand hole income issuance across all US fixed income products over the past few years (we look only
in US fixed income at term issuance, excluding T-bills, discount notes, etc.). Net issuance nearly tripled from
2008 to 2009, from $980bn to $2640bn. But remove Fed/Treasury purchases, and the
amount hitting investors’ screens was actually a third of that, or just ~$875bn. Net supply in
2010 should be almost as high as in 2009, at a total of $2600bn. But without a massive Fed
bid, all of that supply will need to be absorbed by investors. So from their standpoint, net
supply will nearly triple in 2010 (from$875bn in 2009 to nearly $2600bn in 2010).
Figure 1: Absence of Fed to be felt in 2010

2009 2010E
Term FI Supply, Net, $bn 2006 2007 2008 2009 ex Fed/Tsy 2010E ex Fed
Treasury, ex-bills 191 101 332 1,556 1,256 1,840 1,840

Municipal debt 135 156 1 31 31 115 115
Agency debt ex discount notes 85 -60 -44 -14 -174 100 85
Agency MBS 315 538 507 480 -820 350 300
Fixed rate corporate, IG+HY 178 309 454 918 918 545 545
Non-agency MBS+CMBS+ABS 798 441 -267 -332 -332 -250 -250
Total net term supply, $bn 1,702 1,485 983 2,639 879 2,700 2,635
Total 10y equivalents, $bn 1,956 2,135 1,905 3,260 2,359 3,370 3,300
Note: Treasury and agency debt exclude T-bills and agency discount notes, respectively. Corporate supply excludes any floating-rate supply but includes government-
guaranteed and supra/sovereign debt. Municipal debt includes BAB issuance. Source: Treasury, Flow of funds, EMBS, Barclays Capital
Scary as these supply numbers seem, there are some offsetting demand factors. US
households, US banks, and foreign investors should all step up this year. We walk through
the exact details of our demand forecasts in “Supply/demand: Supply deluge” on page 11.
For now, we turn to Figure 2, which shows that these three investor types could add as
much as $1600bn in 2010. US households should lead the way with $600bn in demand,
though most of that will show up through money managers and mutual funds. Despite this,
we still expect a hole of as much as $1trn between supply and demand in US fixed income.
Admittedly, our estimates are based on assumptions, both on the supply and especially on
the demand side. And they could well be off somewhat across the course of 2010. But it is
very unlikely that they will be off by enough to make up a $1trn hole. The conclusion is
inescapable: the supply-demand imbalance is heavily tilted towards higher rates in 2010.
Figure 2: Net fixed income supply should jump next year Figure 3: Rates forecast for 2010
Ex-Ante Demand Total Fed Funds 3m Libor 2y 5y 10y 30y 10y RY

1Q10 0.00-0.25 0.28 1.10 2.50 3.70 4.70 1.40
Overseas +600 2Q10 0.00-0.25 0.27 1.60 3.00 4.20 5.20 1.70
Banks +300 3Q10 0.50 0.75 2.00 3.40 4.50 5.50 1.90
4Q10 1.00 1.20 2.30 3.60 4.50 5.50 1.90
Households +600
Pension allocation +100
Total ex-ante demand 1,600
Total supply 2,640
Imbalance -1,040
Source: US Treasury, Flow of Funds, Barclays Capital Source: Barclays Capital
8 January 2010 5
Rate sell-off could take a That view is reflected in our rates forecasts in Figure 3. While we expect Q1 10 to be a
breather in Q1, before strong quarter, we also expect the market to take a breather from the December sell-off.
picking up in Q2 and Q3 Some investors will also probably be tempted to dismiss Q1 as a short-term bounce. But
continued strong growth in Q2, as per our forecast, should make them sit up and take
notice. Consequently, we have the bulk of the sell-off concentrated in Q2 and Q3. The fourth
quarter should herald the start of curve flattening in 2s10s as the Fed hikes. Even then, our
models expect the 2s10s curve to stay steeper than what the forwards are pricing in. Recent
statements from FOMC officials suggest that when the Fed does start, it could hike more
quickly than in the past. Our economists disagree – they expect the Fed to take a break from
hikes in the first half of 2011. Regardless, the 2y point typically reacts strongly to the start of
hike cycles; this time is unlikely to be different. 10s are a different matter – we expect the 10y
yield to stop rising in the fourth quarter as Fed hikes dampen inflation expectations.
Rates products: Swaps, volatility and the TIPS markets

How will our rates views play out in various rates products? Swap spreads are an easy call.
Across the world, 2010 should see massive amounts of sovereign debt issuance but
relatively little issuance of high grade spread products. Meanwhile, the US banking system
will likely continue to recapitalize itself, taking advantage of cheap funding and a steep
curve. We expect the swap spread curve to flatten as longer spreads tighten, as we discuss
in “Swaps: Weighed down by supply” on page 65.
Swap spreads should tighten, Meanwhile, rate volatility tends to be driven in the medium term by economic uncertainty,
while gamma will likely drop questions about Fed moves, and supply-demand dynamics. The first half of 2010 should see
a Fed on hold, an economic recovery that quashes fears of a double dip, and increased
volatility supply driven by higher yields. In this environment, gamma should drop across
tails. However, longer expiries could stay well bid, given demand from the MBS universe and
very little supply. We talk through our views on volatility in “Vol: Entering a quiet period” on
page 74.
Finally, the TIPS market could be very active in 2010, with interest even from crossover
buyers, as investors concerned about loose monetary policy look for protection. Although
outright TIPS returns may be weak as real rates rise, relative performance should be positive.
Structural allocations, the Treasury’s commitment, the potential for a US VAT, and food
inflation risks should be supportive, while the market reaction to rent and OER CPI may be a
drag; for details, please see “TIPS: Still cheap insurance” on page 42.
Agency debt and the GSEs: Don’t hold your breath

Investors expecting sweeping Unlike the derivatives and TIPS markets, agency debt is one asset class in which our rates
changes to the GSEs are likely to views will not be the primary driver. Instead, policy changes hold the key. As we had expected,
be disappointed in February the administration raised the amount of trust preferred capital upon which the GSEs can draw
down. But the major developments should be in February, when the administration is
supposed to present a plan about the future of the GSEs. Sweeping changes are unlikely, but
the government could lay out a timeline for conservatorship and lower the risk weight on GSE
debt. As for the longer term, there are several options: full nationalization, full privatization, or
hybrid models. Each has pros and cons, and we discussed these in detail in “GSEs: Back to the
future,” Securitized Products Weekly, December 11, 2009. But the history of the GSEs is one of
glacial evolution, with Congress often co-opting them for public policy purposes. This time
should be similar, with the legislative focus on health care and then financial regulatory
reform. The administration will probably preserve the status quo on the GSEs for a long time,
with the final changes emerging many years from now. So for those waiting for the resolution
to the GSE issue, all we can say is: don’t hold your breath.
8 January 2010 6
Agency debt spreads A more immediate concern is finding sources of demand for agency debt to replace the exiting
could widen 15-20 bp Fed bid. The changes in the PSPAs gave the GSEs room to grow their portfolios, but we think
actual growth is unlikely. Rather, the goal was to make sure that they did not have to be active
sellers to meet their previous portfolio targets. We do expect the GSEs to continue terming out
liabilities to reduce rollover risk. Our estimates call for FNM/FRE to term out roughly 3% of
total liabilities in 2010. We look for a nearly 5% term out at FHLB. In total, we expect this to
lead to about $100bn of net long-term debt issuance at the three GSEs, with about $25bn
each at FNM and FRE and $50bn at FHLB. On the demand side, an end to Fed buying will hurt,
though money managers and banks could pick up some of the slack.
US securitized markets: Stepping back from the limelight

The end of the Fed purchase Talk of the GSEs always leads into the US mortgage markets, which have been a big focus
program should push agency for investors worldwide for the past two years. That should change in 2010 – in our opinion,
MBS spreads wider by 25-30bp the systemic risk posed by securitized markets is now lower. This includes agency MBS,
where the Fed has bought nearly $.125trn. Understandably, investors are concerned about
how agency MBS will manage without the Fed. We do expect the end of Fed buying to
pressure spreads wider. But agency MBS spreads have been tight for the past several
months; even if they widen 30bp (as we expect), that should bring them near longer-term
averages. And at these wider spreads, we fully expect buyers. To understand why, consider
the demand-supply dynamics in agency MBS in 2009. The Fed bought $1.1trn, but net
agency MBS issuance was less than $400bn. That means someone sold the Fed the other
$700bn. The big questions are: Who were these sellers, and will they come back to buy
when spreads widen? We think the answer to the second is yes.
As Figure 4 shows, most of the selling was from unleveraged accounts such as money
managers and mutual funds. In fact, money managers, mutual funds and insurance
companies together sold over $300bn in the first six months, or an annualized $600bn. And
many of these investors are benchmarked to the Barclays Capital indices. Consequently,
most top fixed income money managers are now sharply underweight MBS against their
benchmark. But as spreads widen with the Fed’s departure, demand should come back as
index investors move closer to market weight. Our verdict: spreads should widen 30bp
when the Fed walks away before being met with demand from index money and the
banking system. For details on why we expect a 30bp widening, please see “Mortgage basis
outlook: Life after Fed,” Securitized Products Outlook 2010.
Figure 4: Money managers and mutual funds were sellers Figure 5: 60d+ have stabilized for 2003 sub-prime
EOY 2008 9-Jun Diff 60+ (%

Factor
Govt. (Fed+ TSY) 71 612 541 Curbal)
120% 18
GSEs (FN, FH & FHLB) 901 1016 115
Banks (Commercial, 16
100%
S&L, & CU) 1030 1122 92 14
Foreign 645 591 -54 80% 12
Mutual Funds 534 400 -134
Factor 10
Pension + Retirement Funds 564 494 -70 60%
60+ (% Current Balance) 8
Insurance (Life, P & C) 383 304 -78
40% 6
Broker Dealers 152 104 -48
REIT 91 103 12 4
20%
SIV 346 307 -39 2
Households * 201 107 -94 0% 0
Total** 4917 5160 243 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71
Source: Barclays Capital Source: Loan Performance, Barclays Capital
8 January 2010 7
Tail risk has reduced in On the non-agency front, the picture is mixed. There was no primary origination of non-
non-agency MBS losses agency MBS in 2009, but there are finally signs of credit burn-out in legacy non-agency
mortgages. The rate at which current borrowers turn delinquent has stabilized in recent
months in sub-prime, as the weaker borrowers have continued to default and been “burnt
out.” For example, Figure 5 shows that 60-day delinquencies for 2003 vintage sub-prime
loans have stabilized for the last several months. In addition, loss severities are improving;
some servicers are turning to short sales to avoid liquidation costs. Admittedly, our base
case loss forecasts have not improved much; there are still losses to come in non-agency
MBS. But there are signs that both delinquencies and severities have peaked. Unfortunately,
the same cannot be said for CMBS and the commercial real estate (CRE) sector.
CRE will likely be a Unlike residential mortgages, commercial mortgages lag an economic recovery. As Figure 6
source of headlines next shows, demand for CRE should stop shrinking only in the second half of 2010, once y/y
year as losses pick up payroll growth turns positive. And the Fed is worried enough about this sector that it is the
only asset for which it has extended TALF into 2010. Despite this, CMBS was one of the best
performing sectors in the Barclays Aggregate Index in 2009, largely because tail risk (with
CMBS priced to depression-like scenarios) has dissipated. But lower tail risk is probably the
only good news for CRE for the next several quarters.
Home prices and bank losses: Tail risk has declined

We expect further declines in Any non-agency loss projections depend on the home price forecast. Thankfully, tail risk has
home prices, but not enough to also declined in US housing. US home prices, as measured by the Case-Shiller index, have
have a sharp macro effect stabilized over the past few months. But prices could start dropping again soon. While
negative seasonals will probably be the immediate driver (for details, see ‘Housing: Seasonal
HPA biases’, Securitized Products Weekly, July 31 2009), the overhang of foreclosed
inventory could play a role (Figure 7). On the positive side, housing is being helped by an
improving economy, the homeowner tax credit, and better affordability. And unexpectedly,
home prices have been helped by mortgage modification programs such as HAMP.
While many of these modifications might ultimately not work, the process of modification
buys time. It increases the number of months between the borrower turning delinquent and
the home hitting the market. This is shown in the real estate owned (REO) line in Figure 7;
even as foreclosures keep rising, this bucket has gone down. Intuitively, if there are millions
of foreclosed sales to come, it is better to spread them over a few years than a few months –
this prevents prices from over-correcting to the downside. Our forecast calls for prices to
Figure 6: CRE lags an economic recovery Figure 7: Foreclosures have not translated into supply, yet
Sq ft (millions) YOY % Chg in NFP Homes (000s)
300 forecast 4% 3,000
3% 2,500
200
2% 2,000
100
1%
1,500
0 0%
OFF
HOT -1% 1,000
-100
RET
IND -2% 500
-200
APT -3%
YOY % Chg in Payrolls (rhs) -
-300
-4% Jan-05 Jan-06 Jan-07 Jan-08 Jan-09
-400 -5%
90 91 93 94 95 96 98 99 00 01 03 04 05 06 08 09 10 11 Real-estate Owned Foreclosure
Source: BLS, Barclays Capital Source: Loan Performance, Barclays Capital
8 January 2010 8
drop 8% before stabilizing in Q2 10. The macro effect of this decline should be muted. After
all, a house worth $100 at its peak is now worth $67 (prices have fallen about 33% from the
peak in Case-Shiller). A further 8% decline would simply be another $5.3. As every month
passes without a sharp increase in the REO bucket or a sharp drop in prices, the tail risk
posed by housing declines a little further.
Our home price and securitized loss forecasts immediately raise questions about the health
of US banks, which will shoulder many of these losses. The US banking system has over
$13trn in balance sheet. But future losses will come mainly from the $7trn+ loan book. Figure 8
is our forecast for future losses by loan type; it shows that banks have another $850bn+ in
losses. These should be spread over the next few years, at $250-300bn in charge-offs per year.
Any such loss estimate involves many assumptions. But look back at 2009, and our numbers
seem to fit. The banking sector has taken $230bn in losses in Q1-Q3 09, or about $300bn
annualized, very much the pace we forecast at the start of the year, in “Overview: Bad Bank
Blues,” Market Strategy Americas, January 30, 2009.
Figure 8: Bank losses over the next few years ($bn)

Assets ($bn) Proj. Cum Loss(%) Loss 2010 2011 2012
1st lien mortgages 1741 11% 200 75 63 63

2nd lien mortgages & HELOCs 855 21% 178 79 49 49
Multifamily Residential 216 6% 13 4 4 4
CRE mortgages 1090 9% 101 34 34 34
Construction loans 492 40% 195 98 98 0
C&I 1276 7% 85 28 28 28
Cards 393 13% 52 22 15 15
Others 1352 4% 60 20 20 20
Total 7415
Source: FDIC, Barclays Capital
US banks have considerable So is there systemic risk to US banks due to future losses? We think not, assuming 2009 is any
losses to take, but have the indication. Banks have taken $300bn annualized in charge-offs this year yet still reported a
ability to earn their way out small profit. Assuming the funds-10s curve stays steep for another 18 months (we expect it
will), banks should earn their way out of this hole. And the hole is not $850bn, since banks
now have $220bn in loss provisions built up. So if the banking sector can earn $600bn+ over
the next few years, they should be able to absorb losses without eating into equity capital. In
addition, we expect considerable dispersion in losses across banks, with regional banks hit
harder than big money center banks (where systemic risk truly resides). Hence, while losses
might pose challenges to the equity side, they do not appear big enough to be systemic.
Regulatory changes and other wild cards

Liquidity portfolios could push up A big focus for investors across 2010 will be to gauge the effect of regulatory changes.
demand for US fixed income, but Banks being forced to raise liquidity portfolios is one example. The FSA in the UK now
there will likely be a requires British banks to be able to withstand three months of market stress. Unlike in the UK,
phase-in period details in the US and elsewhere are yet to come out. But the FSA rules suggest that banks
might have to increase their holdings of liquid and high-grade assets sharply, though the
funding mix will play a role (the lower the wholesale funding, the lower the liquidity portfolio
needed). The UK does not include agency MBS or debt as liquidity instruments, but we expect
US banking regulators to do so. There will probably be a phase-in period (it is four years in the
UK). But as and when the US rules become clearer, banks should start to position for the new
liquidity regime. One result could be a stronger bid for agency MBS (“Overview,” Market
Strategy Americas, October 22, 2009), but we expect this only from 2011.
8 January 2010 9
Changes in capital requirements and accounting treatment are also expected. H.R. 4173, the
Wall Street Reform and Consumer Protection Act of 2009, is working its way through
Congress right now. Meanwhile, SFAS 166/167 will go into effect next year. The immediate
effect is related to whether securitized trusts need to be moved on balance sheets or not.
The GSEs plan to consolidate their guarantee books in 2010. But this should not materially
affect agency MBS issuance, because GSE regulatory capital requirements have already
been waived. Away from the GSEs, other securitized assets might get affected. But the
language is still in flux. And accounting interpretations will play a big role. For example, it
might seem that banks will have to consolidate all RMBS securitizations. But this can be
accommodated by aggregating a deal with loans from several originators. If no one services
a majority of the deal (since there are many servicers), no firm might have to consolidate.
Credit risk could come back as Meanwhile, a risk to our base case is credit risk. Sovereign credit should be an important topic
an issue in 2010 – on the in 2010. The recent example of Greece, as well as the CDS widening in several developed
sovereign front countries, shows this issue coming to the forefront. Ironically, with the developing world a net
exporter of capital to the developed world, sovereign risks now seem concentrated in some big
developed economies. How will this affect US assets? We believe that increased fears about
G8 sovereign risk will help US Treasuries, which could widen spreads. Clearly, we do not expect
serious concerns about US credit risk in 2010. Our analysis shows that a big driver of ratings
tends to be the share of a country’s currency in total world reserves. By that metric, the US has
a long way to go before other factors (financial stabilization cost, debt/GDP ratios, etc.) drag
down its credit rating (for details, see “How risk-free are US Treasuries?” on page 84). The big-
is-better argument also holds true for US banks. Just as the US rating benefits from the size of
the US economy, so do those of big banks. Consequently, the rating agencies have started
making noises about clauses in H.R.4173 that aim to minimize the effect of failing financial
institutions on the taxpayer. S&P put out a note on December 16, 2009, warning that the bill in
its current form might make it take a second look at the existing ratings of big banks. In other
words, if banks are no longer too big to fail, their credit rating could be negatively affected. We
expect several iterations and discussions between rating agencies and lawmakers before any
language becomes law. But this increases the risk of downgrades to some banks, which
investors should monitor closely.
All these changes could affect the demand for US rates in various ways. Increased sovereign
risk fears in parts of Europe could help the USD and Treasuries, as would any bank
downgrades, while stringent liquidity portfolios might push the bank bid firmly into agency
MBS. Rates investors in 2010 might no longer have to monitor where the next sub-prime
blow-up is coming from. But this will probably be replaced by increased focus on changes in
the regulatory landscape and sovereign risk from parts of the developed world.
8 January 2010 10
US FIXED INCOME SUPPLY/DEMAND
Supply deluge
Anshul Pradhan We discuss the outlook for the supply-demand imbalance in fixed income markets. The
212-412-3681 sharp upward swing in term fixed income supply that the market needs to absorb from
anshul.pradhan@barcap.com 2009 to 2010 should overwhelm the expected increase in private demand. In addition,
the expected demand shortfall is concentrated in Treasuries. We maintain our curve
Amrut Nashikkar steepener view and expect long-end swap spreads to drift tighter through 2010.
+1 212 412 1848
The sharp increase in supply of term fixed income securities in 2009 was offset by the
amrut.nashikkar@barcap.com
Fed’s asset purchase program. The amount of supply hitting the market was actually
less than in prior years (Figure 1).
In 2010, net term supply of fixed income securities is likely to stay as high as in 2009.
But without the Fed bid, the market will have to absorb close to $2.6trn in notional
terms; $1.7trn more than in 2009. 2010 net term notional supply of $2.6trn and
duration supply of $3.3trn in 10y equivalents are well above even pre-crisis levels.
Private demand is likely to pick up, but not enough to fill the void created by the Fed’s
stepping away. We expect overseas investors and households to continue to deploy
foreign exchange reserves and savings respectively. Banks will likely continue to switch
from loans to securities, but we do not expect overall bank credit to increase.
Terming out of short-term holdings, bills and money market shares, should provide a
boost to the pace of purchase of term securities. Demand for term debt still seems to be
falling short of net supply by close to $1trn.
Demand appears to be skewed toward spread products, whereas supply will still be
dominated by Treasuries. In addition, demand appears to be skewed toward the front to
intermediate sector. We recommend forward starting curve steepeners and swap
spreads tighteners in the long end.
Supply-demand imbalances in the fixed income universe have never attracted as much
attention as in 2009. In 2010, the debate is likely to intensify. In the following pages, we
revisit the dynamics of 2009 and discuss our 2010 outlook for the supply and demand of
fixed income securities for issuers and investors, respectively.
Figure 1: Absence of Fed to be felt in 2010

2009 2010E
Term FI Supply, Net, $bn 2006 2007 2008 2009 ex Fed/Tsy 2010E ex Fed
Treasury, ex-bills 191 101 332 1,556 1,256 1,840 1,840

Municipal debt 135 156 1 31 31 115 115
Agency debt ex discount notes 85 -60 -44 -14 -174 100 85
Agency MBS 315 538 507 480 -820 350 300
Fixed rate corporate, IG+HY 178 309 454 918 918 545 545
Non-agency MBS+CMBS+ABS 798 441 -267 -332 -332 -250 -250
Total net term supply, $bn 1,702 1,485 983 2,639 879 2,700 2,635
Total 10y equivalents, $bn 1,956 2,135 1,905 3,260 2,359 3,370 3,300
Note: Treasury and agency debt exclude T-bills and agency discount notes, respectively. Corporate supply excludes any floating-rate supply but includes government-
guaranteed and supra/sovereign debt. Municipal debt includes BAB issuance. Source: Treasury, Federal Reserve, EMBS, Barclays Capital
8 January 2010 11
Treasury supply: As far as the eye can see

The Treasury should continue The Treasury has come a long way in altering the auction calendar. Over the past two years,
gradually to term out debt auction sizes have increased across the curve, although more so in the front to intermediate
sector, especially with the introduction of 3s and 7s. The frequency and sizes of long-end
auctions have also increased, but at a more gradual pace. However, the current average
maturity of outstanding debt of 54 months is still far from the Treasury’s target of 72-84
months. The Treasury has stated that it intends to extend gradually, so we expect only a
marginal increase in auction sizes in the intermediate to long sectors (Figure 2). Front-end
auction sizes should stabilize at current levels before declining toward the end of the year.
Treasury issuance should As a result, gross and net coupon issuance in 2010 should be $2.5trn (up $400bn from
increase to $2.5trn and 2009) and $1.8 trn (up $300bn from 2009), respectively. In 10y equivalents, coupon supply
$1.6trn in gross and 10y should increase to $1.6trn, from $1.1trn in 2009, adjusted for $300bn of Fed purchases. The
equivalent terms bill universe will likely continue to shrink, as we expect financing needs to be close to
$1.4trn (assuming Supplementary Financing Program (SFP)-related borrowing of $300bn,
in addition to our economists’ deficit projection of ~$1.1trn). Were deficits to surprise to the
upside (OMB and CBO latest estimates for 2010 deficits are $1.5trn and $1.4trn,
respectively), bills would likely shrink less, but more important, auction sizes should stabilize
at higher levels, in the intermediate to long sector, than we have pencilled in.
Duration supply should remain Beyond 2010, we expect average maturity to increase gradually, stabilizing in the middle of
close to $1.5trn 10y equivalents the range over the next 5-10 years. While this should help avoid market disruptions, it also
in further out years means that duration supply from Treasuries is likely to remain high. Figure 3 plots our
expectations for gross and 10y equivalent supply of coupon Treasuries for the next ten
years assuming that deficits decline gradually as our economics team expects and the
Treasury continues to term out debt. 2010 should mark the peak of gross coupon issuance at
~$2.5trn and 2013 the trough at $2trn. Gross issuance is unlikely to decline further, as the
amount of debt maturing will be increasing rapidly. Duration issuance should decline even less
because of a gradual terming out and remain at $1.5-1.6trn in 10y equivalent terms over the
Figure 2: 2010 Treasury issuance calendar

2y 3y 5y 7y 10y 30y 5y I/L 10y I/L 30y I/L Total Maturing Net 10yEq
January 44 40 43 33 21 13 10 204 53 151 128

February 44 40 44 34 26 17 9 214 77 137 149
March 44 40 45 35 22 14 200 43 157 126
April 44 40 45 35 22 14 10 8 218 77 141 137
May 44 40 45 35 27 18 209 61 148 140
June 44 40 45 35 23 15 202 44 158 129
July 44 40 45 35 23 15 11 213 44 169 138
August 44 40 45 35 27 18 8 217 63 154 152
September 44 40 45 35 23 15 9 211 47 164 136
October 43 39 45 35 23 15 8 208 47 161 132
November 42 38 44 34 27 18 9 212 53 159 144
December 41 37 44 34 23 15 194 51 143 126
Total 2010 522 474 535 415 287 187 18 47 17 2,502 660 1,842 1,637
Total 2009 499 430 447 306 242 129 15 29 14 2,109 553 1,556 1,070*
% Increase 2009 to 2010 5% 10% 20% 36% 19% 45% 20% 62% 21% 19% 19% 18% 53%
* After adjusting for Fed purchases. Source: Barclays Capital
8 January 2010 12
next several years. Were deficits to follow the path laid out in the CBO baseline (adjusted for
alternative policy measures), overall coupon issuance would stay close to $2.5trn in notional
terms and at a little less than $2trn in 10y equivalent terms for the next several years. While
this may seem extreme, the main takeaway is that Treasury supply is likely to remain an
issue for several years (see the Treasury outlook for details).
Spread products: A sharp swing

Spread product supply to As was evident in 2009, Treasury supply alone is insufficient to determine the outlook for
increase by $1.2trn from 2009 to longer-term rates; one needs to account for supply of spread products as well. Spread
2010 accounting for product supply picked up in 2009 relative to 2008, led mainly by the corporate sector, but
Fed/Treasury purchases was more than offset by the Fed purchases of agency debt and MBS (Figures 1 and 4). We
in both years expect spread product supply in 2010 to decline from the pre-Fed-buying levels of 2009, led
mainly by a reduction in corporate debt, but to increase by $1.2trn from post-Fed-buying
levels. We discuss the details below.
Corporate debt: Corporate supply picked up sharply in 2009 given low interest rates, the
unfreezing of capital markets, and banks’ unwillingness to make loans. Financial issuance
increased, led by the government-guaranteed program and supra/sovereign issuance of
dollar-denominated debt increased, given attractive funding levels in the US market. Of the
$1.1trn in net issuance of spread products before accounting for Fed purchases, corporate
issuance accounted for ~$920bn. Our credit strategists expect issuance to decline by
$350-400bn amid a rebound in corporate profits, rising rates, and elevated cash balances.
Municipal debt: Municipal issuance remained at roughly 2008 levels; however, the
composition shifted toward fixed-rate supply, led by a pickup in Build America Bonds
(BABs). Our municipal strategists expect issuance to pick up in 2010 given strong
borrowing needs and access to the taxable market. In particular, BAB issuance should
increase to $140bn, accounting for most of the ~$80bn increase in net supply from
2009 to 2010 (Figure 1).
Agency debt: While the mortgage portfolios of GSEs should remain unchanged, as it
makes little economic sense for them to grow given current spreads, GSEs should
continue to term out debt and reduce rollover risk by letting discount notes roll off. We
expect $100bn in net term issuance in 2010, which should be offset partly by the $15bn
the Fed has left to buy in Q1 10. This should correspond to a +$260bn swing in the net
Figure 3: Treasury supply: With us forever Figure 4: Spread products: A sharp swing
3,000 40% 2,000 10y Eq

1,500
2,500 35% 1,000
500
2,000 30%
0
1,500 25% -500
-1,000
1,000 20% -1,500
2006 2007 2008 2009E 2009E- 2010E 2010E-
500 15% ex ex Fed
2009 2011 2013 2015 2017 2019 Fed/Tsy
Gross Coupon Supply, $bn Municipal Debt Agency Debt-ex discos
Coupon Supply, 10-yr Eq, $bn Agency MBS Fixed Rate Corporate
% Debt Maturity in < 1yr, RHS Non-Agency MBS+CMBS+ABS
Notes: TIPS included in respective sectors. Source: Treasury, Barclays Capital Source: Flow of Funds, EMBS, Barclays Capital
8 January 2010 13
supply of term agency debt that the market needs to absorb from 2009 to 2010, as net
supply adjusted for Fed purchases was -$175bn in 2009.
Securitized: Securitized issuance was dominated by agency MBS, as net issuance of

non-agency debt remained significantly negative and CMBS and ABS issuance was
slightly negative. Our mortgage strategy team expects net issuance of agency MBS to
decline from $480bn in 2009 to $350bn in 2010, which should be offset further by net
fed purchases of $50bn in 2010 (+$150bn - $100bn paydowns). Net issuance in other
securitized products should remain negative. Overall, after adjusting for Fed and
Treasury purchases in both years, the amount of net securitized issuance that the
market needs to absorb should increase by ~$1.2trn from 2009 to 2010.
Figure 4 shows that the supply of spread products should rise sharply from the post-Fed-
buying levels of 2009. It will likely remain below pre-crisis levels in net terms but around par
in 10y equivalent terms. This is because a large share of issuance in earlier years was in
floating-rate non-agency MBS; this has been replaced by corporate debt, which has a much
longer duration; agency MBS, which has been mostly fixed; and a steady increase in the
fixed-rate component of municipal issuance. Fixed income Issuance in 10y equivalents,
given the increase in Treasury supply, should therefore be higher than pre-crisis levels.
Market needs to absorb $2.6trn Overall, as Figure 1 highlights, we expect net issuance of term debt to increase to $2.6trn in
in net term fixed income supply 2010 from $0.9trn in 2009 after accounting for Fed/Treasury purchases in both years. Net
in 2010 and $3.3trn in 10y issuance of total fixed income securities is likely to be lower, as we expect negative net
equivalent terms; both much issuance in short-term products, such as agency discount notes and T-bills, and floating-
higher than pre-crisis levels rate corporate debt. However, net issuance of such debt was negative in 2009 as well; for
instance, agency discount note issuance was -$440bn, and we expect T-bill issuance to be
similarly negative in 2010. Hence, this has little effect on the swing in net issuance from
2009 to 2010. More important, fluctuations in short-term debt should have little effect on
duration supply, which is what should matter for longer-term rates (and is why it was not
included in Figure 1). In 10y equivalent terms, the market would still need to absorb
~$3.3trn in 2010, much more than pre-crisis levels of $2trn.
Private demand: Not enough to fill the void

Foreign central banks, With the Fed taking out a major chunk of supply, it is not hard to envisage private demand
households, and commercial for less than $1trn in net issuance in the face of the high risk aversion and uncertainty about
banks absorbed most of the the economic outlook. Foreign investors, households, and the banking sector absorbed
supply in 2009 most of the supply in 2009. While private foreign investors were net sellers, mainly of spread
products, official institutions more than made up for them by buying Treasuries.
Households, after paying down their liabilities, deployed the bulk of their savings in fixed
income securities, mostly through mutual funds. The banking sector switched from loans to
securities, as is typical in any recession.
The upward swing in private We expect these investors to continue to buy and to step up the pace of purchases, but not
demand should fall short of that enough to fill the void created by the Fed’s stepping away. We expect overseas investors
in term fixed income supply and households to continue to deploy foreign exchange reserves and savings, respectively.
The terming out of short-term holdings, bills, and money market shares should provide a
boost to the pace of purchases. Banks should continue to switch from loans to securities,
but we do not expect overall bank credit to increase. We estimate that the demand shortfall
for term debt is still close to $1trn in notional terms, mostly in Treasuries. We discuss the
expected purchases from each of these investor classes below.
8 January 2010 14
Foreign demand: FX reserve growth aided by terming out

We expect $600bn in overseas Demand for US securities from foreign central banks has remained robust over the past
demand as the increase in several years, although unsurprisingly, the composition has shifted to Treasuries in the past
foreign exchange reserves and year (Figure 5). In late 2008, official investors moved out of agency securities, in response to
terming out of bill holdings the crisis, and into T-bills, accounting for almost all Treasury purchases. Risk aversion has
offsets the diversification certainly declined, as the composition has shifted from bills to coupon Treasuries. In the
trend that is under way third quarter, official investors bought $100bn in coupon Treasuries and only $25bn in T-
bills, and recent TIC data show that bill holdings have stabilized. We expect the trend of
terming out to gain a foothold in 2010, which, coupled with the deployment of foreign
exchange reserve growth in USD assets, should contribute significantly toward absorbing
fixed income supply. We estimate ~$600bn in coupon Treasury and agency securities over
the next one year, with the bulk in Treasuries. Private investors have been net sellers even as
recently as Q3 09, a continuation of which may offset some of the demand from officials.
FX-reserves should grow by Figure 6 shows that FX reserves, adjusted for currency movements, have begun to grow
$1trn in 2010 as global growth after remaining flat over the 1-year period ending March 2009 (actually declining in
resumes, but $400-450bn is to quarterly terms in Q4 08 and Q1 09). While global growth has picked up marginally, the
be invested in USD assets as the bigger contributor has been the depreciation of the dollar. Our currency strategists expect
diversification trend continues the dollar to strengthen, but the pickup in global growth should continue to result in higher
reserves. We estimate that reserves should grow by 15% in 2010, or $1trn in notional terms,
were the dollar to remain at current levels, or $0.8trn and $1.25trn were the dollar to
strengthen by 10% or weaken by 10%, respectively. How much of this will be invested in
USD assets? Historically, the figure has been 60-70%, but recently, foreign central banks
have been diversifying away from the dollar. Figure 7 shows the percentage of reserves held
in dollars both as reported and adjusted for currency movements. While the reported share
has remained unchanged between June 2008 and June 2009 (latest available), the adjusted
share has gradually declined. Assuming that the same pace of diversification continues,
~$400-450bn should be invested in US dollar assets (Figure 8).
Risks are to both sides of these estimates, a weaker dollar accompanied by a reduction in
the pace of diversification should lead to larger investments in dollars and vice versa. In
particular, if the USD depreciates 10% and the pace of diversification is halved, presumably
as risk aversion returns, foreign investors could buy up to $725bn; at the other extreme, if
the USD appreciates 10% and the pace of diversification doubles, they may buy only $90bn.
Figure 5: Overseas official demand for Treasuries robust Figure 6: Reserve growth to resume
200 30%
150 125
25%
100 20% $1.25tr
50
15% $1.0tr
0
10% $0.8tr
-50
5% 5%
-100
-150 0%
Q4-08 Crisis
-200 -5%
Sep-07 Mar-08 Sep-08 Mar-09 Sep-09 Dec-00 Dec-02 Dec-04 Dec-06 Dec-08 Dec-10
Qtrly Tsy Purchase, $bn of which, T-Bills, $bn Reserve Growth, yoy Estimated
Qtrly Agency Purchase, $bn Weaker USD Stronger USD
Source: Federal Reserve Flow of Funds, Treasury Source: IMF, Barclays Capital, Haver Analytics
8 January 2010 15
Terming out of bill holdings We believe the former is unlikely, as foreign central banks were diversifying even at the peak
should add another $150bn to of the crisis and our currency strategists expect the dollar to appreciate. While the latter
potential demand from estimate seems extreme, because non-US demand of ~$700bn would be tough to satisfy, it
overseas investors highlights that overseas demand is also susceptible to downward surprises. Another source
of foreign demand for long-term securities is terming out from bill holdings. As discussed
earlier, official investors mainly bought T-bills in late 2008 and early 2009. This is slowly
beginning to reverse, but the share of bills in the Treasury portfolios of officials, at 25%, is
still much higher than the pre-crisis average (Figure 9). A reversal to pre-crisis levels would
generate $150bn in demand over 2010, in addition to $400-450bn from reserve growth.
The bulk is to be invested in front How will that be invested? The bulk of the investment is likely to be in Treasuries, as has
to intermediate Treasuries, been the case over the past year; however, custody holdings of agency securities have
although some may find its way started to inch higher over the past month, and some may find their way into agency MBS.
into agency MBS The annual survey data from the pre-crisis period show that agency MBS accounted for
20% to 35% of the total increase in securities held by officials from June 2006 to June 2008.
Within Treasuries, the bulk is likely to go into the front to intermediate sector; the annual
survey of holdings indicates that overseas investors tend to hold mostly in the <5y sector,
and recent auction statistics confirm that most of the demand is in that sector.
Figure 7: Diversification under way Figure 8: $400-450bn to be invested in USD
74% Pace of diversification

vs last year
72% USD inflows, Reserve growth,
$bn $bn Half Same Double
70%
+10% 800 355 275 90
68%
USD 0% 1,000 525 435 260
66% -10% 1,250 725 630 440
64%
62%
60%
Sep-99 Sep-01 Sep-03 Sep-05 Sep-07 Sep-09
% Reserves Held in $, Reported
Currency Adjusted
Source: IMF, Barclays Capital Source: Barclays Capital
Figure 9: Expect terming out to continue Figure 10: Banks security purchase breakdown, YTD
35% Other Treasuries

Securities $54 , 15%
30% $79 , 21%
25%
Agency
20% Munis Debt
$13 , 4% $43 , 12%
15%
10%
Corporate Agency MBS
5%
$47 , 13% $56 , 15%
Jul-04 Jul-05 Jul-06 Jul-07 Jul-08 Jul-09
as a % of Tsy Holdings, Pvt,RHS Foreign
as a % of Tsy Holdings, Off ,RHS $71 , 20%
Source: Treasury Source: FDIC
8 January 2010 16
Banks: Safety of liquid securities

Banks should continue to switch The banking sector helped absorb fixed income supply in 2009 as it deployed cash from
from loans to securities and to loan runoffs to purchase securities after paying down short-term liabilities. In 2009 up to
reduce short-term liabilities; we September (latest data available), bank securities portfolios grew by $360bn (some of the
expect $300bn in demand for growth was due to an increase in market value), of which only $50bn was invested in
fixed income securities Treasuries, with the rest in higher yielding securities; mostly agency debt/MBS (including
CMOs) and foreign and domestic government-guaranteed debt. Interestingly, not all of the
difference between deposit growth and loan growth (reduction in this case) was deployed
in securities; banks allocated 60-65% to securities and used the rest to pay down short-
term liabilities, such as FHLB advances. We expect these trends to remain in place in 2010,
with bank credit to remain flat and net total purchases of $300bn, mainly in high grade
spread products.
Historical experience and recent Bank loan portfolios should continue to shrink, although at a slightly slower pace. Figure 11
lending conditions argue for a shows that banks continue to favor securities over loans even after recessions have ended.
further shrinkage of $300bn in For instance, in the early 1990s, the share of loans in bank credit declined by an average of
2010 in the loans portfolio 2% per year for two years after the recession ended. Current conditions argue for the same;
we estimate that given the continued tightening of credit conditions by banks and weak
final demand as proxied by ISM and consumer confidence, the share of loans in bank credit
should shrink by another 3 percentage points, or ~$300bn. This should be aided by deposit
growth as individuals move out of money market funds, partly into deposits.
Banks are likely to prefer the How is this likely to be invested? Potential liquidity guidelines would encourage banks to
safety of liquid securities and the own liquid securities, but the need to own assets in unencumbered form could prove to be a
carry of agency securities over limiting factor. Issuing, say, 5y term debt and buying 5y Treasuries would actually be more
Treasuries and cash of a drag on net interest margins (NIMs) than raising deposits and keeping the money in
cash. Banks would therefore be inclined to buy agency securities, which would be allowed in
liquid portfolios, but again, they would be limited by new supply. We expect net agency
MBS issuance of $300bn and agency term debt issuance of $85bn after adjusting for Fed
purchases. Only if banks were allowed to take a higher duration gap, which seems unlikely,
would they move away from cash. Hence, the cash sitting on individual bank balance sheets
is likely to go toward paying down liabilities, such as FHLB advances or other short-term
borrowings, which would also help reduce liquidity needs. We illustrate this in Figure 13 by
optimizing bank balance sheets to maximize net interest margins while ensuring that liquid
Figure 11: Loans to continue to shrink Figure 12: Household savings rate to stabilize
78% 16
14
76%
12
74% 10
72% 8
6 6.1
70% 4
68% 2
0
66%
-2
64% -4
Sep-73 Sep-79 Sep-85 Sep-91 Sep-97 Sep-03 Sep-09 Sep-56 Sep-65 Sep-74 Sep-83 Sep-92 Sep-01
Recessions Loans / Bank Credit, % FOF yearly savings Rate, % Estimated
Source: Federal Reserve Source: Federal Reserve, Barclays Capital
8 January 2010 17
assets match liquidity needs. At current leverage, banks should pay down their liabilities,
(mostly short term volatile liabilities) and prefer agency MBS and cash to Treasuries in their
liquid portfolios to maintain their NIMs. The preference for cash should vary with allowed
leverage; higher leverage would mean more term outs and vice versa. We believe a more
plausible scenario for terming out is a steeper curve, which would increase the
attractiveness of Treasuries. Finally, cash holdings at banks should begin to decline as the
Fed begins to drain reserves starting in the second half of 2010. Hence, we don’t expect
banks to step up purchases in any significant manner in 2010. (see “The Potential Effect of
Bank Liquidity Regulations on the Fixed Income Landscape,” October 16, 2009).
Figure 13: Treasuries are not the preferred asset class for banks
Unchanged
Unchanged Lower Higher Duration Gap
Duration Gap Duration Gap Duration Gap steeper curve
Assets
Reserves
Cash and cash eq. ↑↑ ↑↑↑ ↓↓↓ ↓
Tsy/agency debt ↓ ↓ ↑↑↑ ↑↑
Agency MBS ↑↑ ↑ ↑↑↑ ↑↑↑
Other securities ↓ ↓ ↓ ↓
Loans ↓↓ ↓↓ ↓↓ ↓↓↓
Other assets
Liabilities
Deposits ↑↑↑ ↑ ↑↑↑ ↑↑↑
FF/repo ↓↓ ↓↓ ↓↓ ↓↓
Advances ↓ ↓ ↓ ↓
Unsecured debt ↓ ↑↑ ↓ ↑
Other liabilities/equity - - - -
Net interest margin ↓ ↓↓ ↑ -
Source: Barclays Capital
Individuals: High savings aided by terming out

Households should continue to The household category absorbed a fair share of supply, Treasuries in particular, in 2009,
deploy savings into the term and we expect a pickup in demand in 2010 as the savings rate remains high and households
fixed income universe, which term out of their bills and money market holdings. In aggregate, we estimate that $600 will
should be aided by a terming out be available to invest in fixed income securities, directly and indirectly. The asset
of short-term investments. composition is likely to change sharply, as almost all Treasury buying in 2009 was a result of
allocation out of agency securities, which has run its course. Our model suggests a
preference for spread products.
The savings rate is likely to Figure 12 shows that the savings rate measured from the flow of funds data, which had
stabilize at current levels, with risen to 10%, has gradually declined to 6.5% over the past year. We expect the savings rate
the risk of a decline as asset to stabilize at current levels, and it may even decline as household net worth continues to
prices continue to increase rise. Figure 12 also shows our estimate of the savings rate using the net worth-to-income
ratio and the level of rates; the higher the net worth, the lower the need to save, and the
lower the level of rates, the lower the incentive to save. Current conditions argue that the
savings rate should be 6%. How it will evolve depends largely on the outlook for asset
prices; were home prices to stabilize and equity prices to rise by 10% over the next year, the
saving rate should decline to 5.5% but were home and equity prices to decline by 20%
8 January 2010 18
each, the savings rate could increase to 8.5%. Assuming 5.5-6.0% for the savings rate,
$600bn should be available for the net acquisition of financial assets.
Terming out from money market Another source of demand for longer-term securities is the cash held in money market
investments and bill holdings funds or T-bills as households term out. Households, directly and through life and pension
should continue as the curve accounts, increased their holdings of money market shares sharply, to $1.95trn by the end
stays steep of 2008, but have since reduced them to $1.73trn. We expect the trend to continue as risk
aversion declines and the curve remains steep. Figure 14 shows the y/y growth in
household holdings of money market funds explained by the current and lagged shape of
the 3m10y curve, which argues that there should be 10-15% reduction in 2010 as well, or
~$200bn. This pace is slightly lower than the $300bn annualized in 2009, which we believe
was high because of the reversal of the overshoot in late 2008. Bill holdings may also be
termed out by as much as $100bn as the bill universe shrinks, bringing the total to $900bn
($600 + 200 + 100bn) available to invest. The key question is how it will be invested.
Asset composition of investment

Households using savings to pay First, let’s examine the evidence so far. Figure 15 tabulates how households have deployed a
down liabilities total of ~$650bn of their savings and the money from terming out as of the third quarter of
2009. They have used part to pay down their liabilities, as a result of banks’ unwillingness to
lend, as well as less demand for loans. The difference between direct and indirect
investments beyond that is stark. The net direct purchase of Treasuries in 2009 was
completely an asset allocation out of agency securities. This is a little puzzling, as individuals
do not hold as much in agency securities, especially agency MBS.
The household category, in addition to individuals, includes domestic hedge funds and non-
The increase in Treasury
profit organizations. We believe the Fed purchases of agency MBS displaced these investors
holdings was largely due to the
into Treasuries. Also, net issuance of agency discount notes and callables has been negative,
Fed’s displacing investors from
pushing investors in Treasuries (we estimate that 30% of Treasury purchases were in bills).
agency MBS and negative net
Overall, there was little direct demand for fixed income securities; out of $170bn, equities
issuance of agency debt
accounted for $230bn. Given these one-time effects on direct holdings, indirect investments
are a better gauge of the desired composition of individual investments, in our opinion. Of
the $335bn invested through mutual funds, 70% found its way into spread products, 20%
into equities, and only about 10% into Treasuries; these trends have held fairly constant
over the past few months even though spreads, particularly corporate spreads, have
compressed and the equity market has rallied off its lows.
Figure 14: Households to continue to terming out Figure 15: Household indirect investments a better gauge
30% Household investment 2009 up to Q309
25% a. Net savings 436

20% b. Short-term outflows 224
15% c. Net Increase in Liabilities (157)
10% Net Acq. of Assets (a+b+c) 502
5%
Direct Indirect Total
0%
-5% Treasury (including bills) 529 21 549
-10% Agency debt/MBS (657) 48 (609)
-15% Other spread products 62 151 212
Dec-95 Dec-00 Dec-05 Dec-10 Equities 233 63 296
yoy growth, HH MMF shares, % Retirement A/C - 53 53
Estimated Using Current and Lagged Curve Total 167 335 502
Source: Federal Reserve, Barclays Capital Source: Federal Reserve
8 January 2010 19
Figure 16: Households: Treasury not the preferred asset Figure 17: Defined benefit equity allocation already at 40%
Short term 150 65%

8% 100 60%
Treasury 50 55%
17% 0
50%
Life+Pension -50
45%
45% -100
Spread 40%
-150
Products
8% -200 35%
-250 30%
2000 2001 2002 2003 2004 2005 2006 2007 2008
Equities
22% Equity Flows, $bn FI Flows, $bn
Equity/Assets, %, RHS
Source: Federal Reserve. Desire investment composition of households. Source: Federal Reserve
We expect a total of $600bn in If we extrapolate from this trend, households should continue to pay down their liabilities.
demand for fixed income As discussed earlier, we expect banks to remain wary of making loans, and we have already
securities, with the bulk in accounted for the demand for securities from banks. Households should then have
spread products and through ~$750bn ($900 - $150) to invest in securities, of which ex-ante $525bn should find its way
mutual funds into spread products, $150bn into equities, and the remainder into Treasuries. The bulk of
this is likely to be channelled through mutual funds, with some funnelled through
retirement accounts. A model of the household investment decision using factors such as
the ratio of net worth to income, the level of short rates, and consumer confidence yields
similar results. Figure 16 shows that 17% should be invested in Treasuries and 22% in
equities. Of the remainder, historical data suggest that the bulk should go into retirement
accounts, as was the case in earlier episodes of higher savings, but we haven’t seen
evidence of this so far in this episode. Nonetheless, these investors typically also prefer
spread products/equities.
Defined benefit pensions: Increasing allocation to fixed income

Another source of demand for fixed income is defined benefit plans’ continuing to allocate
out of equities and into fixed income, which could absorb another $100bn or so.
Defined benefit pension plans Figure 17 shows that this allocation has been under way for the past several years, and the
have been allocating away from share held in equities has already declined substantially. In 2007 and 2008 (latest data
equities into fixed income; the available), defined benefit pension plans were net sellers of equities to the tune of $200bn
share of equities has already each year and net buyers of fixed income (directly and indirectly) to the tune of $100bn. The
declined to 40-45%. sharp correction in equity prices has also increased the allocation to fixed income. Figure 17
shows that the share held in equities fell from an average of 60% during 2004-2006 to 40%
by the end of 2008. Assuming that similar flows continued into 2009, the share would have
risen marginally, to 40-45%, after accounting for the rally in the stock markets. Were these
trends to remain in place in 2010, the share of equities could fall to just 30-35%, suggesting
a slowdown in allocation out of equities in 2010.
Another reason for a slowdown in 2010 is that during most of 2007 and 2008, when these
plans were allocating out of equities, the funded ratio was close to 100% (Figure 19).
Currently, the funded ratio is at 75%, as the rally in the equity markets has been offset by an
increase in the present value of liabilities as corporate spreads have tightened. Such a large
8 January 2010 20
shortfall is much harder to bridge when heavily invested in fixed income. Hence, demand
from this investor base is unlikely to exceed $100bn.
Conclusion: Steeper curve/Tighter spreads

Ex-ante demand shortfall To summarize, demand for term fixed income securities seems to be close to $1.6trn, with
of close to $1trn, concentrated overseas and households accounting for 35-40% each and banks and pension funds should
in Treasuries absorb another 20% and 5%, respectively (Figure 18). While net demand of $1.6trn is a
large upward swing from the $875bn in 2009, it still falls short of total net term fixed
income supply of $2.6trn in 2010. Also, the ex-ante demand shortfall is mostly in Treasuries.
Other than for overseas official investors, the desired investment seems to be spread
products.
Market implications
Rates curve to stay steeper than First, rates (particularly longer-term rates) need to rise so that issuers such as corporates
the market is pricing in have less incentive to issue, the Treasury lowers the pace of terming out, or investors term
out assets at a faster rate. With front-end rates pegged to the fed funds rate, the curve
should be steeper. Figure 20 shows that the market is pricing the 2s10s swap curve to
flatten by 80bp over the next one year. With Fed hikes still nine months away according to
our economists, such a pre-emptive flattening is unlikely; in the past five hiking cycles, the
2s10s curve has, on average, flattened by just 10bp over the year leading into the first hike.
Long-end swap spreads Second, spreads should be tighter, as demand is skewed toward spread products and
should drift tighter, as the supply is skewed toward Treasuries. In the rates market, this should translate into tighter
demand shortfall is concentrated swap spreads. We believe longer swap spreads, 10y and 30y, could tighten even further
in Treasuries from these levels. The Treasury supply expectations embedded in longer-dated spreads are
optimistic, in our opinion. In addition, growing concerns about the sovereign rating with
deficits persisting at high levels for several years should put further tightening pressure on
long-dated spreads (see the “Swap Spread: Weighed Down by Supply” article for details).
Figure 18: Supply-demand imbalance of term-fixed income securities

Desired investment
Ex-ante demand Total Tsy Spread pdts. Source
Overseas +600 +600 0 Reserve growth/terming out

Banks +300 0 +300 Loans to securities
Households +600 +100 +500 Savings/terming out
Pension allocation +100 0 +100 Equity to fixed income
Total demand 1,600 +700 +900
Total supply 2,640 +1,840 +800
Imbalance -1,040 -1,140 +100
8 January 2010 21
Figure 19: Defined plans still well underfunded Figure 20: Forward curves are too flat
110% 300
265
105%
250
100%
200 182
95%
90% 150
113
85%
100
80%
75% 50
70% 0
Nov-05 Nov-06 Nov-07 Nov-08 Nov-09 Spot Jun10 Dec10 Jun11 Dec11
Funded Ratio 2s10s Tsy Curve, bp
Source: Milliman Source: Barclays Capital
8 January 2010 22
MONEY MARKETS
Short rates: Choppy seas ahead

Joseph Abate Markets needed relief after the sturm und drang of 2008. A long period of low, stable,
+1 212 526 6815 and predictable interest rates in 2009 was exactly what the short rates markets
joseph.abate@barcap.com required. But this sweet spot is about to end in 2010, in our view.
Two key themes are likely to emerge next year: the Fed’s exit from credit easing and the
shift to greater regulation.
We look for reserves draining operations to cause a sharp steepening in short rates
beginning early next spring.
Supply and demand dynamics next year are least favorable for the repo market but
probably much better for bills.
Regulatory reform has only just started. A lack of coordination, together with significant
uncertainty about timing and implementation, is likely to widen spreads and increase volatility.
While we do not look for anything like 2008’s storm and fury, 2010 is likely to be much
choppier than 2009.
Decision time
Although the FOMC retained its “extended period” language at the December 2009
meeting, the time for exiting credit easing is quickly approaching. Our economists are
projecting strong growth in Q4 to carry over into 2010, giving the Fed more confidence that
the economy will be able to stand on its own without the extraordinary support necessary
since 2008. A lingering output gap should keep inflation (and inflation expectations) low,
enabling the Fed to move to the sidelines gradually – taking back a portion of the deflation
insurance pushed into the economy last year.
The FOMC has a lengthy agenda Indeed, the Fed has begun discussing the mechanics of its exit strategy – and even testing
to work through before exiting some of the necessary tools. The November FOMC minutes contained a fairly detailed
credit easing description of some of the issues that the Committee will need to address over the next few
months. Leaving aside the macroeconomic considerations of projected growth and
inflation, some of the leftover items on the Fed’s “to do” list include: the sequencing of rate
hikes and reserve drains, the timing of future asset sales, and the size and types of tools to
be used. Less prosaically, the FOMC needs to decide what policy variable it wants to use: the
effective fed funds rate – like past cycles – or the rate on bank reserves.
How much to drain?

The Fed’s huge balance sheet All these considerations combine with the Federal Reserve’s balance sheet swelling to
raises the issue of what to do $2.5trn by March 2010 from under $1.0trn before the crisis and with an unprecedented
with $1.3trn in bank reserves level of bank reserves. In past rate hike cycles, the level of bank reserves has not mattered
much because balances have been small (less than $50bn) and banks actively minimized
them. But at $1.3trn, the level of bank reserves is so large that it needs to be taken into
consideration – especially since the quasi-monetarists on the FOMC argue that a pile this
big is potentially very inflationary (Figure 1). As a result, these members would argue for a
quick and decisive move to drain reserves from the banking sector. Some members even
argue that the Fed’s balance sheet should be returned to its pre-crisis level and composition
– just $1trn made up of only Treasury securities. However, we do not think this is the
8 January 2010 23
consensus view on the FOMC. Rather, the Fed is probably content to keep its balance sheet
large and just reduce the level of bank reserves – probably in advance of hiking the fed
funds rate target in September 2010 as we expect.
We look for the Fed to leave Our working assumption is that the Fed will reduce reserve balances by between $800bn
several hundred billion in and $1000bn – leaving between $300bn and $500bn of excess in the banking system. Some
reserves in the system portion of the excess is likely to be used to satisfy banks’ lingering precautionary demand
for reserves, which has been shocked higher after the financial crisis. Moreover, the Fed
might also want to leave a few hundred billion dollars in bank reserves at banks to help
reduce intraday credit risk. Intraday credit risk crops up when banks are temporarily
overdrawn in their security (and cash) accounts at the Fed from clearing and settling
transactions. This exposes the Fed to significant counterparty risk. Since the Federal Reserve
began paying interest on reserves last fall, however, the level of daylight overdrafts has
fallen sharply and clearing systems have become more efficient (Figure 2).
What tools?
Reverse repos and term deposits The Fed has outlined three principal tools it plans to use to drain bank reserves. All three
are likely to be the Fed’s most tools share the same basic principle: reallocate the Fed’s liabilities away from bank reserves.
important tools And each tool has been mentioned by Fed officials in roughly the same order, indicating a
hierarchy of sorts: reverse repos, term deposits, and the joint Treasury-Federal Reserve
Supplemental Financing Bill (SFB) program. 1
Reverse repos have received the bulk of press attention largely because the Fed has already
begun testing these transactions – swapping some of its collateral against cash from its
traditional counterparties (the 18 primary dealers). However, we believe that the capacity of
the primary dealer community is too limited to enable the Fed to drain the $500bn we expect
in reverse repos. Instead, we expect the Fed will expand the list of its reverse repo
counterparties to include money market funds (MMFs) and the GSEs. In our opinion,
regulatory relief for primary dealers on balance sheet netting is unlikely to make reverses with
the Fed sufficiently attractive to push dealers into $500bn worth of transactions with one
counterparty – albeit the Fed. After all, reverse repos with the Fed are not generally huge
money makers for primary dealers.
Figure 1: Bank reserves ($bn) Figure 2: Bank daylight overdraft ($bn)
1,600 300 Fed begins paying

interest on reserves
1,400
250
1,200
200
1,000
Forecast
800 150
600 100
400
50
200
0 0
Jul-08 Nov-08 Mar-09 Jul-09 Nov-09 Mar-10 Jan-06 Oct-06 Jul-07 Apr-08 Jan-09
Peak overdraft Daily average
Source: Federal Reserve, arclays Capital Source: Federal Reserve Bank
1
For more details on the Fed’s exit strategy, its tools and the mechanics of reserve management, please see Heading
for the Exit, September 28, 2009
8 January 2010 24
Arbing the Fed

One issue that will likely need to be addressed is the spread between the effective fed funds rate and the interest rate on
reserves. In theory, a Lombard facility, like those in the UK and Canada should work just as well in the US. In such a facility, the
central bank corrals the overnight call money rate by establishing both a floor (the interest rate paid on reserve balances held at
the central bank) and a ceiling (the cost to banks to borrow reserves – the discount rate). Typically, both rates are set as a
symmetric spread to the overnight target – say at a spread of plus/minus 100bp. As discussed by various Federal Reserve
officials, the ability to pay interest on bank reserves sets the floor for all interest rates – excess bank cash simply flows back to
the central bank and out of the call money market. However, it works best when the floor is lower than the target. In the US, the
target is set as a range from 0 to 25bp, while the interest rate on reserves is 25bp. This arrangement creates a bit of a problem
for establishing a true Lombard facility in the US.
Banks have a strong incentive to borrow at the overnight call money rate (fed funds effective) and park the borrowed cash at
the Fed to earn a positive spread – generally between 10 and 15bp. There is no way of estimating the size of this activity,
although we suspect it is sizeable given the amount of fed funds listed as sold in the 10-Qs of the three largest GSEs – at least
$150bn. Figures from the Federal Reserve indicate that roughly 85% of the growth in bank reserves this year has accumulated
on the balance sheets of the 25 largest US commercial banks – the institutions most likely to be borrowing and depositing
funds at the Fed. A portion of their nearly $900bn hoard is used as precautionary insurance against the type of liquidity runs
seen in 2008. But there is undoubtedly a large portion (above $150bn) that is probably being used to arb the spread between
the effective fed funds rate and the interest on reserves.
This would be less of an issue if there was a way for the Fed to control the steady supply of cash coming into the fed funds
market. After all, without any additional cash, banks would arbitrage the spread between the fed funds and the interest rate on
reserves to 0. However, the fungability of reserves along with the strong incentive the GSEs face to sell their cash into the fed
funds market (since they can’t earn interest from the Fed) creates a steady flow of cash into the market that the largest
commercial banks simply borrow and park at the Fed.
We believe that in order for the Fed exert some traction on rates, at a minimum, it will need
to remove the cash flowing into the fed funds market from the GSEs – perhaps by doing
reverse repos with them. Alternatively, the Fed could lower the rate paid on bank reserves to
less than the effective fed funds rate. 2 But with the effective now trading at 12bp,
attempting to eliminate that spread might require ceasing the payment of interest on bank
reserves. Instead, it seems like the only opportunity for the Fed to open up a spread
between the two rates will come through holding the rate on reserves constant as the funds
target is boosted. As a result, the bank reserve arbitrage trade is likely to continue through
September 2010.
At the end of December, the Federal Reserve circulated some details about its proposed
term deposit facility. The basic structure of the program was not much different than what
markets anticipated. In it, the Fed would transfer balances from regular reserve accounts
into a term deposit that would be the functional equivalent of a certificate of deposit for
households. These deposits could not be used to meet reserve requirements, they would
have a zero risk weighting and would be auctioned through a competitive bidding process.
The Fed’s proposal was not very detailed with respect to the maturities or the maximum
bidding rate. We assume that the maximum maturity will be 84 days – mainly to match the
duration of the analogous TAF program or it could be set to match the 65 business day limit
on the Fed’s reverse repos. However, there is no clear benefit to the Fed (or banks) from
matching the maturities of either program.
With respect to the maximum bid rate at the term deposit auctions, we assume the Fed will
set the rate equal to the discount rate – although the language of the December proposal
2
The mood in Congress toward the Fed and the GSEs is frosty. As a result, we doubt the Fed would be able to get
permission to pay interest on GSE deposits held at the central bank.
8 January 2010 25
referred to a variety of short-term interest rates with maturities up to a year. As a practical

matter, the interest rate ceiling at the term deposit auctions should be equal to the discount
rate to prevent bank arbitrage. And while there might not be much arbitrage between the
discount window and the term deposit program given the still-present borrowing stigma,
the ability to “arb” the Fed via the TAF program will need to be addressed. Our sense is that
the Fed will either put the TAF program into quasi-permanent hibernation (to be
resuscitated at the next financial crisis) or it will increase the borrowing cost for TAF loans.
Either way, before the Fed can begin draining bank reserves via term deposits, it will need to
work through some of the mechanics of its interactions with other Fed programs.
However, we expect the SFB Finally, we expect the SFB program to come back early next year – although in a somewhat
program to ramp back up to limited capacity. Political noise around increasing the debt ceiling has significantly reduced
$200bn after March the SFB program’s attractiveness to the Fed. Nevertheless, between February and October
2009, the program helped offset a constant $200bn worth of Fed asset expansion. Since the
program was not entirely abandoned in October, we suspect that despite the Fed’s distaste,
it will once again be boosted to $200bn after another, more permanent, debt ceiling
increase in March 2010.
Fed asset sales are unlikely By contrast, the only permanent way to drain bank reserves requires the Fed to sell assets.
in 2010 The minutes of the November FOMC meeting mention selling some of the more than $2trn
worth of securities in the Fed portfolio as a means of shrinking its balance sheet and
draining bank reserves. However, while we expect the Fed to eventually begin offloading
some of the MBS and Treasury securities in its portfolio, such action is unlikely in 2010 with
a still precarious housing market. Instead, the Fed will need to continuously reallocate its
liabilities away from bank reserves, while waiting for a profitable time to sell off assets.
Spring steepening
The first draining operations are We expect Fed rate hikes to be preceded by reserve draining operations – by about three
likely to start in May or months (Figure 3). Thus, by late spring 2010, the Fed will commence draining bank reserves
June 2010 through reverse repos. Initially, these operations will be small and involve the primary
dealers. We also expect the first operations will use Treasury collateral. Only late in the
fourth quarter do we expect the Fed to expand its reverses to include MBS collateral along
with a broader array of counterparties. At the moment, it is unclear if term deposits will be
used by late spring 2010 as well.
The first draining operations are Short rate markets will steepen sharply immediately following the first reserve draining
likely to start in May or operation – regardless of the tool used or its size. 3 The steepening may begin weeks before
June 2010 the first operation – the first reserve drain will confirm it. At the extreme short end, the
overnight space, the Fed’s operation will immediately push the effective fed funds rate to
the upper boundary of its 0-25bp range. Similarly and consistent with history, the OIS curve
should steepen as well – in anticipation of a long series of rate hikes and potentially more
aggressive action from the Fed once it begins raising the target rate. Further out, term repo
markets and commercial paper, which all trade as spreads to OIS, should move higher.
Uncertainty around the Fed’s path and the speed with which it intends to take the target
back to neutral is likely to make next spring’s steepening fairly volatile.
3
The late fall surge in bank reserves – from $800bn to $1100bn – has had a surprisingly muted effect on the effective
fed funds rate. Instead of pushing the fed funds rate sharply lower, funds have stayed locked in a narrow range. This
suggests that the relationship between the level of bank reserves and the funds rate is not linear. Consequently, the
Fed’s first draining operations are likely to have disproportionally large effects on the funds rate – regardless of their size.
8 January 2010 26
Supply and demand factors

Supply and demand factors in various short rate markets are likely to cause some rates to
back up further and more abruptly than others, causing spreads to widen. For instance,
since the Fed’s first draining operations are likely to be reverses, we expect a
disproportionate back-up in Treasury collateral rates relative to other repo product and
other short rate markets. Indeed, pushing potentially $500bn worth of collateral into the
repo market (accounting for 20% of the amount outstanding across Treasuries, agencies,
and MBS) could easily cause repo rates to back up to 20bp or more over the fed funds rate
by the end of 2010. Most of the back-up will be in GC, since we do not expect the Fed to
start performing reverses against MBS until late 2010. As a result, the spread between MBS
and Treasury collateral should remain fairly narrow for most of 2010.
Bills may retain much of their By contrast, there is likely to be comparatively less steepening in the bill market. The Treasury
scarcity premium in 2010 recently affirmed its commitment to lengthen the average maturity of the US debt. This will
require issuing more intermediate paper, and, most significantly, substantially reducing gross
bill supply. Indeed, we look for gross bill issuance to be cut from $2.0trn in 2009 to $1.2trn in
2010 (excluding an expected $200bn in SFBs). Such a sharp reduction could mean that
Treasury bills will retain much of their scarcity premium next year relative to other short rate
assets – even if the 2008-09 flight to quality move is a distant memory (Figure 4).
Meanwhile, in other short rate markets, we do not expect much of a shift in supply.
Commercial paper outstanding, which has shrunk by 30% in the past year, is expected to
increase only slightly in 2010. In the past year, banks have termed out their debt and shifted a
significant portion of their funding toward more sticky retail deposits. Regulatory emphasis is
likely to keep the pressure on banks to issue less short-term debt. Similarly, new accounting
rules around off-balance sheet vehicles (Fin 166 and 167) are expected to dampen new
issuance of asset backed CP next year. But a robust economic recovery should boost business
demand for working capital, which in turn, is likely to push up non-bank CP issuance. Overall,
we expect commercial paper supply to increase by about 5% in 2010 – or $60bn.
Figure 3: Federal Reserve operations timeline Figure 4: Term GC and bills (bp)
180
1-Feb Fed liquidity programs end
160
31-Mar Fed asset purchases end
140
May/Jun Fed reverse repos begin draining liquidity
120
-- against TSYs and with dealers only
100
Reverses increase steadily
80
Summer Fed begins doing term deposit auctions
60
Sep Fed hikes funds rate
Oct-Dec Reverses are expanded 40
-- to include other collateral and counterparties 20
2012? Fed starts selling assets 0

-20
Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09
Source: Barclays Capital Source: Federal Reserve Bank, Barclays Capital
8 January 2010 27
As risk aversion declines, we On the demand side, we expect a clear shift away from low rate, short paper next year.
expect demand for short rate Measures of financial stress have already returned to late 1998 LTCM levels and are
investments to erode expected to rapidly sink back to pre-crisis conditions (Figure 5). Accompanying the decline
in risk aversion, we expect the demand for higher-yielding investments to increase. This
should be most apparent in the reduction in the demand for money fund deposits. At the
height of the financial crisis, money funds held $3.8trn in assets under management – up
52% from pre-crisis levels. And while balances have steadily eroded since early spring, we
look for additional shrinkage early next year – with overall money funds perhaps shrinking
another 22% to $2.5trn (Figure 6). So long as the erosion in balances is steady (as it has
been this year), money fund managers can breathe a sigh of relief. Super low yields, a lack
of investment product, and massive inflows practically eliminated their fees and made
operating these super-safe funds extremely complicated. The return of higher rates and
smaller balances should enable the funds to at least start collecting operation fees again.
Repo and CP rates should widen Assuming the money funds do not adjust their investment mix much in the coming year, a
relative to bills 22% reduction in aggregate assets under management will reduce the demand for repo,
bills, and commercial paper by a proportional amount. In dollar terms, we expect money
fund appetite for all three investments to decline $150bn, $100bn, and $160bn,
respectively. The combination of increased supply and reduced demand for repo and
commercial paper should push their rates higher than bills – i.e., their spread to bills should
widen several basis points. The Treasury’s reduction in bill supply, however, is larger than
the anticipated reduction in bill demand from money funds.
Foreign official demand for bills Money funds, of course, are not the only source of demand for bills. Foreign official
is likely to erode further in 2010 institutions, among others, have been heavy buyers of short-term US government debt. As
noted recently by Anshul Pradhan, however, their demand for Treasury bills has slackened –
slowing to just $25bn in Q3 from about $150bn at the end of 2008. 4 Instead, these
investors are terming out – with demand for coupons in Q3 rising to $100bn compared with
very little outside the bill sector at the height of the financial crisis in Q4 08. We expect the
terming out and reduction in foreign official demand for bills to continue in 2010.
Figure 5: KC Financial Stress index (Q4 98=100) Figure 6: Money market fund balances ($bn)
800 4,500
700 4,000
Tax free
600 3,500
500
3,000
400 Prime
2,500
300
2,000
200
1,500
100
0 1,000 Gov-only
-100 500
-200 0
Feb-90 Feb-94 Feb-98 Feb-02 Feb-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09
Source: Kansas City Federal Reserve Bank, Barclays Capital Source: imoney.net
4
See “Robust overseas demand”, Market Strategy Americas, December 17, 2009.
8 January 2010 28
Households have emerged as a In the past year, the household sector has ramped up their demand for Treasuries – of all
large demand segment maturities. Indeed, at the end of Q3, their holdings had more than doubled from a year
for Treasuries earlier, and the sector now owns more Treasuries than money market funds (Figure 7).
Likewise, the dealer community has shifted from an overall short base in the Treasury
market to a long position in the past year (Figure 8). Dealers are typically short Treasuries
against their inventories of MBS, corporate bonds, and agency debentures. But, with the
level of risk aversion ebbing next year, household and dealer Treasury positions may return
to pre-crisis levels – accounting for a reduction in aggregate Treasury demand of several
hundred billion dollars with a significant component coming from the bill sector. Thus, the
key uncertainty for the bill sector next year is whether the reduction in bill demand coming
from households, dealers, and money funds will be larger than the expected supply
reduction. Our sense is that unlike commercial paper and certainly repo, the bill sector has a
greater chance of richening (in a relative sense) based on its supply and demand dynamics.
Regulatory reform: Unknown variable for 2010

A slate of regulatory changes coming in 2010 is likely to produce volatile short-end markets
– regardless of the Fed’s monetary policy actions and the supply-and-demand dynamics.
Upcoming regulations run the gamut: from how money funds operate, to how the tri-party
repo market is organized. Besides reducing systemic risk, the regulatory changes are likely
to have two significant consequences: increasing funding costs (particularly for banks) and
– given their as-yet-unresolved state and in some cases conflicting nature – increasing
market volatility.
Most of the regulatory efforts are Most of the regulatory changes planned for 2010 involve efforts to reduce systemic risk at
aimed at banks banks. FDIC bank deposit insurance premiums have been boosted for funding other than
traditional insured deposits. The intent is to push banks to rely less heavily on repo and
wholesale funding markets – markets that either precipitated or exacerbated liquidity runs
during the crisis. At the same time, regulators are also pushing banks to raise capital and
establish thicker liquidity buffers. These buffers (designed from stress test results) are
meant to enable banks to operate on their own for several months if liquidity pulls away and
they lose a significant portion of their external financing. The buffers will be composed of
government securities and cash.
Figure 7: Household and MMF Treasury holdings ($bn) Figure 8: Dealer Treasury and bill positions ($bn)
700 150
600 100
500 50
400 0
300 -50
-100
200
-150
100
-200
0
Mar-00 Mar-02 Mar-04 Mar-06 Mar-08 -250
HH MMFs Bills All TSYs
Source: Federal Reserve Source: Federal Reserve
8 January 2010 29
Pending legislation in the House goes even further – establishing limits on the amount of
short-term funding banks can use. HR4173, however, does not provide numerical
guidelines – presumably these would be set once the legislation makes it through the
Senate and is signed into law. It would, however, give the Federal Reserve and other
regulators the ability to put a systemically important institution into receivership or
liquidation. Moreover, a macroprudential regulator – set up with participants from the Fed,
Treasury and FDIC among others, would have the ability to require a bank to cease certain
activities and even order it to divest certain businesses.
Bank funding costs will increase It is difficult to estimate the expected increase in bank funding costs that will occur as a
result of this legislation. Banks have already shifted away from wholesale funding sources –
their participation in the repo and fed funds markets is down about 66% since September
2008, while their use of retail deposits has largely supplanted wholesale large time deposits.
With forward LOIS expected to We expect the increase in bank funding costs will push LOIS spreads wider with the 3m rate
widen in 2010 widening from 9bp currently to potentially 40bp or more by the end of 2010. However,
there is a strong counter case that the reduction in systemic risk caused by the tightening of
bank funding sources and the strengthening of their balance sheets and liquidity sources
should result in a narrowing in future LOIS spreads. The market seems conflicted – pushing
the current 3m spread below the 12.5bp pre-crisis average but widening it to 25bp at the
end of 2010.
Bank heads were recently criticized by the president for not making more small business
loans and re-working the terms of home mortgages in default. Efforts to push banks to raise
cash, hold higher liquidity buffers and move away from wholesale funding while all the
while increasing their lending is a tall order – particularly for institutions that just
experienced their biggest crisis in 80 years.
Meanwhile, money market funds have become another focus of regulatory attention since
the Primary Reserve Fund “broke the buck” in September 2008. The Federal Reserve and the
President’s Working Group on Financial Markets view the funds as potential hazards to
financial markets since they act like banks – intermediating between lenders and borrowers
– but are not regulated and do not hold capital like banks. Their stable NAVs are
backstopped by the reputations of the fund complexes managing the money funds. In a
flight like the one in September 2008, despite the reputational hazard, fund complexes may
not have pockets deep enough to take on the assets of a faltering money fund to prevent its
NAV from drifting below $1/share. Instead, investors who withdraw their money first are
more likely to get $1/share than those who wait.
The SEC draft would reduce Details of the planned regulatory changes to MMFs have been delayed until sometime early
maximum fund WAMs from 90 in the first quarter. The initial SEC draft released last summer included tighter restrictions on
to 60 days the types of assets money funds could purchase, as well as shorter-weighted average
maturities (WAMs). In addition, it would establish liquidity requirements based on the type
of investors in the funds. Those whose investors are primarily institutions would be required
to hold larger cash buffers (30% of assets in securities maturing in less than seven days).
While not specifically addressed in the initial SEC draft, we believe that regulators are
interested in the establishment of a “liquidity bank” structure that could provide
collateralized loans to money funds in distress. Such a bank would necessarily have to be
quite large given the size of the money fund industry (at more than $3trn) – perhaps
between 2 and 5% of assets under management. Details could be included in the next SEC
draft to be released early in the first quarter.
8 January 2010 30
We believe these restrictions would push money funds all into the same types of trades with
a heavy concentration of repo and some light seasoning in longer maturity, higher yielding
commercial paper or other assets. This could create some issues in future financial crises,
especially as funds faced with redemption all seek to unload their repo positions at the same
time. More significantly, shifting the composition of money funds assets will reduce liquidity
in other market and change issuance patterns. For instance, with money funds less able to
purchase commercial paper – where they make up 44% of the market – issuers looking for
short-term money will have to turn to other sources, namely more expensive short-term
bank loans. Similarly, spread WAM limitations based on a security’s final maturity (rather
than its first reset date) will require some floating rate note issuers to change the maturities
of their paper in order to accommodate the money funds.
These changes are expected to reduce systemic risk – but they will change how markets
operate and raise funding costs. Until the rules are clearly established and markets adjust to
the new regulatory regime, we expect spreads and rates to become more volatile. This,
coupled with the uncertainties surrounding the Fed’s exit strategy, is likely to make 2010
quite a bit choppier than the calm seas of 2009.
8 January 2010 31
TREASURIES
Pick your poison

Anshul Pradhan 2010 should be a challenging year for US Treasuries, though not as bad as 2009.
212-412-3681 Regardless, Treasury investors might have to make tough decisions on where to
anshul.pradhan@barcap.com maintain exposure. In other words, it’s time for Treasury investors to pick their poision.
2y rates should move gradually higher through 2010. However, given a cautious Fed and
Amrut Nashikkar
the high level of excess reserves, the risk is toward 2y rates staying below what the
+1 212 412 1848
market is pricing in, at least during the first half of the year.
The Treasury should continue to term out debt by letting bills roll over and marginally
increasing auctions sizes in the intermediate sector and the long end. Average maturity
should increase to ~60 months by the end of the 2010 fiscal year.
Increase in Treasury supply, coupled with a sharp swing in the supply of spread
products hitting the market, should outweigh the increase in demand from private
investors. We recommend steepeners, as long-term real rates are still low and the
flattening priced into the forward curve appears excessive
Within curvature trades, the 2s5s10s butterfly should cheapen as the Fed remains on
hold amid strong data, the supply-demand imbalance keeps the curve steep, and
mortgage convexity hedging cheapens the 5y sector further relative to the wings. The
10s20s30s fly is trading in line with market stress variables but has room to richen as
issuance in the 10y and 30y sectors increases.
Most relative value trades have gradually corrected; however, some still stand out. High-
coupon Feb19s and Aug17s are trading cheap to their low-coupon counterparts
compared with other such pairs in the 7-10y sector. Also, the on-the-run 10y seems to
be trading rich and should cheapen as it ages away from the 10y sector. 10-11y issues
still offer good rolldown potential into the 10y sector, and REFCO P-STRIPS still appear
attractive compared with Treasury P-STRIPS. The introduction of the ultra-long contract
in January should lead to a cheapening of the 15-20y sector as liquidity moves further
out the curve.
The Treasury market was a tale of two tails in 2009; while 2y rates remained largely pegged
around 1%, 10y and 30y rates drifted higher by 160bp and 200bp, respectively. As a result, the
2s10s curve steepened by 125bp, to 269bp, ending the year close to the peak levels of the past
two decades. The Barclays Capital Treasury Index returned -3.6% in 2009, its worst
performance since 1973 (-3.4% in 1994 and -2.6% in 1999). Going into 2010, we expect this
trend to continue, with rates gradually inching higher and the curve staying steeper than what
the market is pricing in.
Figure 1: 2010 rates outlook

Market Q1 10 Q2 10 Q3 10 Q4 10
2s 1.02 1.10 1.60 2.00 2.30

5s 2.60 2.60 3.10 3.50 3.60
10s 3.82 3.70 4.20 4.50 4.50
30s 4.68 4.70 5.20 5.50 5.50
2s10s 2.80 2.60 2.60 2.50 2.20
Source: Barclays Capital. As of January 7, 2010
8 January 2010 32
Short-term rates: Fed policy and excess reserves

2y rates traded around 1% 2y rates traded around 1% for most of 2009, with a few exceptions around payroll releases
for most of 2009, and toward the end of the year, when they rallied to close to 60bp before ending the year at
with few exceptions 1.1%. As we move through 2010, 2y rates should rise, as the 2-year period will cover a larger
share of the hiking cycle that our economists expect to begin in September. The market is
pricing a move higher in 2y rates; however, given a cautious Fed and the high level of excess
reserves, which has been an important driver of 2y rates in 2009, the risk is towards 2y rates,
staying below what the market is pricing in, at least during the first half of 2010.
An increase in excess reserves in Figure 2 shows 2y Treasury rates against the level of excess reserves; a sharp rise in excess
the banking system was reserves has been associated with a rally in 2y rates and vice versa (with an r-square of
associated with a rally in 2y rates 0.8%). The rally in 2y rates in the fourth quarter coincided with a sharp rise in excess
reserves as the Treasury unwound the Supplementary Financing Program (SFP) and the Fed
continued to buy securities through the asset purchase program. Any forecast of 2y rates
over the next few quarters should therefore take into account the path of excess reserves, in
addition to the path of the fed funds rate. Figure 3 plots the path of 2y rates under two
scenarios for excess reserves, superimposed on the path of the fed funds rate.
With elevated economic With the unemployment rate remaining well above what the Fed would consider neutral
uncertainty and excess reserves and the core inflation rate below target, the hiking cycle should be gradual. Our economists
remaining high, 2y rates may expect the Fed to begin hiking in September 2010; to raise the funds rate to 1% by the end
remain below forwards for the of 2010; to remain on hold during the first half of 2011; and to end 2011 and 2012 at 2%
first half of 2010 and 3%, respectively. We therefore expect the average fed funds rate for the subsequent 2-
year period to increase from 0.8% to 1.9% over the course of 2010 (Figure 3). Were the Fed
not to change the liability side of its balance sheet, excess reserves would increase gradually
during the first half of 2010 as agency MBS settle, keeping 2y rates below what is implied by
the average of the fed funds rate, as well as the forwards. In the alternative scenario, we
assume that the SFP is revived in the second quarter (we think it is unlikely in the first
quarter, given that the debt ceiling was raised only modestly) and reserves are gradually
drained to $500bn by the end of 2010. Even under this scenario, 2y rates should remain
below the forwards for the first half of the year.
Figure 2: Short-term rates: Timing of reserve drain is key Figure 3: 2y rates in different scenarios
2.1 1,400 2.50
1.9 1,200 2.25
1.7 2.00
1,000
1.5 1.75
800
1.50
1.3
600 1.25
1.1
400 1.00
0.9
0.75
0.7 200
0.50
0.5 0 current Q1-10 Q2-10 Q3-10 Q4-10
Sep-08 Dec-08 Mar-09 Jun-09 Sep-09 Dec-09
Avg FF rates, 2y 2y Rates w/o reserve drain
2yr Rates, LHS Excess Reserves, RHS 2y Rates with reserve drain Market
Source: Federal Reserve, Barclays Capital Source: Barclays Capital
8 January 2010 33
Treasury supply: Steady march toward the target

The outlook for longer rates should be driven by the supply-demand imbalance, which we
discuss in detail in “US Fixed Income Supply/Demand: Supply Deluge.” Here, we look at the
Treasury component of fixed income supply in detail.
Auction sizes should increase The Treasury has come a long way in altering the auction calendar. Over the past two years,
marginally in the intermediate auction sizes have increased across the curve, although more so in the front to intermediate
sector and the long end; sector, especially with the introduction of 3s and 7s. The frequency and sizes of long-end
bill universe should continue auctions have also increased, but at a more gradual pace. Having done that, the Treasury
to shrink seems to be stabilizing front-end auctions at current levels: it kept the 2y and 3y sizes
unchanged for the past few auctions, although it continues gradually to increase auction
sizes further out the curve. We believe this trend will continue for most of 2010, with
issuance continuing to shift out the curve (Figure 4). We expect 2s and 3s to remain at
current levels till the end of fiscal year 2010 and to decline gradually from then on. 5s and
7s should rise until Q1 and stabilize at those levels before declining marginally later. 10s and
30s should continue to increase until the middle of 2010 before stabilizing at peak levels.
Gross and net coupon issuance As a result, gross and net coupon issuance should be $2.5trn (up $400bn from 2009) and
should increase to $2.5trn $1.8trn (up $300bn from 2009), respectively. In 10y equivalents, coupon supply should
and $1.8trn in 2010 increase to $1.6trn from $1.1trn in 2009, adjusted for $300bn of Fed purchases. The bill
universe will likely continue to shrink, as we expect financing needs to be close to $1.4trn
(assuming SFP borrowing of $300bn, in addition to our economists’ deficit projection of
$1.1trn). How does this tie in with the goal of increasing the average maturity? The Treasury
recently stated that it intends to increase the average maturity of outstanding debt to 6-7
years but that it will do so over several years. Our stated issuance mix would help increase
the maturity to ~60 months from the current 53 months (Figure 5). Were deficits to surprise
to the upside (OMB and CBO latest estimates of 2010 deficits are $1.5trn and $1.4trn,
respectively), bills would likely shrink less and, more important, auction sizes might stabilize
at higher levels, at least in the intermediate and long sectors, than we have pencilled in.
Figure 4: 2010 Issuance Calendar

2y 3y 5y 7y 10y 30y 5y I/L 10y I/L 30y I/L Total Maturing Net 10yEq
January 44 40 43 33 21 13 10 204 53 151 128

February 44 40 44 34 26 17 9 214 77 137 149
March 44 40 45 35 22 14 200 43 157 126
April 44 40 45 35 22 14 10 8 218 77 141 137
May 44 40 45 35 27 18 209 61 148 140
June 44 40 45 35 23 15 202 44 158 129
July 44 40 45 35 23 15 11 213 44 169 138
August 44 40 45 35 27 18 8 217 63 154 152
September 44 40 45 35 23 15 9 211 47 164 136
October 43 39 45 35 23 15 8 208 47 161 132
November 42 38 44 34 27 18 9 212 53 159 144
December 41 37 44 34 23 15 194 51 143 126
Total 2010 522 474 535 415 287 187 18 47 17 2,502 660 1,842 1,637
Total 2009 499 430 447 306 242 129 15 29 14 2,109 553 1,556 1,070*
% Increase 2009 to 2010 5% 10% 20% 36% 19% 45% 20% 62% 21% 19% 19% 18% 53%
* After adjusting for Fed purchases. Source: Barclays Capital
8 January 2010 34
Beyond 2010, we expect average maturity to increase gradually, stabilizing in the middle of
the range over the next 5-10 years (Figure 5), but Treasury supply is likely to remain high.
The issuance mix and actual auction sizes in years further out also depend on deficits. We
consider two scenarios: first, the base case, in which deficits follow our economics team’s
projections, declining to $1.0trn in 2011 and eventually stabilizing at 5% of GDP. The
second scenario is the CBO’s latest baseline projection, but adjusted for alternative policy
measures. The CBO’s baseline factors in current laws and policies but does not take into
account possible future changes. We adjust this using the CBO’s own estimate of the effects
of added policy measures, such as changes to the tax code. 5 Deficits stay at $1.4trn in 2010
under this scenario and decline gradually to $1.1trn (or 7% of GDP) before marching up
again. In both scenarios, we assume that the added borrowing needs linked to the SFP
program of up to $300bn in 2010 will be reversed over the subsequent three years at a pace
of $100bn each year.
Beyond 2010, the Treasury In the base case, we expect issuance in the front end to decline steadily beginning in fiscal
should be able to reduce year 2011, with 2s averaging $35bn in FY2011 and $26bn in FY2012 before stabilizing at
auction sizes in the front end if $20bn; the same should be the case with 3s (Figure 6). Further out, 5y and 7y auction sizes
deficits follow our economics should decline only marginally, and 10s and 30s should stabilize at peak levels before
team’s forecast eventually rising further. As a result, supply should gradually shift out the curve, with 10s and
30s contributing one-third of coupon issuance, compared with 20% in 2009. 2010 should
mark the peak of gross coupon issuance at ~$2.5trn and 2013 the trough at $2trn (Figure 7).
Gross issuance is unlikely to decline further, as the amount of debt reaching maturity should
increase rapidly. Duration issuance should vary less because of a gradual outward shift and
remain at $1.5-1.6trn in 10y equivalent terms over the next several years. One consequence of
the terming out should be a gradual reduction in the amount of debt maturing in less than one
year, declining from 38% to 20% over the next several years before eventually rising as the
Treasury stabilizes the average maturity. This is well below historical averages, but the
Treasury’s stated target of 6-7 years is also well above historical averages.
Figure 5: Projected path of the average maturity Figure 6: Issuance shifting to the long end, average size, $bn
85 50
80 45
Sep-19,
75 40
80M
70 35
Sep-10, Sep-12,
65 61M 71M 30
25
60
20
55 Sep-09,
53M 15
50
10
45
5
40 0
Sep-01 Sep-04 Sep-07 Sep-10 Sep-13 Sep-16 Sep-19 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Average Maturity, Months 2y 3y 5y

7y 10y 30y
Note: TIPS included in respective sectors. Source: Treasury, Barclays Capital Source: Treasury, Barclays Capital
5
We have adjusted the CBO’s baseline primary deficit using its own estimate of the effect of alternative policy
measures as detailed in the July estimate of the President’s budget. The interest cost is then projected assuming that
rates move along the forward path and the average maturity of debt increases as shown in Figure 5.
8 January 2010 35
Figure 7: Coupon issuance peaking in 2010 Figure 8: No relief in CBO’s scenario
3,000 40% 70
60
2,500
35%
50
2,000
40
30%
1,500 30
25% 20
1,000
10
500 20%
2009 2011 2013 2015 2017 2019 0
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Gross Coupon Supply, $bn
Coupon Supply, 10-yr Eq, $bn 2y 3y 5y
% Debt Maturing in <1y, RHS 7y 10y 30y
Source: Treasury, Barclays Capital Source: Treasury, Barclays Capital
Were CBO forecasts to be Under the second scenario – the CBO baseline adjusted for alternative policy measures – any
realized, there is likely to be little relief in terms of coupon size would be marginal. Figure 8 shows that 2y and 3y auction
relief in auction sizes sizes should remain at high levels and 5s should rise until the middle of next year before
stabilizing for the next several years. Auction sizes further out should keep rising steadily,
with 10s increasing to $40bn per month over the next five years. Bond auction sizes should
also increase to the average size of the current 10y. Overall coupon issuance would stay
close to $2.5trn in notional terms and at a little less than $2trn in 10y equivalent terms for
the next several years. While this may seem extreme, the main takeaway is that Treasury
supply is likely to be an issue for several years.
Changing demand dynamics

Overseas investors are giving While the Treasury has been changing the auction calendar gradually, demand composition
way to domestic investors in the has shifted rapidly in the past year. Figure 9 tabulates net coupon Treasury purchases from
Treasury market various investors. The main point is that overseas investors, which used to absorb almost
the entire coupon supply, have given way to domestic investors. In 2007 and 2008, the
foreign sector absorbed the entire coupon supply, with Treasury holdings among domestic
investors absorbing only 25%. The slack has been picked up by the Fed, which absorbed
another 25% through its asset purchase program, and households, which bought the largest
Figure 9: Coupon Treasury supply/demand: Overseas investors giving way to domestic investors
Net coupon Foreign Federal Money Banking Pension/ Broker
issuance sector Reserve market Households* sector insurance dealers Others
2007 183 201 -3 21 -133 15 41 0 42

2008 377 378 -42 49 -230 -25 45 218 -16
Q1 2009 300 71 16 -12 216 24 7 -52 29
Q2 2009 366 103 164 2 37 12 7 6 34
Q3 2009 406 104 113 3 99 58 16 -19 31
Note: Net coupon purchases have been estimated by subtracting net bill purchases from total Treasury purchases for each investor. Unaccounted bill demand has
been assumed to be absorbed by households. * The household category, in addition to individuals, includes domestic hedge funds and non-profit organizations such
as charitable organizations, private foundations, schools, churches, labor unions, and hospitals. Analysis Date: September 30, 2009. Source: Flow of Funds, Federal
Reserve H4 data, ICI, New York Fed, Treasury, Barclays Capital
8 January 2010 36
share at close to 40% (the household category, in addition to individuals, includes domestic
hedge funds and non-profit organizations). The remainder has been split among the
banking sector and other intermediaries such as mutual funds and retirement accounts.
With net coupon Treasury supply set to increase to $1.8trn in 2010 and the Fed no longer
buying, the question is whether overseas investors and households will continue to
purchase Treasuries, as well as whether the banking sector will step up.
As was evident in 2009, supply-demand dynamics in the Treasury market alone do not drive
rates and the curve. The supply-demand imbalance in other fixed income sectors, such as
mortgage-backed and corporate bonds, must be accounted for as well. We discuss this in
detail in the article “US fixed income supply/demand: Supply deluge.” Our main conclusion
is that the Treasury-led sharp increase in the supply of term fixed income securities in 2009
was offset by the Fed’s asset purchase program. The amount of supply hitting the market
was actually less than in prior years. In 2010, the net term supply of fixed income securities
is likely to stay as high as in 2009; however, in the absence of Fed buying, the market will
have to absorb close to $2.6trn in net term fixed income supply and $3.3trn in 10y
equivalent terms; both much higher than pre-crisis levels.
Overall, the sharp swing in net Private demand is likely to pick up, but not enough to fill the void created by the Fed’s stepping
fixed income supply from 2009 away. We expect overseas investors and households to continue to deploy foreign exchange
to 2010 should overwhelm the reserves and savings, respectively. The terming out of short-term holdings, bills, and money
increase in private demand market shares should provide a boost to the pace of purchases. Banks will likely continue to
switch from loans to securities, but we do not expect overall bank credit to increase. We
estimate the ex-ante demand shortfall for term debt to be close to $1trn in notional terms. The
bulk of this shortfall is expected to be in Treasuries, as the asset allocation out of agency
securities and into Treasuries by households is unlikely to continue, and banks will likely
continue to prefer spread products to Treasuries. Overseas should still favor Treasuries.
Curve should remain Hence, long-term rates should drift higher through 2010. 5y5y forward real rates have risen
steeper than the market is sharply from the lows of late 2008 but are still well below the levels of prior years (Figure
pricing in 10). At slightly below 2%, they are still 60bp below the average during 2005-2007 and
175bp below the average during 1990s. This is also reflected in the flatness of the forward
curve. Figure 11 shows that the market is pricing a curve flattening of 80bp over the next
one year. With the Fed still nine months away from hiking, in our view, we believe such a
Figure 10: Forward real rates still low Figure 11: Forward curve too flat
6.0 300 280
5.0 250
4.0 200
200
3.0
1.94 150 130
2.0
1.0 100
0.0
50
-1.0
Dec-95 Dec-97 Dec-99 Dec-01 Dec-03 Dec-05 Dec-07 Dec-09 0
5y5y Real Rates, % Spot Jun10 Dec10 Jun11 Dec11
2s10s Tsy Curve, bp
Source: Barclays Capital. Real rates measured using real swaps since 2005 and as Source: Barclays Capital. Forward 2s-10s Treasury curve has been computed
a difference between 5y5y nominal rates and realized 1y core inflation prior. from the swap curve assuming that spreads stay at current levels.
8 January 2010 37
flattening is unlikely; in the past five hiking cycles, the 2s10s curve has, on average,
flattened by just 10bp in the year leading into the first hike. We therefore recommend curve
steepeners.
Curvature trades
Our outlook for curvature, mainly 2s5s10s and 10s20s30s, is motivated by the evolution of
Fed policy, supply-demand imbalance, and improving conditions in the financial markets.
2s5s10s: Expect cheapening in H1 09

The 2s5s10s gross butterfly has exhibited a wide range in 2009; richening from -20bp in the
first half to +60bp at the peak of the convexity episode in the middle of the year, before
richening back to +10bp in December and finally ending the year at +35bp. We expect the
fly to cheapen to +50bp in the first half of the year before richening to the low 30s. In
addition, the fly has exhibited strong cyclicality around auctions in 2009 – cheapening going
into 5y auctions and richening again as the auction concession shifts to the long end of the
curve – which should continue this year as well.
Level of short term rates and Figure 12 plots the 2s5s10s gross fly against the output of our fair value model. The three
slopes of the 3m2y and 2y10y factors incorporated in the model are the level of 3m T-bill rates and the slopes of the 3m2y
curve should drive the fly and 2y10y curves. In addition to controlling for the level of rates, the model controls for Fed
expectations and term premium. The rationale for including the 3m2y curve is that the
market is pricing in gradual Fed moves, and if hikes are priced in over the near term on
strong economic data, some spill over beyond the 2y point, resulting in a cheapening of the
fly as the 3m2y curve steepens and vice versa. For instance, over the 1-year period leading
up to the first Fed hike in the previous hiking cycle, the fly cheapened by ~50bp as the
money market curve steepened sharply. The slope of the 2s10s curve, in conjunction with
the 3m2y curve, should control for term premium. Figure 12 shows that these factors
explain the fly to a large extent and that the current level, at 38bp, is slightly cheap.
Figure 12: Explaining the 2s5s10s fly Figure 13: Convexity paying flows cheapen the fly
100 40
80 35
60 30
40 25
34
20 20
15
0
10
-20
5
-40
0
-60
Mar-03 Jul-03 Apr-04 Feb-05 Apr-09 May-09
Dec-89 Dec-93 Dec-97 Dec-01 Dec-05 Dec-09
2s5s10s Fly Residual Change, bp in bp/yr extension
2s5s10s Fly, bp Estimated
Source: Barclays Capital. Estimated based on 3m bill rates, 3m2y slope and Source: Barclays Capital
2y10y slope.
8 January 2010 38
The fly should cheapen as the The fly should cheapen to ~45bp by the middle of this year as the Fed stays on hold amid
Fed stays on hold amid strong strong data and a supply-demand imbalance, respectively, steepen the 3m2y curve and
data and supply/demand keep the 2s10s curve at current levels, before richening to low 30s by the end of the year as
imbalance keeps the curve steep the aforementioned curves flatten when the Fed actually begins to hike.
The risk is of even cheaper levels in the first half, as our mortgage strategists expect
mortgage rates to move higher given the current tight spreads. Since convexity flows are
concentrated in the 5-10y part of the curve, an extension of the mortgage universe typically
cheapens the fly. Figure 13 shows that the fly cheapened more than implied by the model in
prior extension episodes. Even if we exclude the May-June selloff, the beta from prior
episodes suggests that the fly could cheapen 5-10bp more than implied by the above
model, or to 50-55bp, for the 1-year extension that we expect in the first half of the year.
However, we do not expect the fly to cheapen monotonously but to exhibit strong
cyclicality around auctions. Figure 14 shows that the fly cheapened an average of ~6bp
more than implied by the model heading into 5y auctions in 2009, but gave back the entire
cheapening after the auction as the concession shifted to the long end. Such price action
should continue, but with the fly cheapening gradually through each auction cycle.
10s20s30s: Expect further richening

The 10s20s30s butterfly has richened steadily through 2009 from the extreme levels
reached during the financial crisis in late 2008. At current levels, the fly is in line with market
stress metrics, but it could richen further as the Treasury continues to term out debt by
increasing issuance at the 10y and 30y maturity points.
10s20s30s is trading in line Figure 15 shows that the 10s20s30s fly cheapened sharply in late 2008 and in earlier
with market stress metrics but episodes of stress as well. The 20y sector, which is less liquid than the 10y and 30y sectors,
should richen as the Treasury cheapens in periods of high market stress as investors demand a higher discount (or higher
increases auction sizes for the yield) to own illiquid issues. Modelling the fly using market stress metrics – such as the L-FF
10y and 30y sectors basis, realized volatility, and the percentage of variance in the yield curve unexplained by the
first three PCA factors of the yield curve – explains the cheapening to a large extent. In late
2008, the fly cheapened beyond what can be explained by the model, and as we expected, it
richened gradually through 2009. The current state of these variables argues that the fly
should trade at 34bp, pretty much in line with where it is trading. With these stress variables
back to a neutral, we do not expect them to drive near-term performance; however, the fly
could richen as the Treasury terms out debt by increasing issuance of 10y and 30y.
Figure 14: 2s5s10s fly around auctions Figure 15: 10s20s30s fly appears fair
4 70 Stock market Current

3 9/11 crash crisis
2 60 Bear
1 Stearns
50
0
-1 40
-2
-3 30
-4
20
-5
-6 10
-7
-10 -8 -6 -4 -2 0 2 4 6 8 10 0
00 02 04 06 08
Days from the 5y Auction
Average Model Residual, Jan-Nov 5y Auctions, bp 10-20-30 fly (bp) Model (bp)
Source: Barclays Capital Source: Barclays Capital
8 January 2010 39
Figure 16 shows that terming out of debt is associated with the richening of the 10s20s30s
fly. We expect 30y issuance to increase to $190bn in 2010 from $130bn in 2009, an
increase of 45%. 10y issuance should also increase, by 20% or so (Figure 4). Such a large
jump in issuance at the long end would suggest that the fly should richen by 18bp. Hence,
while the fly has richened steadily through 2009, we expect the trend to continue, although
on a smaller scale.
Liquidity premium trades

Most relative value trades Most relative value trades that became dislocated in the fourth quarter of 2008 have
have corrected, but some gradually corrected and are now trading close to pre-crisis averages. However, some trades
still stand out still stand out. High-coupon Feb19s and Aug17s are trading cheap to their low-coupon
counterparts compared with other such pairs in the 7-10y sector. Also, the on-the-run 10y
is trading rich and should cheapen as it ages away from the 10y sector. 10-11y issues still
offer solid rolldown potential into the 10y sector; and REFCO P-STRIPS are still trading at a
well above pre-crisis discount to Treasury STRIPS. Their P/L potential pales in comparison
with that at the peak of the crisis, but they are still attractive relative to their pre-crisis
behavior.
A few high-coupon issues look Figure 17 shows the spread of asset swap spreads between high-coupon and low-coupon
relatively cheap; issues in the 7-10y sector; the more negative the spread, the cheaper the high-coupon is to
REFCO P-STRIPS look cheap to low -coupon issue. The illiquidity discount has declined substantially from the peak of 60-
Treasury P-STRIPS 70bp to 5-10bp, but some issues stand out. Aug17 high-coupon issues are trading 5bp
cheap to their low-coupon counterparts compared with HC May18s, which are trading right
on top of LC May18s, and May17s, which are trading 2bp cheap. Investors should consider
switching out of HC May18s into HC Aug17s. Similarly, HC Feb19s are trading 8bp cheap to
LC Feb19s, whereas the aging away from the sector implies that they should trade 4bp
cheap. The on-the-run 10y (or LC Nov19) is also trading rich on the curve, even though it
has little repo specialness and will be reopened for the second time next week. 2021-22
issues should richen as they roll toward the 10y sector; they offer 4bp of rolldown potential
over a 1-year period.
Figure 16: 10s20s30s fly versus long-end supply Figure 17: Low-coupon vs high-coupon issues
70 0
60 -2
50 -4
10-20-30 (bp)
40 -6 -5
30 -8
2 -8
R = 0.4656
20 -10
10 -12
-12
LC-HC Spread of Asset Swap Spread, bp
0 -14
-15 -10 -5 0 5 May Nov May Aug May Nov Feb Aug Nov
yoy change in average maturity of debt (months) 16s 16s 17s 17s 18s 18s 19s 19s 19s
Source: Barclays Capital, Haver Analytics Source: Barclays Capital
8 January 2010 40
Similarly, REFCO P-STRIPS still look attractive relative to their Treasury counterparts, as they
have richened from their peaks but are still trading far from pre-crisis levels. Figure 18
shows that Oct19 REFCO P-STRIPS are trading 40bp above Treasury STRIPS, compared
with the pre-crisis average of 15bp. The mispricing further out the curve is not that high,
but even Apr30 REFCO-STRIPS look 13bp cheap. We therefore continue to recommend
overweighting REFCO STRIPS versus Treasury P-STRIPS.
Introduction of the ultra-long The introduction of the ultra-long bond contract next week should also lead to some
should cheapen the 15yr sector interesting trades, which we discuss in detail in “A new bond contract may change
dynamics.” In summary, the 15y sector should cheapen relative to the 25y sector as liquidity
moves further out the curve. This should also manifest in Ps’ cheapening to Cs in that
sector. Figure 19 shows that Ps are trading 7-10bp rich to Cs in the sector, particularly
Nov24s, compared with an average of 2-5bp for shorter maturity pairs.
Hence, while 2009 proved to be an easy ride for relative value trades, in 2010 investors will
have to pick and choose opportunities and accept much smaller returns, hopefully for much
smaller volatility as well.
Figure 18: REFCO vs Treasury P-STRIPS Figure 19: Cs vs Ps in the 10-15y sector
80 72 67 10
65 67
62 63
70 9
60 8
50 40 7
34 31 34
40 28 28 6
30 5
20 4
15 3
10
2
0
1
RFCO RFCO RFCO RFCO RFCO RFCO
0
Oct 19 Jul 20 Oct 20 Jan 21 Jan 30 Apr 30
Aug Nov Aug Nov Feb Aug Nov Feb Aug Feb
RFCO-Tsy Ps,bp Peak during Crisis, bp 21 21 22 22 23 23 24 25 25 26
Pre-Crisis Average, bp Cs-Ps, YY spread, bp
8 January 2010 41
INFLATION-LINKED
TIPS: Still cheap insurance

Michael Pond Although outright TIPS returns may be weak as real rates rise, relative performance
+1 212 412 5051 should be positive and breakevens still offer cheap inflation insurance.
michael.pond@barcap.com
Structural allocations, the Treasury’s commitment, the potential for a US VAT, and food
inflation risks should be supportive for breakevens
Chirag Mirani
+1 212 412 6819 Market reaction to rents and OER CPI may be a drag and the Fed’s reaction function to a
chirag.mirani@barcap.com rise in breakevens could limit upside potential
“By a continuing process of inflation, governments can confiscate, secretly and unobserved,
an important part of the wealth of their citizens” – John Maynard Keynes, The Economic
Consequences of the Peace
What's on the horizon for returns?

Expect real yields to rise slowly After a rough 2008, 2009 was a great year for TIPS. The Barclays Capital TIPS index gained
as the Fed begins to reverse 10.5% on the year after falling 1.7% the year before. Relative performance was even
some of its stimulus stronger because a basket of nominal comparators fell 5.7% in 2009. We do not expect
these gains to be repeated in the year ahead. We have argued that the Fed, through its
quantitative easing program, has pushed 10y real yields down 70-80bp. As the Fed ends its
asset purchase program and begins to reverse some of the stimulus put in place when the
economy and markets were in a crisis, real yields should slowly rise.
Combining our interest rate forecasts with our economists' projection of about 1.7% NSA CPI
inflation accretion next year, we find that only TIPS out to TIIJul12s have a positive 1y horizon
return and 20y TIPS are expected to lose 4-5% (Figure 1). However, we expect nominal yields
to rise even further as Treasury supply pressure continues and the market reacts to positive
economic data and medium-term inflation risks. Our forecasts imply breakeven total returns
of 0.9-1.2% from the 2y to the 10y sector and 1.3-1.9% in the long end. This implies that from
an outright return or risk-adjusted breakeven return, investors should favor the 3-5y sector.
Figure 1: 1y horizon returns using Barclays Capital forecasts
3%
2%
1%
0%
-1%
-2%
-3%
-4%
-5%
-6%
Apr 10
Apr 11
Apr 12
Apr 28
Apr 29
Apr 32
Jan 11
Jan 12
Jul 12
Apr 13
Jul 13
Apr 14
Jan 14
Jul 14
Jan 15
Jul 15
Jan 16
Jul 16
Jan 17
Jul 17
Jan 18
Jul 18
Jan 19
Jul 19
Jan 25
Jan 26
Jan 27
Jan 28
Jan 29
TIPS Total Return Breakeven Return

8 January 2010 42
The right allocation is not zero

Expect real rates to lag the move We expect real rates to lag the move higher in nominals – particularly in the first half of the
higher in nominals as structural year – in part owing to expectations that structural inflows will continue to come into the
allocations continue asset class. A broad base of real money investors, including foreign central banks, domestic
institutional accounts and retail investors, helped push 10y breakevens up about 230bp over
the past year, and we believe that support will continue.
Before 2009, foreign central bank involvement in the TIPS market was largely tactical. There
has been a clear shift, though, to using the asset class as a de facto currency hedge, as
insurance against inflation relative to nominal Treasuries, and simply as a diversification tool
within a broad portfolio of assets.
Domestic accounts are also realizing that they should have at least some of their portfolio
hedged against inflation because of the medium-term risks coming from stimulative
monetary and fiscal policies. We are seeing investors shifting from 0% to 2%, 0% to 5%, or
0% to 10%, but regardless, there is growing consensus that a 0% weight is not the optimal
one. Along these lines, investors are wondering whether TIPS should be part of a broad high
grade index. There are pros and cons regarding this issue; however, many investors appear
to be moving toward an Aggregate Index type weight (they would be about 4%) regardless
of whether they are actually in the index or not. Our own efficient frontier/optimal portfolio
analysis using TIPS returns over the past ten years indicates that including TIPS in a well-
diversified portfolio pushes the efficient frontier to the left, reducing risk and increasing
overall returns. During this time-frame, TIPS delivered the highest annualized total returns
among the asset classes noted in Figure 2.
Figure 2: Asset class returns and volatility through December 2009

S&P
High 500 Global
TIPS Treasury Yield Corporate MBS TR US Inflation
Index Index Index Index Index Index GSCI Aggregate Index
1y 10.5% -3.6% 58.2% 18.7% 5.9% 26.5% 13.5% 5.9% 13.6%

3y 6.7% 6.1% 6.0% 5.7% 7.0% -5.6% -6.9% 6.0% 4.6%
Annualized Returns
5y 4.6% 4.8% 6.5% 4.6% 5.8% 0.4% -3.0% 5.0% 4.9%
10y 7.7% 6.2% 6.7% 6.6% 6.5% -0.9% 5.1% 6.3% 7.8%
Since Inception 6.7% 6.1% 6.3% 6.4% 6.4% 4.5% 2.6% 6.2% 7.7%
1y 7.9% 5.5% 13.0% 6.8% 2.5% 22.3% 24.5% 3.3% 10.5%
Annualized Vol 3y 8.7% 5.7% 17.2% 9.0% 3.3% 19.9% 31.1% 4.2% 11.8%
of monthly returns 5y 7.2% 4.8% 13.5% 7.4% 3.0% 16.0% 28.2% 3.7% 9.8%
10y 6.7% 5.1% 11.5% 6.3% 2.9% 16.1% 25.6% 3.8% 8.6%
Since Inception 6.0% 4.8% 10.4% 5.9% 2.8% 16.6% 24.6% 3.7% 8.0%
AnnualizedSharpe Ratio 10y 1.15 1.22 0.59 1.04 2.23 -0.06 0.20 1.65 0.91
Since Inception 1.10 1.26 0.60 1.07 2.29 0.27 0.11 1.67 0.96
Note: Since inception data are from March 1997. Source: Barclays Capital
Retail TIPS mutual funds had strong inflows flows in 2009 as investors became more
concerned about medium-term inflation risks. These flows may continue in the first half of
the year as individual investors are drawn in by historical performance, but if real yields rise
as we expect, that cash flow may dry up or even reverse. However, structural flows can
continue from foreign and domestic institutional accounts because their valuation decisions
are typically on a breakeven basis, rather than on outright performance expectations.
8 January 2010 43
No fear of commitment
Starting with the August refunding announcement, it was clear that the Treasury had
Treasury has recommitted to the recommitted itself to the TIPS program. This is important because investors and other
TIPS program; demand should market participants might be hesitant to enter into an asset class in which there were
easily absorb our expected gross concerns over its longevity. That should no longer be a concern, in our view.
supply of $80-$85bn in 2010
One of the ways the Treasury is showing its commitment and trying to improve the liquidity
of this market is by gradually increasing auction sizes. In 2004, when it showed its
commitment to the program by expanding issuance maturity points and increasing sizes,
TIPS had one of their best performance years on an outright and relative basis, and average
daily trading volume nearly doubled.
The Treasury is taking care not to overwhelm the market with too much TIPS supply, and
we expect demand easily to absorb the additional gross supply of $22-27bn and have noted
that net issuance will actually decline despite the increase in gross issuance (Figure 3). For
the past couple of years, auction sizes have been steady at $8bn on new issues and $6bn at
reopenings. The Treasury increased the October reopenings to $7bn, and we expect
continued gradual increases in sizes throughout 2010, with additional auctions added to the
schedule in the second half for total gross issuance $80-85bn. However, if the Treasury
believes continued structural demand can support greater supply than this, we expect it not
to hesitate to increase issuance more rapidly.
Figure 3: Annual TIPS issuance
90
75
60
45
30
15
0
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
(expected)
Total Gross Issuance ($bn, Excluding SOMA) Net Issuance Ex-SOMA

Source: US Treasury, Barclays Capital
Treasury will reintroduce 30y The Treasury will also reintroduce issuance of 30y TIPS at the end of February in order to
TIPS at the end of February; we capture a broader demand base and potentially improve liquidity at the long end. We expect
expect the first auction to that first auction to go very well because we expect there is significant pent up demand for
go very well a long duration real instrument.
VAT (Very Attractive for TIPS?)

One factor that poses upside risk to TIPS performance is the potential for the introduction in
the US of a value added tax (VAT). Congress and the administration are looking for ways to
show that, at least eventually, fiscal discipline will bring deficits down. One proposal being
discussed by various policy makers is the implementation of VAT similar to that already in
8 January 2010 44
existence in about 130 countries. 6 A VAT is seen as an efficient source of revenue mainly
because it applies to a broad taxpayer base and is not easily evaded or avoided. Although
the main argument against one is its regressive nature, Democratic leadership in the House,
Senate, and the administration have all expressed a view that it should be on the table in
discussions about deficit reduction and fiscal reform.
Introduction of value added tax Implementation would give TIPS investors at least a one-time windfall because the tax would
(VAT) would give TIPS investors be captured in the CPI. A CBO study concluded that “adopting a 5 percent VAT whose tax
at least a one-time windfall base included 60 percent of consumption would initially increase the CPI by about 3 percent.” 7
It would likely apply to less than 100% of the CPI basket because some categories, such as
food at home, would probably be taxed a lower or even zero rate to make it less regressive and
because it is difficult to measure or tax value-added in some services categories. For example
in Europe, while most goods and services are taxed at the standard rate, reduced and super-
reduced levies are applied based on largely on social merit. For instance, Spain taxes shoes and
leather goods at the standard 16% rate, but subjects social services and social housing to 7%
and 4% rates, respectively. In addition, some goods and services are zero-rated (exemption
with refund for tax paid at the preceding stage) and exempted. For example, the UK zero-rates
water supplies and medical equipment for disabled persons and exempts medical/dental care.
The average standard VAT rate in Europe is about 18% (Figure 4), and we estimate that the
weighted average rate on the CPI basket would be about 9% if the US were structured similar to
Europe, rather than the 3% provided as an example in the CBO study. It could also be lower
though if pass-through rates to consumers are low or more categories had zero tax rates.
However, even if a one-time lift to the CPI index was just 3%, this would be worth about 30bp
to 10y breakevens and 60bp to 5y breakevens. From a timing perspective, we would not expect
implementation of a VAT while there is still concern about the sustainability of consumer
spending and the health of the economy, so such a levy, if it were to be enacted, may still be
years away. However, with Congress likely wanting to show some fiscal discipline before the fall
elections, legislation may go through well before implementation. Investors should watch this
debate closely in the months ahead because the market will likely price in a VAT well before it
affects the CPI index.
Figure 4: VAT rates in Europe

Super Reduced Reduced Standard
UK 5% 17.5%
Germany 7% 19%
Italy 4% 10% 20%
Spain 4% 7% 16%
Holland 6% 19%
France 2.1% 5.5% 19.6%
Avg. 3.4% 6.8% 18.5%
Source: European Commission
Food, glorious food

Another factor that TIPS investors should be watching is the potential for a spike in food
CPI. The surge in agricultural futures in 2007 and the first half of 2008 caused food CPI to
6
http://www.nytimes.com/2009/12/11/business/11vat.html?pagewanted=1&em&adxnnl=1&adxnnlx=1260620335-
CMLGz7NjjY4HHXvBAq0a5A.
7
http://www.cbo.gov/ftpdocs/102xx/doc10288/1992_02_effectsofadloptingavat.pdf.
8 January 2010 45
Figure 5: Crops are rising up Figure 6: Livestock prices are coming back
300 140
135
250
130
200
125
150 120
115
100
110
50
105
0 100
Oct-06 Apr-07 Oct-07 Apr-08 Oct-08 Apr-09 Oct-09 Jan-06 Oct-06 Jul-07 Apr-08 Jan-09 Oct-09
Cocoa Corn Rice Sugar Prices received by farmers: livestock and products
Source: Bloomberg Source: Haver Analytics
increase 6.3% y/y in October 2008. The lag was due to base effects, as well as the
processing time from raw to retail for many food items. Many agriculture futures declined
along with energy in the second half of 2008; as a result, food CPI came in at -0.6% y/y in
the November reading.
Food price pressures have now been building the other way. One of the biggest movers has
been sugar, where futures are up about 130% since April, but cocoa is up about 40% over a
similar period and corn prices have risen 40% since the September 2009 lows (Figure 5).
While rice prices are important in many emerging market economies, corn is important to
US food prices because it and its by-products are used heavily in the retail food chain. Our
commodity group expects emerging strength in grains, particularly corn, as feed demand
picks up and the EPA makes a decision this year on whether to increase to the US ethanol
blend rate from the current 10%. In addition, after decades of being a large exporter,
China’s trade balance in corn continues to reflect increasing imports and decelerating
exports, and our commodities research group expects this trend to continue. Price
Food price pressures have been
pressures appear to be building in categories other than crops as well, and the US
building; our commodity group
Department of Agriculture’s index of prices received by farmers on livestock and related
expects increases in grain prices
products was up 9.3%, unannualized, from September through December (Figure 6).
Rents – Fly in the ointment or canary in the coal mine?

One factor that could hamper TIPS performance in the near term is investors’ reaction to
a continuation of the recent trend in shelter inflation (Figure 7). Rents and owners’
equivalent rent (OER) are certainly important categories within CPI because they make up
about 40% of core. After the trend in rates averaged above core and about 3% from the
beginning of 2008 through the first half of 2009, the 3m annualized trend through
November was -0.9%, and that on the more heavily weighted OER was -1.1%. Our
economics group expects these components to remain soft over the coming months
Soft OER/rent related breakeven
because of lags, but looks for other components to provide a positive offset and for y/y
weakness should be seen as a
core of about 1.1% for 2010. However, the market may extrapolate the trend in shelter
buying opportunity
inflation further than we expect, which could dampen TIPS performance. However, if it
does spook the market and leads to a pullback in breakevens, we would view that
cheapening as an opportunity to add to existing longs because we believe the positive
factors discussed above will dominate trading in 2010.
8 January 2010 46
Figure 7: Rent and OER CPI (y/y % chng)
5%
4%
3%
2%
1%
0%
Aug-04 May-05 Feb-06 Nov-06 Aug-07 May-08 Feb-09 Nov-09
OER Rents
Source: BLS
The Fed watching the markets watching the Fed

Breakevens will likely rise in a While we expect breakevens to rise, we expect them to do so in a relatively modest fashion,
modest fashion when compared particularly compared with performance over the past year. One reason is the potential for
with last year, as a sharp rise this higher breakevens to provoke a reaction from Fed officials. Forward breakevens have risen
year may provoke a hawkish nicely over the past several months, to the upper end of historical ranges (Figure 8). Since
tone from Fed early October, our measure has gone up 50bp, and the Fed’s version has increased 40bp.
Despite the increase, Fed officials and statements have continued to describe longer-term
inflation expectations as stable. While we believe longer breakevens have plenty of room to
move higher before the Fed considers them “unmoored,” there is a risk that a continued
sharp rise might cause a hawkish response from some Fed officials, which may limit how
much breakevens can increase this year.
Figure 8: 5y5y forward breakevens
4.0%
3.5%
3.0%
2.5%
2.0%
1.5%
Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10
BarCap 5y5y BEI Fed's 5y5y BEI
Source: Federal Reserve, Barclays Capital
8 January 2010 47
AGENCIES
The long and winding road

Rajiv Setia We examine the scope of potential changes to the GSEs but do not believe anything
+1 212 412 5507 substantive will be decided in 2010. A more pressing concern is finding sources of
rajiv.setia@barcap.com demand to replace the exiting Fed, as net term supply should remain brisk.
Resolving the debate on the GSEs will likely take several years as Congress hammers out
James Ma
the future of US housing finance, preserving the status quo in the meantime.
+1 212 412 2563
james.ma@barcap.com By raising the portfolio and debt caps, and removing the limit on equity infusions
through 2012, the Treasury has taken immediate nationalization off the table.
Despite higher cap limits, FNM/FRE portfolio growth should flatline unless the
MBS/agency basis widens materially. FHLB advance activity should stabilize.
We expect net term issuance of $100bn, driven by term out. The term funding mix
should shift toward callables early in the year as excess liquidity spurs bank demand.
With $140bn in maturing bellwethers, the funding calendar should remain active.
We believe a spread widening of 10-15bp across the bullet curve is likely from YE09
levels. However, cheapening should gain momentum only when the Fed starts to drain
liquidity from the system. In the early going, bank demand should offset reduced money
manager appetite, keeping spreads in check.
3y paper appears cheap to surrounding sectors; we also recommend TLGP names and
bank debt guaranteed by selected foreign governments. We are more cautious on
callables, but believe short-lockout, short-maturity structures make sense for investors
less bullish on the economic recovery.
Future of the GSEs

We expect the administration to Questions about the fate of Fannie Mae (FNM) and Freddie Mac (FRE) broadly center on
maintain the status quo at two different timetables. In February, in conjunction with the President’s FY11 budget, the
FNM/FRE for several years administration is expected to unveil its vision for the GSEs. As we have detailed in prior
publications, 8 we believe that the status quo will hold for several years:
Despite lacking an explicit guarantee and likely posting quarterly losses for the next few
years, FNM/FRE still have a smoothly functioning guarantee business. In the near term,
there simply is no viable alternative to the GSEs for housing finance.
While conservatorship technically implies that FNM/FRE are being healed slowly and run
for profit, it is clear that Congress will use FNM/FRE for public policy purposes in the
near term, even if such a decision is not economic. In fact, Representative Barney Frank
recently stated that the entities already “have become kind of public utilities.” 9
Longer term, regardless of what the administration outlines, it is Congress that must decide
the extent of government involvement in housing policy and what role, if any, the GSEs or
their successors play in fulfilling their vision. We expect the debate to be long and
protracted and would not be surprised to see it stretch out over many years.
8
“GSEs: Back to the future,” 11 December 2009
9
Interview on CNBC, 5 January 2010
8 January 2010 48
Treasury backstops: ?00 billion dollar baby

The Treasury has greatly One of our concerns about the GSEs in late 2009 centered on the adequacy of the preferred
strengthened capital support for capital backstop in place to absorb future losses. The ability of Treasury to alter the size of
the GSEs the backstop was due to expire at YE09, but in a show of solid support for debt-holders,
Treasury altered the terms of the Preferred Stock Purchase Agreements (PSPAs) radically on
Christmas Eve:
The maximum amounts of preferred stock FNM/FRE can draw from Treasury are now
the greater of $200bn, or $200bn plus the cumulative amount drawn from 2010
through 2012. Any extant surplus at December 31, 2012, will reduce the cumulative
limit by that amount.
Instead of forcing FNM/FRE to shrink their portfolios and debt outstanding by 10% per
year starting in 2010, the limits on FNM/FRE’s portfolios and debt will shrink by 10% per
year. So, FNM/FRE’s portfolios will be capped at $900bn as of December 31, 2009, then
$810bn as of December 31, 2010, and so on. Debt is still limited to 120% of the
portfolio cap. We address the implications for debt supply later in this article.
In our view, these actions mitigate the tail risk of an unplanned GSE receivership in a stress
scenario:
The cumulative preferred stock backstops expand with any new draws over the next
three years. This time window is relevant because the bulk of credit-related expenses
will be provisioned for by then, and there will still be $140bn plus in capital support per
entity left to absorb potential losses beyond 2012.
GSE credit should no longer be a concern for investors for the foreseeable future.
Although Treasury pushed back the commitment fee by a year, keeping the senior
preferred coupon at 10% ensures that lower portions of the capital structure do not
benefit from the extraordinary support the government has extended to debtholders. It
also ensures that conservatorship will last as long as the government wants it to.
Relaxing portfolio limits also ensures that the GSEs will not be forced to sell existing MBS
holdings into a market already dealing with the aftermath of the Fed’s QE program.
Although we find the possibility unlikely, it also allows the GSEs to grow in 2010, serving
as a backstop bid for MBS in the event of inordinate market stress.
If our estimates 10 are accurate, draws on the PSPA will continue for the next few years. In
our base case scenario, we expect cumulative draws of $130bn and $100bn at FNM and
FRE, respectively; in a stress scenario, required infusions may rise to $180bn and $130bn,
respectively. To the extent the GSEs are used to accomplish non-economic policy initiatives
going forward, more support might prove necessary.
So what actually happens in February?

Nationalization is unlikely
In light of our loss estimates, it should not be surprising that we expect conservatorship to
be indefinite. So the question remains when, not if, the GSEs will be put on the
government’s balance sheet, as suggested by OMB Director Peter Orszag. While we expect
the GSEs to come onto the government balance sheet eventually, we disagree with the
notion that this is to be expected in February:
10
“GSEs: Back to the future,” 11 December 2009
8 January 2010 49
Balance sheet assets and liabilities at FNM/FRE would be subsumed into government
assets and liabilities, similar to TVA. Also, the guarantee businesses would be included
at their net present values, similar to FHA. Given how much difficulty Congress is having
in raising the debt limit just to keep the government operating, we doubt that an
additional $2trn increase in the ceiling is even on the radar screen (Figure 1).
Adding $2trn to the public debt would increase the debt-to-GDP ratio of the US from 54%
to about 65% immediately (Figure 2). Even without the GSEs, CBO projects debt-to-GDP
to rise to 80%-100% over the next decade, which at some point could expose the US to
the possibility of a ratings downgrade. Waiting 5-10 years would reduce the risk: not only
would the portfolios be more tractable after they shrink, but GDP would also be far larger.
Heading into midterm elections, incumbent Democrats in Congress may be unwilling to

give their opponents ammunition to attack a perceived lack of fiscal responsibility.
While this may seem prosaic, talent retention is an important consideration. As part of
the government, FNM/FRE would have a difficult time retaining or attracting talent to
manage the $1.7trn portfolios and trillions in associated derivative hedges.
Immediate nationalization is off As outlined earlier, this rationale does not change the endgame of putting the GSEs into
the table, but could still be the receivership eventually, wiping out the existing preferred and common equity holders.
endgame further down the road Senior debt holders and MBS holders should have no worries even in this scenario, given the
pledge made by the government with the PSPAs.
If the status quo holds, what else could the administration announce?
Our sense is that the administration will not change the status quo with its announcement
in February; there is every indication that the agency MBS market is not only functioning but
also vital to the securitization of conforming loans, and drastic changes could threaten the
already-fragile recovery.
The administration could therefore try to frame the discussion of the GSEs’ structure several
years out by announcing the following:
A firm timeline for the end of conservatorship; e.g., a sunset possibly 7-10 years from
now. This would put pressure on Congress to make a decision on the future of housing
finance before too long, and highlight that the administration is not simply kicking the
can down the road.
Figure 1: Politics limits headroom to raise debt ceiling Figure 2: Rising debt/GDP may pose sovereign risks
$trn % of GDP
18 90%
16 80%
14 70%
12 CBO projection
60%
10
50% 54% as of
8
11/30/09
6 40%
4 30%
Jan-96 Jan-99 Jan-02 Jan-05 Jan-08 Jan-11 Sep-06 Sep-09 Sep-12 Sep-15 Sep-18
Debt Subject to Limit Plus FNM/FRE Limit Public debt Plus FNM/FRE
8 January 2010 50
Reduction of FNM/FRE risk weight from 20% to 0-5%, which would serve as another
show of support, but stop short of nationalization. Recall that in late 2008, the Office of the
Comptroller of the Currency discussed reducing the risk weight for FNM/FRE to 10%, but
the proposal was shelved. Also, in October 2009, the risk weight on TLGP paper was
reduced from 20% to 0%, in light of the sovereign guarantee backing that paper. Note that
in this scenario, we do not think that risk weights on other GSEs will change.
Endorsement of alternatives to traditional mortgage finance, including covered bonds,

through a legislative framework.
A few favored options for restructuring FNM/FRE, letting Congress sort it out.
Regulatory changes: There’s too much confusion, I can’t get no relief

Financial regulatory reform Reports have also emerged that House Financial Services Committee Chairman Frank is
poses outside risks to the GSEs pushing GSE reform to the forefront of his 2010 agenda. However, any serious effort on this
topic will likely be deferred until legislation dealing with healthcare reform and financial
regulatory overhaul (H.R. 4173) is enacted. Our view is that the debate between
nationalization of the GSEs and other alternatives 11 will be contentious and play out over a
five-year plus timeline, and any dramatic shifts in housing finance will only occur over a
decade or longer.
The advent of prescriptive liquidity regulations for UK banks may foreshadow similar measures
being adopted in the US 12. While we expect liquidity regulations to enhance the appeal of GSE
MBS and debt for bank portfolios, the real changes for banks will likely occur on the liability
side of the balance sheet. If liquid holdings are required to be unencumbered, as in the UK,
banks may need to make major changes in their funding mix, which could permanently
reduce reliance on wholesale funding, including FHLB advances.
The House version of the Wall Street Reform Act also bears careful watching as it poses
significant risks for secured creditors. If enacted, it would allow a new financial regulator to
supersede the interests of secured creditors and enforce a haircut on their secured claims,
which would be replaced with an unsecured claim. The legislation exempts trades with more
than 30 days original maturity, and has a carve-out for debt issued by the Federal government
and its agencies 13. Importantly, FHLB advances would also be exempt from this requirement.
The Senate version of the legislation does not have any such provisions for secured lending,
and it remains to be seen what form the final legislation will take.
Term supply in 2010: More than we bargained for

For 2010, net GSE supply should hinge on two main factors:
The extent to which FNM/FRE use their newly granted cap room to grow, or at least not
shrink the retained portfolios, and
The amount and pace of liability term out at all three major GSEs.
Under the prior PSPA guidelines, both FNM and FRE would have been required to shrink
their portfolios by 10% of YE09 levels, or around $75bn each, to a target level of
approximately $675bn. Under certain rate scenarios, this shrinkage may have been difficult
to achieve without necessitating sales of existing holdings. Further, forced GSE sales would
11
See complete discussion of alternatives in Part II of “Back to the Future”
12
“The Impact of New Bank Liquidity Regulations on the Fixed Income Landscape,” 16 October 2009
13
”Short rates: Choppy seas ahead,” this annual
8 January 2010 51
have been procyclical and highly disruptive in a market environment already searching for
equilibrium after the Fed’s purchase programs end in Q1 2010.
Although Treasury has changed We expect the changes to the PSPAs announced in December 2009 to result in more stable
how it calculates the portfolio portfolio sizes, but not necessarily outright growth. Under the new guidelines, the GSEs can
caps, we don’t expect growth carry portfolio balances of $810bn and debt of $972bn at YE10. 14 Given FNM/FRE’s current
portfolio size of about $750bn, in theory, both entities can grow their portfolios by $60bn
each (Figure 3).
However, barring a major widening in MBS spreads, growth is unlikely. Treasury’s stated
goal was to provide the GSEs “with some additional flexibility to meet the requirement to
reduce their portfolios,” and it “does not expect Fannie Mae and Freddie Mac to be active
buyers to increase the size of their retained mortgage portfolios, but neither is it expected
that active selling will be necessary to meet the required targets.” 15
The case against growth is mainly economic

Despite being in conservatorship, the GSEs have behaved in an economically consistent
manner in some respects. In 2009, FNM/FRE were net sellers of agency MBS despite having
plenty of room to grow under the prior cap ($900bn). As the Fed’s QE richened MBS
valuations beyond where the GSEs could effectively fund, FNM/FRE allowed agency MBS
balances to pay down (Figure 4).
Presently, this basis has improved somewhat, and current-coupon MBS valuations at L-5bp
are only modestly wider than spreads on a blended GSE funding level incorporating
callables and shorter maturity bullets. As a result, absent a sustained 20-25bp cheapening
of the MBS/agency basis, we expect the portfolios to flatline in 2010.
As far as the FHLB system is concerned, it clearly displayed its countercyclical prowess
during the credit crisis, providing funding to member institutions via advances as credit
conditions tightened throughout 2007-08. As markets normalized in 2009 and the Fed and
FDIC supplied the markets with liquidity, FHLB lending took a backseat. Banks have elected
to pay down wholesale funding rapidly, and advances outstanding are now back to
Figure 3: Effect of raising the GSE portfolio caps Figure 4: Portfolios shrank in 2009 even with cap room
Retained portfolio, $bn $ bn bp

900 60 80
60
850 40
40
20 20
800
0 0
750 -20
-20
-40
700 -40
-60
650 -60 -80
Dec-07 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Apr-06 Dec-06 Aug-07 Apr-08 Dec-08 Aug-09
Portfolio FNM Portfolio FRE Change, Agency MBS in Portfolio (L)

Old Limit FNM Old Limit FRE Average 5y Agency ASW, fitted (R, 2m lag)
New Limit Both Average Agency MBS LOAS (R, 2m lag)
14
“Treasury changes Preferred Stock Purchase Agreements,” US Agencies Intraday Comment, 25 December 2009
15
“Treasury issues update on status of support for housing programs,” US Treasury press release, 24 December 2009
8 January 2010 52
pre-crisis levels. For full year 2010, we also expect balance sheet growth to be flat, although
over the next few quarters shrinkage may continue, as banks remain flush with cash.
The case for term out is both economic and prudential

The GSEs should continue to We expect all three major GSEs to continue terming out their liabilities, following the pattern
term out their funding to reduce of 2009 but to a lesser extent. In 2009, the GSEs made a significant push to term out their
rollover risk liabilities, particularly after having their access to the capital markets cut off in late 2008.
Also, recall that for a large part of 2008 and 2009, the GSEs were able to issue short-term
debt at considerably better funding levels than term; now, with the collapse of front-end
Libor rates, term funding levels are practically even with discount notes’ (Figure 5).
Furthermore, we believe there is a significant prudential reason for the GSEs to term out
their funding and reduce their rollover risk. Although FNM and FRE in particular have made
strides in lowering the current portion of their funding (i.e., discount notes plus the part of
term debt maturing in one year), they are still well above pre-crisis levels (Figure 6).
More prosaically, but equally important, is that failing to term out debt in a favourable
market environment when every other prudent issuer (including the US Treasury) is terming
out debt is sure to raise eyebrows.
As such, we expect FNM/FRE to term out roughly 3% of total liabilities in 2010. We look for a
closer-to-5% term out at FHLB, which has farther to go given its prior heavy reliance on 18m
floaters.
Net long-term issuance could Putting together our projections for zero growth and continued term-out at all three major
total $100bn at the three GSEs, we project $100bn of net long-term debt issuance, with about $25bn each at FNM
housing GSEs and FRE, and $50bn at FHLB. Note that this increase in net term issuance should be
countered by a corresponding decrease in short-term debt outstanding, given our
expectations for flat overall portfolio growth (and net issuance).
In order to determine the effect on the bellwether market, we must further decompose our
projected $100bn in net term supply into bullets and callables. Similar to the imbalance in short-
versus long-term funding, which the GSEs are still working through, callables have become a
much smaller portion of the GSEs’ funding mix in recent quarters (Figure 7). With callable
funding levels currently rich to bullets on an LOAS basis, we expect all three GSEs to continue
Figure 5: Term and discount note funding levels converge Figure 6: Current portion of term debt remains elevated
bp % total debt maturing in 1y
100 70%
50 65%
0 60%
-50 55%
-100 50%
-150 45%
-200
40%
-250
35%
-300
30%
-350
Dec 05 Dec 07 Sep 08 Mar 09 Sep 09
Jan-07 Sep-07 May-08 Jan-09 Sep-09
3m Agency ASW 3y Agency ASW FRE FNM FHLB
8 January 2010 53
emphasizing callables within their funding mix, especially in 1H10. We reckon that of the $100bn
in term supply we anticipate, around $80bn will be funded via callables, and $20bn in bullets.
Sizable bellwether maturities in Our revised net supply expectations, coupled with the fact that the three GSEs will likely replace
2010 could make placements most of the $142bn in bellwether bullet redemptions slated for 2010, lead us to project a brisk
more difficult pace of bellwether bullet issuance this year. In previous years, gross issuance of bellwethers
remained at about $120bn, only leaping up to $160bn in 2009 with the advent of the Fed’s QE
program (Figure 8).
Of course, in 2010, we expect placing $140bn plus in bullet supply to be far more of an
uphill battle. As we will detail in the next section, we do not expect this to occur without a
substantial widening of agency-Treasury and agency-Libor spreads from current levels.
Handicapping investor demand

Who will replace the Fed?
It is hard to overstate the importance of the Fed purchase program for agency debt in 2009.
Even though the GSEs came to market with $160bn of bellwether supply, the Fed removed an
almost-similar amount by making $145bn of purchases. As a result, the market had to take
down only $15bn in bellwethers, setting the stage for a sustained tightening in spreads. In
addition, Fed involvement encouraged previously dormant investors to re-enter the agency
market, especially domestic money managers. This influx of domestic demand helped to
offset the credit- and uncertainty-driven exit of foreign/official investors (Figure 9).
In 2010, the Fed will have just $15bn of capacity to buy agency bellwethers, versus our
expectation of bellwether bullet supply of $140bn plus, which places the onus squarely on
three different investor classes to step in and replace the lost demand:
Money managers: At the forefront of demand for agencies in 2009 (Figure 10), these
investors are the most valuation-sensitive of potential Fed replacements. With agency-
Treasury spreads very snug across the curve (and 6m breakeven spread widening
cushions versus Treasuries at a couple of basis points), we believe demand from money
managers will dissipate and only re-emerge at substantially wider spread levels.
Figure 7: Callables have reduced as a % of funding Figure 8: Bellwether supply should remain brisk
% funding in callables Gross bellwether issuance, $bn

50% 180
45% 160
40% 140
35% 120
30% 100
25% 80
20% 60
15% 40
10% 20
Feb-01 Aug-02 Feb-04 Aug-05 Feb-07 Aug-08 0
2004 2005 2006 2007 2008 2009 2010 2011
FNMA FHLMC FHLB mat mat
8 January 2010 54
Figure 9: Primary distribution data shows domestic demand Figure 10: Flow of funds shows money manager demand
Agency debt outstanding owned as of:
100% 14% 14%
14% 13% 15% $ bn 2Q08 4Q08 2Q09 3Q09
19%
80% Fed 0 15 97 131
23% 19%
30% 33% 39% Banks 222 222 261 268
60% 43% 12%
19% Foreign 865 758 725 648
21% 18%
40% 15% Mutual funds 59 89 154 274
8%
55% Money mkt 425 756 733 635
20% 44%
35% 34% 33% 30% Dealers 220 130 124 109
0% Other 1313 1215 856 735
2004 2005 2006 2007 2008 2009 Total 3104 3185 2950 2800
Fund Managers Commercial Banks Of this, ST 870 1155 862 797
Central Banks Other Of this, LT 2234 2030 2088 2003
Bank portfolios: Traditionally yield focused, banks may (rightly) view agencies as cheap
Treasuries and should continue to add to holdings even at current valuations. Callables
in particular should remain in vogue, as long as MBS valuations remain ultra-rich and
the Fed stays on hold. Forthcoming prescriptive liquidity regulations should also
enhance the appeal of agency debt/MBS for bank portfolios. Importantly, banks are
awash in cash (almost $1.2trn as of YE09) from all the excess liquidity being kept in the
system by the Fed. Therefore, bank demand should remain robust at least early in the
year, which should help support spreads. However, as the Fed begins to remove excess
liquidity from the system (in H2 10), bank demand should wane.
Foreign investors: A true unknown, we believe there could be an outside chance that
the new higher PSPA limits could provoke higher demand from foreign investors. Higher
yield levels should also help, as well as any reassuring plan of action from the Obama
Administration in February.
Figure 11: Evolution of agency-Treasury spreads in 2009 Figure 12: Evolution of agency asset swap levels in 2009
Spread to Treasuries (bp) Spread to Libor (bp)

250 200
200 150
100
150
50
100 0
-50
50
-100
0
-150
3m 6m 2y 3y 5y 10y 30y
3mo 6mo 2yr 3yr 5yr 10yr 30yr
9/4/08 11/24/08 1/4/10
9/4/08 11/24/08 1/4/10
8 January 2010 55
Outlook for spreads

In late 2008, agency-Treasury spreads reached 100bp in the days leading up to
conservatorship, then proceeded to widen out even further on worries about competing
supply from the FDIC’s TLG program, before sharply turning around after the Fed
announced its purchase program for agency debt under QE. Spreads across the curve have
compressed to the 10-25bp area versus Treasuries, not far from historic tights (Figure 11).
On an asset swap basis, agencies are about 5-10bp rich to pre-conservatorship levels, with
longer tenors still in Libor plus territory, but largely as a function of how tight long maturity
swap spreads are in a historical context (Figure 12).
Spreads could widen 10-15bp Given our revised supply outlook, and our expectation that all three GSEs will have steady
from YE09 levels as some funding needs throughout the year, we believe a spread widening of at least 10-15bp across
investor classes exit the market the bullet curve is likely from YE09 levels. From a historical standpoint, in our view, equilibrium
spread levels for agency-Treasury spreads in 2010 should be closer to those seen in 2005-06,
when GSE credit was not an issue and net term issuance was in the same ballpark as current
expectations (Figure 13). To be fair, Treasury issuance today is far higher than it was back
then, which would argue for tighter fair-value spread levels. But countering this dynamic is the
fact that the buyer base has also shifted away from central banks to domestic money
mangers, which are far more sensitive to relative valuations. Without the Fed’s backstop bid,
dealers’ smaller balance sheets and reduced ability to take risk also cannot be ignored.
Figure 13: Clearing spread levels for 2010 likely similar to 2005-06 levels
Agency Net Issuance ($bn) Treasury Net Average Agency Spread to

Year Callable Non-callable Term Short-term Total Term Total 2y 5y 10y
2001 -32 208 176 69 245 -148 16 32 58 75
2002 162 82 244 -39 205 145 222 13 42 63
2003 154 24 178 89 267 330 370 8 27 46
2004 85 31 116 -28 88 310 385 12 33 48
2005 68 -12 56 -128 -72 263 224 19 29 36
2006 78 7 85 -70 15 178 158 25 33 37
2007 -119 59 -60 345 285 135 195 38 42 46
2008 -250 204 -46 273 227 398 1261 80 87 76
2009 10 -24 -14 -456 -470 1541 1845 11 21 23
2010 est 80 20 100 -100 0 1900 1250 20-25 30-40 35-45
2011 est 1400 1000
*Spread levels are as of December 31 for 2009, not an average. Source: Barclays Capital
However, we expect any spread widening to gain momentum only once it becomes clear
that the Fed is going to start draining liquidity from the system, which we peg at sometime
in Q2 2010 (to be clear, actual draining is likely a H2 2010 event, but the market will
anticipate this earlier). In the early going, we suspect robust demand from banks should
mitigate reduced money manager appetite, keeping spreads in check. Finally, if MBS widen
substantially and GSE portfolios actually grow, we could see an even greater cheapening in
spreads, as supply would exceed our projections.
Relative Value
3y agencies versus surrounding sectors: Relative to matched-maturity Treasuries, this
part of the agency curve offers the best rolldown (into 2s) and is relatively cheap to
surrounding sectors while still offering 28-30bp of pickup over Treasuries (Figure 14).
On the supply front, the GSEs should be less focused on the 3y sector in 2010 as the
8 January 2010 56
Figure 14: Agency adjusted-Treasury spreads Figure 15: Agency asset swap spreads
Adjusted Tsy Spread (bp) MM Libor Spread (bp)

40 20
35 15
30 10
25 5
0
20
-5
15
-10
10 -15
5 -20
0 -25
0 2 4 6 8 10 0 2 4 6 8 10
FNMA FHLMC FHLB FFCB C FNMA FHLMC FHLB FFCB C

asset swap curve has flattened out, and could expand to 2s and 5s (Figure 15). This
would remove some of the supply-related pressure on the 3y sector.
Overweight TLGP versus GSE names, particularly C: Despite the end of TLGP supply
and a new lease on life for GSE debt issuance, FDIC-guaranteed paper remains
stubbornly cheap to agencies. While some liquidity premium is certainly justified, note
that Citigroup names typically trade the cheapest within the TLGP space despite being
the most active issuer by far. Thus, in our view, it should not be surprising to see TLGP
valuations converge to that of agency paper.
We would recommend buying TCCULGP paper, as there may be a few more deals
before the expiration of the program on June 30.
EIB/KfW versus FNM/FRE: Although this trade has converged much of the way from its
historical wides in mid-2009 (Figure 16), EIB and KfW still trade 10-15bp cheap to
similar-maturity FNM/FRE debt. Our AAA research counterparts in Europe believe that
USD will become a less-important part of EIB/KfW funding programs in 2010, and as
gross funding should remain at or near-2009 levels, supply technicals are comparable to
Figure 16: Add to EIB/KFW over agencies in primary market Figure 17: Foreign GGB levels cheap
Asset swap spread, 5y, bp GGB ASW Sovereign CDS

100
bp 3y 3y 5y
80
60 US -16 (FRE) 33 37
40 -3 (TLGP)
20 UK 25 51 78
0 France 3 23 31
-20 Germany -6 (KfW) 19 26
-40 Netherlands 18 21 30
Fed purchases begin
-60 Sweden 20 38 52
Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Australia 21 33 39
EIB KfW FNMA FHLMC
Source: Barclays Capital As of January 6, 2010. Source: Barclays Capital
8 January 2010 57
FNM/FRE. Finally, on a sovereign-risk-adjusted basis, EIB/KfW should trade through

FNM/FRE, especially in the case of KfW as it is owned by the German government. 16 We
recommend buying these names at new issue.
Foreign GGB versus TLGP: Continuing the above theme, we recommend owning certain
foreign government-guaranteed bank paper versus TLGP and FNM/FRE names. As we
have detailed in our prior comments on sovereign risk 17, we particularly believe that
France, Germany, the Netherlands, and Australia have more stable fiscal situations as
compared to the US and UK than is reflected in their relative ratings. As such, using
sovereign CDS levels as a proxy, we believe that debt guaranteed by these governments
trades relatively cheap to their embedded sovereign risk (Figure 17).
Recent events, such as the Greece downgrade, may have helped cheapen foreign GGBs
versus US agencies further, and this dynamic could certainly continue in H1 2010.
However, over a longer-term perspective, the above governments have a much wider
fiscal moat, and we would aggressively add on any market overreaction.
Covered bonds: While we do not expect much supply, the push to develop this alternative
to securitization could result in some investor-friendly pricing for deals that come to
market in 2010. Investors hardly need to be reminded of the spectacular value offered by
TLGP deals in late 2008/early 2009 when the market was still relatively new.
Front-end callable structures, 6m locks: 3nc6m and 5nc6m European structures offer
about 20-30bp of cushion versus 6m forwards (Figure 18). At a 2.00% coupon for 3nc6m
and just north of 3% for 5nc6m Europeans, these structures make sense for yield-focused
investors who expect the Fed to stay on hold in 2010, as they offer a substantial pickup
over 6m discount notes if they are called away. They also outperform similar-OAD bullets
substantially in a modest sell-off over the next six months (Figure 19).
However, given current valuations, our expectations for increasing supply, and our view
that the Fed will withdraw liquidity in 2H10, we would tread carefully, adding exposure
opportunistically after large backups in rates, and/or sharp upticks in implied volatility.
Figure 18: Cushion of callables to the forwards is limited Figure 19: Callable excess return
Agency yield, % Excess return, bp

4.5 150
4.0
3.5 100
3.0 50
2.5
2.0 0
1.5 -50
1.0
0.5 -100
0.0 -150
0 2 4 6 8 10 -150 -100 -50 0 50 100 150
Maturity Parallel rate shift, bp; 6m horizon
Spot 6m fwd
6m fwd coupons 1y fwd coupons 3nc6m 3nc1 5nc6m
16
See Deconstructing Sovereign Ratings, July 17, 2009, and ”Are US Treasuries really risk-free?” in this Annual
17
Ibid.
8 January 2010 58
TREASURY FUTURES
A new bond contract may change dynamics

Amrut Nashikkar In early 2010, there will be a new contract at the long end of the curve. We expect the
+1 212 412 1848 contract to find favor with hedgers, to cheapen the 15y sector and to richen the 25y
amrut.nashikkar@barcap.com sector. Treasury supply driven trades should continue to offer value.
The recent trend of increased volume and improved liquidity in the Treasury futures
market should continue into 2010.
The ultra-long bond contract should find favor with investors who wish to hedge long
duration demand. The richening of 25y sector securities is likely to continue, and there
may be some widening pressure in long-end swap spreads.
The existing bond contract should cheapen as demand moves further out the curve and
a higher level of rates makes the switch option more valuable.
We expect the TY contract to trade cheap relative to cash. Trading around supply should
continue to be a theme in the TY, TU and FV contracts.
The return of futures market volume

Liquidity has returned to the The futures market had large dislocations in early 2009, because of a lack of liquidity and
futures market and should the scarcity of relative value capital. But financial sector balance sheet constraints have
continue to improve, relaxed and liquidity has returned to this market; the CME reported a surge in volumes in
accompanied by an interest rate futures in December. Figure 1 shows the total open interest in the first two
increase in volumes contracts for US, TY, FV and TU. After falling to levels that existed in 2003-04, open interest
in futures has increased. The same story can be seen in a measure of futures market
illiquidity: the monthly average ratio of absolute daily price changes to daily futures volume
during non-roll months. Large price changes accompanied by low volume mean the market
is more illiquid. The measure shows that illiquidity has declined substantially after reaching
its peak in January 2009. However, there remains scope for improvement before we get
back to pre-crisis levels. Given the steady improvement in financial balance sheets, we
Figure 1: Open interest has been rising after falling to multi- Figure 2: Illiquidity spiked in early 2009 but is improving
year lows
Millions Abs price change (cents) /log volume

3.5 12 6
3.0 10 5
2.5
8 4
2.0
6 3
1.5
1.0 4 2
0.5 2 1
0.0 0 0
Jan-00 Aug-01 Feb-03 Aug-04 Feb-06 Jul-07 Dec-08 2004 2005 2006 2007 2008 2009
US TY FV TU US TY FV, RHS TU, RHS
Source: Bloomberg, Barclays Capital Source: Barclays Capital
8 January 2010 59
expect liquidity to continue to improve through 2010. This should be accompanied by

increases in volume.
Ultra-long bond futures: A successful new contract?

The introduction of the ultra-long bond contract in January 2010 is likely to change some of
the dynamics of the curve at the long end. Bonds with maturities greater than 25 years will
be eligible for delivery under the new contract.
What will the new contract look like?

CTD securities for the ultra-long Figure 3 shows the deliverable basket for a hypothetical ultra-long bond futures contract as
contract have been switching of December 31, 2009. In computing the delivery probabilities, we have calibrated our
between Feb 36s, May 37s model to March option vols on USH0. At current levels, if the contract price were the same
and Feb 37s as our model implied price, the cheapest to deliver (CTD) would be the Feb 36 bond.
However, the CTD securities for the contract have been switching, and there are three close
contenders – Feb36s, May37s and Feb37s. The switch option value is not likely to be very
high, however, at 4.5 ticks, because of the similar durations of the securities in the basket.
The recent sell-off in rates and the accompanying increase in volatility imply that delivery
probabilities are now much more spread out across the basket than when rates were lower
at the end of November. However, given the similar durations of the securities, switch
option value has not increased too much. Relative value trading activity should likely focus
more on curve than on basis trades and could act as a potential source of shorts if the
contract consistently trades rich relative to the long end of the curve.
Significantly higher DV01 of the The significantly higher DV01 of the contract means that for every position in the current
new contract means liquidity contract that is replaced by the new ultra-long contract, there will need to be 40% fewer
may suffer contracts in the ultra-long. Thus, liquidity in the contract may suffer unless volumes from
relative value trading and the possible substitution of long-end receivers by the ultra-long
contract in the swaps market make up for the reduction.
Figure 3: Delivery analysis of hypothetical ultra-long bond contract

Model Switch option
price 120-233 (ticks) 4.5
DV01 187 CTD 4.5 Feb 36

Conv Gross Net Deliv
Coupon Maturity Yield fact. basis basis prob
4.5 2/15/2036 4.615 0.8045 33.2 0.7 39.3

4.75 2/15/2037 4.622 0.8344 35.9 1.6 0.2
5 5/15/2037 4.613 0.8671 38.1 1.6 41.1
4.375 2/15/2038 4.645 0.7816 42.4 10.8 -
4.5 5/15/2038 4.649 0.7978 40.2 7.4 9
3.5 2/15/2039 4.658 0.6594 62.6 37.4 4.2
4.25 5/15/2039 4.658 0.7609 49.9 19 0
4.5 8/15/2039 4.661 0.7943 46.2 13.7 6
4.375 11/15/2039 4.661 0.7765 50.5 18.3 0.2
Note: Closing prices for January 4, 2010. Source: Barclays Capital
8 January 2010 60
The new contract may be attractive to hedgers

We believe that there may be excess demand for long duration in off-balance-sheet
instruments at the long end of the curve. Figure 4 shows our model estimate of the richness
in the bond contract (in cents per $100 notional), together with 30y headline swap spreads
(in bp). These measures are related, and both indicate that many investors prefer to hold
long positions in the long end in off-balance-sheet instruments. Given the tightness of 30y
swap spreads, it appears that this demand has been expressed largely through 30y swaps.
This suggests that there may be interest in higher duration bond futures. If the contract is
successful, this may have a widening effect on long end swap spreads.
Implications for the long end

CTD of the current contract To analyse the possible implications of the introduction of the contract for the curve, we
should cheapen and that of the may draw analogies from the behaviour of CTDs when the notional coupon on futures
new should richen contracts was last reduced in March 2000. The new contract and the current contracts are
substitutes. The introduction of a new higher maturity bucket bond futures contract is like a
reduction in the notional coupon of an existing contract in terms of its effect on the
underlying basket.
Figure 5 shows the richness or cheapness of likely deliverables into the front bond futures
contract from February 1999 to May 2000. The first contract to trade with a 6% notional
coupon was the March 2000. The CTD for the June, Sep and Dec 99 contracts was the
February 2015 bond. Since a notional coupon of 8% strongly anchored the February 2015
bond as the CTD, it traded rich to the curve, while securities with higher maturities (for
instance, the February 2019 bond) traded cheap.
The decision to change the notional coupon was made in late February 1999 and was
implemented for the March 2000 contract. Figure 5 shows that the CTD security cheapened
significantly on announcement, since it was obvious that CTDs in the new contracts would be
further out the curve. A more interesting dynamic was observed in October 1999 to December
1999. As the March 2000 contract became increasingly active, February 15s cheapened
significantly. Securities further out the curve, such as the February 2019 bond, richened.
Figure 4: Demand for long end off-balance-sheet duration is Figure 5: In 1999’s notional coupon change, the bonds
evident in the richness of the bond contract and in 30y swap further out the curve richened
spreads
20 3 bp Roll into new

3 Notional 6% contract
coupon change
0 0
2
-20 -3 1
-40 -6 0
-1
-60 -9
-2
Last 8% coupon
-80 -12
-3 front contract
Feb-99 May-99 Sep-99 Dec-99 Apr-00 Jul-00
10s-30s spread curve (bp)
US contract richness (32nds, LHS) 11.25 2/15/2015 8.875 2/15/2019
8 January 2010 61
Applying this experience to a new 25+y bond futures contract implies that the CTDs of the
current contracts (15-16y sector) should cheapen and the February 2036 bond should
richen. Both these effects seem to have already been priced into the market somewhat,
giving us comfort in the other effects that we expect. Figure 6 shows the asset swap spread
curve in the delivery basket for USH0. Clearly, bonds in the 2025-27 sector, where the CTD
traditionally lies, are trading cheap relative to the 2036 sector, where the CTD of the new
contract is expected to be. If the new contract is a success, we would look for this process
to continue, with the 2036 sector richening further, as this may be necessary to bring shorts
into the picture to meet long demand.
US: Currently too rich

The existing long bond contract traded somewhat rich relative to our fair value model
through most of 2009, as not enough switch optionality was priced in for a given level of
rate and vol, as seen in Figure 4. We believe that this was the result of demand for off-
balance-sheet long-duration instruments. Even though ex-post rates never sold off to the
extent that would have led to realizations of CTD switches, we do not expect the same over
the next year. As discussed in our supply and demand outlook, we look for rates to sell off
significantly and the curve to steepen over the next year. Furthermore, as mentioned in our
discussion of the 10s20s30s curvature trade in the Treasury article, excess long-end supply
usually leads to a steepening of the 20-30y part of the curve and a richening of 20s relative
to 10s and 30s. Higher rates and a steeper curve in the delivery basket should imply greater
switch option value over the next year. Further, with the introduction of the ultra-long bond
contract, some of the pressures that have kept the US contract rich should be alleviated.
Long CTD basis position offers The richness of the bond contract allows trades that benefit from rising long-end rates. Even
an inexpensive way to express a after the recent sell-off, there has been little widening in the CTD basis for USH0. Our switch
conditional view on a sell-off option model predicts an option value of about 5 ticks, which is much higher than implied
by the net basis of the CTD. We believe that a long CTD basis position in USH0 offers an
inexpensive way to express a conditional view on a sell-off. The pay-off profile is shown in
Figure 7.
Most of the roll cycles in the bond contract in 2009 were characterized by a dynamic in
which the richness of the front contract translated into a richening of the back contract,
Figure 6: 2036-37 bonds trading rich relative to 2025-26 Figure 7: Long CTD basis trade in USH0 offers attractive
bonds payoff profile in case of a large sell-off in long-end rates
18 P&L (ticks) Change in slope

16 30
14 25
12 20
10 15
8
10
6
5
4
0
2
0 -5
Aug-23 Aug-25 Nov-26 Nov-27 Feb-29 Feb-31 May-37 -100 -50 0 50 100
Change in level (bp)
ASW (bp) -4 -2 0 2 4
8 January 2010 62
while the front contract normalized to fair value toward the end of the roll. This led to
cheapening of the US roll through most roll cycles. We expect this effect to become more
subdued in 2010.
TY – Basis trades expected to be active

We do not expect any systematic The TY contract tended to trade fair through most of 2009, in part because of its liquidity, as
dislocations in the TY contract, well as the fact that it lies in a liquid part of the cash curve. The basis on the contract has
given the liquidity widened significantly through the recent sell-off in rates and is at a level that we believe to
be close to fair value. Given the liquidity of the sector, we do not expect any systematic
dislocation in the contract. The sell-off in rates has made two securities close to being
cheapest to deliver in the contract – the oldest new 7y note in the basket and the oldest 10y
note in the basket. Given the level of rates that we expect, new 7y notes are likely to be
issued at higher coupons than recent 7y notes. This increases the likelihood that they will be
CTD, since a higher coupon also implies a lower duration. This may be a feature of most TY
contracts through 2010. Given our outlook on rates, some switch optionality should return
to the TY contract, and it should, in general, trade somewhat cheap to cash.
One interesting relationship in 2009 was the negative correlation in the richness of the TY
and TU contracts through the roll cycle (Figure 8). We believe that this was related to
steepener/flattener dynamics between speculators and hedgers that are apparent through
CFTC positioning data. Figure 9 shows that the relative positioning of speculators between
TY and TU indicates trend-following behavior. Given our outlook for the curve to stay steep,
we do not expect this dynamic to reverse. As a result, steepening and flattening flows
should continue to play a major role through most TY roll cycles.
TU and FV: Supply driven trades offer value

The TU contract traded The TU and FV contracts should continue to have little switch optionality. Even though we
cheap at the beginning of the expect higher rates in 2010, we believe that this should result in a steeper curve, as a result
roll cycle in 2009, which of which front and intermediate rates are unlikely to sell off significantly enough for the
could continue in 2010 switch option to come into play. Of the two, the TU contract could be more interesting, as
the trend of higher coupons in 2y notes means that newly issued 2s could again become
viable candidates for the cheapest to deliver. A distinguishing characteristic of the TU
Figure 8: Negative correlation between TU and TY rolls Figure 9: Speculator positions show trend-following
because of rolling of steepeners and flatteners behavior; currently in steepeners
5.5 TUU-Z 5.00 45 3.0

richens
5.0 30 2.5
4.5 4.75 2.0

15
4.0 TYU-Z 1.5
cheapens 4.50 0
3.5 1.0
-15
3.0 0.5
4.25
2.5 -30 0.0
2.0 4.00 -45 -0.5

10-Aug-09 20-Aug-09 30-Aug-09 9-Sep-09 Jan-05 Oct-05 Jul-06 Apr-07 Jan-08 Oct-08 Jul-09
2y - 10y (% spec positions, LHS) 2s-10s (%,RHS)

TY richness (32nds, LHS) TU richness (32nds, RHS)
8 January 2010 63
contract in 2009 was its tendency to trade cheap relative to fair value at the beginning of
the roll cycle (and for calendar rolls to trade rich). This led to the contracts’ rallying as they
neared expiration. We would not be surprised if this trend continues.
A curve flattener using TU or FY Another attractive trade in 2009 for TU and FV was trading around supply. Figures 10 and
CTDs and respective on-the-run 11 show how profitable the trade has been. It plots the slope of the yield curve between the
issues post-auctions has front contract TU CTD and the on-the-run 2y security against the number of days from
been a profitable trade which there was a 2y auction. The jump on the auction date is because of the change in the
on-the-run. For the 2y note, there does not seem to be much of a pre-auction concession
building up. However, it is clear that putting on a flattener between the CTD and the on the-
run 2y once the 2y auction is through has been an attractive trade. A similar dynamic can be
seen around 5y note auctions, when we study the slope between the CTD of the 5y contract
and the on-the-run 5y (Figure 11). Given that supply should continue to overwhelm
markets in 2010, we would expect this trade to continue to play out every month.
Figure 10: Trading around auctions in TU against the on-the- Figure 11: Trading around auctions in FV against the on-the-
run 2s has been an attractive trade run 5s has been an attractive trade
bp bp
10 29
9 28
8 27
7
26
6
25
5
4 24
3 23
2 22
1 21
0 20
-10 -8 -6 -4 -2 0 2 4 6 8 10 -10 -8 -6 -4 -2 0 2 4 6 8 10
Days from auction Days from auction
TUCTD - CT2 slope (bp) FVCTD - CT5 slope (bp)
8 January 2010 64
SWAP SPREADS
Weighed down by supply

Anshul Pradhan The swap spread curve should flatten and long-end spreads tighten through 2010.
+1 212 412 3681 Treasury supply expectations embedded in long end spreads are optimistic and related
anshul.pradhan@barcap.com sovereign risk is underpriced. Spread widening related to convexity paying or
introduction of the ULB contract should provide good entry points for these trades.
The supply/demand imbalance in short-term markets should remain in favor of excess supply
of funds, keeping the Libor-OIS basis in check. Banks should continue to reduce reliance on
short-term capital markets by switching to core deposits. Money market funds should
continue to move away from government paper. Regulation, however, remains a risk.
As the Fed begins to drain reserves through reverse repurchase agreements, the OIS-
Treasury basis should compress, providing an offset to a widening Libor-OIS basis. The
short-term Libor-Treasury spread should, therefore, remain rangebound at 20-25bp.
Treasury supply expectations embedded in longer-dated spreads are optimistically low;

budget deficits for the next several years are likely to persist at levels higher than what the
market is pricing in. Long-end spreads should drift tighter and the spread curve flatten as
the Treasury continues to term out debt.
Growing concerns about the sovereign rating of US could reduce the relative
attractiveness of long-dated Treasuries. This should put further tightening pressure on
long-dated spreads relative to front/intermediate sector spreads. The latter may be
buoyed by investors’ switching into front-end Treasuries from other governments’ debt
and by potential negative implications for ratings of large financial institutions.
Convexity paying as the Fed steps back from the mortgage market or the introduction
ultra-long contract could slow the pace of tightening early in the year. Spreads should
eventually tighten to levels consistent with Treasury supply expectations, with such
events providing attractive entry points for spread tighteners and spread curve flatteners.
We recommend 5s30s spread Swap spreads gradually tightened across the curve through 2009 from the high levels at the
curve flatteners and beta beginning of the year. The tightening in front-end spreads was led by the Libor-OIS basis as
weighted 10y spread tighteners bank funding concerns alleviated. Intermediate and long-end spreads tightened as long-
as core positions for 2010 term supply expectations of Treasury increased with worsening deficits. Convexity-related
flows added quite a bit of volatility to intermediate spreads, especially during the middle of
the year. Going into 2010, we expect front-end spreads to remain rangebound around
current levels but long-end spreads to drift tighter, with the trend accelerating in the second
half of the year as convexity paying buoys spreads in the first half (Figure 1).
Figure 1: 5s30s Spread Curve Flatteners and Beta Weighted 10y Spread Tighteners
Market Q1 10 Q2 10 Q3 10 Q4 10
2s 26 28 30 30 30
5s 29 30 30 28 25
10s 10 10 5 0 -5
30s -12 -15 -20 -28 -35
5s-30s -41 -45 -50 -56 -60
8 January 2010 65
Libor-GC: A tug of war

The path of front-end spreads in 2010 should be governed by expectations of how Libor-OIS
basis and OIS-Treasury basis are likely to evolve. The L-OIS basis for the 3m tenor has already
compressed to 10bp, below the pre-crisis average of 11bp. but it is priced to widen to 20-25bp
over the next year. Supply/demand dynamics in short-term markets should remain in favor of
excess supply of funds in 2010, keeping the L-OIS basis in check. However, regulation remains
a risk. The OIS-Treasury basis should tighten, and possibly invert, as the Fed begins to drain
reserves, given that the primary reserve draining tool would be reverse repos via Treasuries.
Hence, the short term Libor-Treasury basis should not sway from the current 20-25bp.
Supply-demand dynamics
Banks are relying less on capital Banks are gradually moving away from short-term capital markets, both for secured and
market for short-term unsecured funding, reducing the demand for short-term funds. Figure 3 shows that from
borrowings as they are September 2008 to September 2009 (latest FDIC data available), liabilities of large banks
gradually shifting to deposits have declined by $350bn, suggesting lower funding needs. In addition, core deposits
(domestic, small time) have increased by $500bn, indicating an even lower requirement to
tap the capital markets. Reverse repos/FF outstandings have shrunk 20%, FHLB advances
have declined 40%, and other short term borrowings have decreased 45% from September
2008 levels. The former two indicate less reliance on secured funding and the latter on
unsecured short-term funding. We expect this trend to continue.
Banks should continue to reduce Potential liquidity guidelines should encourage banks to own liquid assets in an unencumbered
reliance on short term form to cover liquidity needs, which are governed by the share of short-term liabilities. Since
borrowings as the share of holding liquid securities would be a drag on NIMs, banks will likely try to reduce liquidity needs.
volatile liabilities is still quite high Figure 3 shows that while banks have been moving in this direction, the share of volatile
liabilities (secured/unsecured, <1y maturity) at 35% is still high and core deposits still form only
half of total liabilities. The need to hold assets in an unencumbered form could also mean
higher issuance of unsecured debt, but we believe this should mostly be either in core deposits
or long-term debt (more so of the former), therefore ensuring that liquidity needs do not rise.
Optimizing bank balance sheets to maximize NIMs while guaranteeing that liquid assets meet
liquidity needs results in the same conclusion (see “The Impact of New Bank Liquidity
Regulations on the Fixed Income Landscape”, October 16, 2009).
Finally, the expiration of Fed facilities in February should have a limited effect on Libor, as
the banking system is not relying on them. CPFF outstanding has declined to just $14bn.
Figure 2: Banks relying more on core deposits… Figure 3: …More room to go
600 502 55%
400 50%
200
45%
0
40%
-200 -81
-137 -119
-180 -170 -173
35%
-400 -357
Other ST
FF/Repo
Deposits
Liabilities
deposits
Advances
Liabilities
LT Borr.
Other
30%
Borr.
Core
Other
Total
Sep-08 Dec-08 Mar-09 Jun-09 Sep-09

Volatile Liabilties, % of Total Liabilities
1y Change, Sep08-Sep09 Core Deposits, % of Total Liabilities
Source: FDIC Source: FDIC
8 January 2010 66
TSLF-schedule 1 was suspended in the middle of 2009, and TSLF-schedule 2 has not
received any bids since August 2009. TAF has been undersubscribed since late 2008, and
total outstanding loans under TAF have shrunk to $76bn or 15% of the peak value. Overall,
the banking system is already more or less independent of the Fed facilities.
Moving to the supplier of funds, money market funds have been shifting away from the
Expiration of Fed facilities should
safety of government paper toward CDs, and we expect that trend to continue. Figure 4
have a limited impact on Libor
shows the change in various assets held by money market funds from the peak in February
2009 to November 2009 (latest data available). Even as money market assets have shrunk
by $500bn, their investment in CDs has increased $100bn and most of the decline has been
Money market funds are
T-bills or short-term agency securities. One reason for this is that outflows from money
migrating toward bank paper
market funds have been concentrated in government-only funds, biasing the shrinkage in
from the safety of government
that sector. However, even within prime funds, the change in asset composition has been
securities, and there is more
similar with a preference toward bank paper. The share of government paper in prime funds
room to go
is still quite high, suggesting there is more room to go (Figure 5).
Outflows from money market As we move through 2010, the steepness of the curve should encourage investors to
funds are partly being channeled continue moving out of money market funds, reducing the available supply of funds in the
back into bank despots short-term market. We believe the effect will likely be small. First, the bulk of the shrinkage
should be in government funds, due to low yields and the potential for even lower yields,
given the shrinking size of the bills and agency discount universe. Second, recent data
suggest some of the money market outflows are being plugged back into bank deposits. For
instance, households have invested 30% of money market outflows during the first three
quarters of 2009 into bank deposits; this ratio compares favorably with 20% of total money
market assets in CDs. Overall, the banking sector should be well supplied with funds.
Regulation, however, remains a risk. While the aim is to reduce systemic risk, an unintended
consequence would be to increase the uncertainty about future government support to the
banking system. Bank ratings get an uplift based on the implicit government support; if this
were to be become uncertain, it could hurt ratings and increase borrowing costs. With
regulation still pending, we would look for any development on this front.
Draining of reserves should Draining reserves

cheapen Treasury collateral to
Whether draining reserves will widen Libor with respect to OIS is not quite clear, as both
swaps offsetting any widening in
rates should move higher but it should certainly cheapen GC relative to OIS. The Fed has
the L-OIS basis
mentioned reverse repos as its primary tool in draining reserves, and we believe it will likely
Figure 4: Money market funds favoring bank paper… Figure 5: …More room to go
200 18
100 16
0 14
-100
12
-200
10
-300
8
-400
-500 6
-600 4
Corproate
Tsy/Agency
CP
Repo
CD
Total
2
Dec-07 Jun-08 Dec-08 Jun-09 Dec-09
% Assets in Govt Securities

Change, Feb09 to Nov09, $bn
Source: ICI money market Source: imoney.net. Data: composition of prime money market funds
8 January 2010 67
begin with the Treasury collateral. Figure 6 shows that the excess reserves may already be
responsible for the richening of bills with respect to the OIS rate. With the agency MBS
program still underway, reserves are likely to increase first, arguing for a further richening,
though that would be limited by the 0 floor on bill yields. As we move through H2 10,
reserves should decline substantially, possibly inverting the OIS-GC basis or raising the
funding cost of Treasuries above the OIS rate.
Short term Libor-Treasury basis Overall therefore, the short-term Libor-Treasury spread should remain rangebound at 20-
should not sway away from 25bp in 2010, as any potential widening in the L-OIS basis should be gradual and offset by
current levels the tightening of the OIS-GC basis towards the second half of the year.
Treasury supply and sovereign risk

Long-end spreads should drift While front-end spreads are somewhat in line with pre-crisis levels, 10y and 30y spreads are
tighter on Treasury supply trading well through. We believe this is a reflection of high budget deficits or Treasury
expectations and related supply expectation, which is likely to persist in 2010 as well. The market seems to be priced
sovereign risk for modest supply, highlighting that the risk is toward positive supply shocks and tighter
long-dated spreads. In addition, growing concerns around sovereign rating should put a
tightening pressure on long-dated spreads, further flattening the spread curve.
Deficit projections: No relief

Expectations of persistent Figure 7 shows 10y swap spreads against the 5y forward-looking average of the CBO’s
deficits tighten long-end swap baseline deficit to GDP projections, the rationale being that a higher relative supply of
spreads as the scarcity premium Treasuries should reduce the scarcity premium in Treasuries and tighten swap spreads, and
of Treasuries declines vice versa. In 1999-2000, when the CBO was forecasting budget surpluses, 10y spreads
widened to above 100bp as the scarcity premium in Treasuries increased sharply. The
current scenario is exactly the opposite, with the CBO forecasting average deficits of -5% of
GDP for the next five years. Hence, even as measures of risk aversion have normalized,
bringing front-end spreads more in line with pre-crisis levels, supply expectations are still at
an extreme in the sample, justifying why longer spreads are trading well through. Our fair
value framework suggests 10y spreads should trade at 10-15bp for current TED spreads,
latest CBO baseline projections, and the recent move in mortgage durations. We have
established above that TED spreads are unlikely to move far from current levels, and barring
a move in mortgage durations, which we discuss in detail later, the outlook for longer
spreads should be determined by how supply expectations evolve.
Figure 6: Excess reserves exerting widening pressure Figure 7: 10y spreads and Treasury supply
1200 10 160 8
1100 140 6
5
120 4
1000
100 2
900 0
80 0
800 -5 60 -2
700 40 -4
-10
600 20 -6
0 -8
500 -15
Dec-89 Dec-94 Dec-99 Dec-04 Dec-09
Jan-09 Mar-09 May-09 Jul-09 Sep-09 Nov-09
10y Spreads, bp, LHS
Excess Reserves, $bn, RHS Fair Value, bp, LHS
3m Bill vs 3m OIS, bp, RHS Average Deficit/GDP %, 5yr, RHS
Source: Haver Analytics, Barclays Capital Source: Barclays Capital
8 January 2010 68
CBO baseline forecasts are We believe CBO baseline forecasts are understating deficit projections and the market is priced
understating deficit projections to baseline forecasts, suggesting the risk is to tighter spreads. Figure 8 tabulates the deficit to
as they are based on current GDP ratio for the next several years in the baseline CBO scenario and the baseline adjusted for
laws and policies expected policy changes. The CBO’s baseline is based on current laws and policies and does
not account for possible changes, even they are if highly likely. For instance, in the baseline
scenario, the CBO assumes that various tax cuts (individual and corporate) will expire by the
end of 2010, boosting revenues. In the CBO’s estimate of the president’s budget, published in
June 2009, it assumed that most of the tax cuts would be extended, with the exception of
those taxpayers in the high income bracket. This would, therefore, result in a higher deficit
than that assumed in the baseline. In addition, the indexation of AMT to inflation would also
reduce revenues, as it would capture a smaller share of taxpayers, resulting in higher deficits.
On the outlay side, the CBO assumes that discretionaries will grow with inflation in the
baseline, whereas they have historically moved with nominal GDP growth. CBO baseline
forecasts are, therefore, understating outlays and deficits.
Realistic deficit projections Adjusting for the above with the CBO’s own estimates for each of these categories results in
would suggest that 10y and 30y higher but more realistic primary deficits. Figure 8 shows that primary deficits should trough
spreads should be -10bp and at $650-700bn. Adding on the interest cost (computed assuming the Treasury continues to
-35bp respectively. gradually term out and rates move along the forward path), the trough in total deficits
should be close to $1.1trn, or 7% of deficits. How should this be incorporated in the model?
We have used the 5y average of the CBO’s deficit/GDP projections as one of the explanatory
variables in our swap spread model, which for 2010-14 is -5.1%. As we move through 2010,
the CBO’s baseline should move higher to the alternative scenario of -7.4%, not lower as
implied in the CBO’s baseline. Figure 9 shows the level and shape of the spread curve under
these deficit scenarios (after removing the effect of convexity, which we adjust for later).
The market is priced to the current baseline scenario, which as migrates to the alternative
scenario, should tighten 10y and 30y spreads to -10bp and -35bp, respectively.
And the spread curve should be The spread curve in this scenario is not much flatter than current levels, which, in our
flatter as front to intermediate opinion, is not the true representation. The supply-demand imbalance in Treasuries is more
spreads stay wide due to strong pronounced further out the curve. The Treasury should continue to term out debt in 2010
foreign demand by keeping front to intermediate sizes constant but issuing more at the long end. At the
same time, potential demand seems to be in the front to intermediate sector. Recent auction
statistics show good sponsorship in that sector, while long-end auctions have suffered
lately; the steepening of the rate curve last year was testament to that. Hence, 2y and 5y
spreads are likely to tighten less than indicated in Figure 9.
Figure 8: Deficits could persist at high levels

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
CBO baseline primary deficit, $bn -1,184 -681 -293 -165 -82 -16 -29 15 53 0
a. EGTRRA/JGTRRA Tax Cut Extensions 0 -102 -185 -199 -210 -221 -230 -240 -250 -260
b. Other Tax provisions -12 -67 -151 -151 -160 -165 -173 -183 -188 -194
c. Indexing the AMT to inflation -7 -69 -31 -34 -37 -41 -46 -53 -60 -70
d. Interaction of a and c 0 -18 -39 -43 -46 -49 -52 -54 -57 -60
e. Appropriations grow @ nominal GDP -8 -33 -74 -120 -159 -194 -228 -261 -294 -328
Total primary deficit, $bn -1,210 -970 -773 -711 -695 -686 -758 -777 -796 -913
Interest cost, $bn -179 -267 -364 -447 -530 -621 -703 -793 -878 -975
Total deficits, $bn -1,389 -1,237 -1,137 -1,158 -1,225 -1,306 -1,461 -1,570 -1,674 -1,887
Deficit/GDP, % -10% -8% -7% -7% -7% -7% -8% -8% -8% -9%
Deficit/GDP (CBO baseline), % -9% -6% -4% -3% -3% -3% -3% -3% -3% -3%
Source: CBO, Treasury, Barclays Capital
8 January 2010 69
Sovereign risk
Growing concerns about Another factor supporting a flatter spread curve is growing worries about sovereign risk. We
sovereign risk also support believe the US sovereign rating could become a matter of concern over the next few years if
tighter long-end spreads and a the key fiscal metrics, such as interest coverage ratio and debt/GDP, rise as projected by the
flatter spread curve CBO (see “Sovereigns: How risk free are US Treasuries?” for details). In addition, foreign central
banks have been slowly diversifying away from US dollar assets, a continuation of which
would diminish the status of the US dollar as the world reserve currency, a factor which
supports the high rating, in our opinion. In other words, while short-term Treasuries have a
strong credit rating, the same may not be said about long-term Treasuries.
Cross-country snapshot of Figure 10 plots the 5s30s spread curve across countries against their 10y sovereign CDS
spread curve suggests high spreads. As expected, the relationship is strongly negative; countries with higher sovereign
negative correlation with CDS spreads have a much flatter spread curve, which supports the theory that long-term
sovereign CDS government bonds are not seen as creditworthy as short-term ones, particularly for countries
with lower perceived ratings. The US is at the upper left corner of the chart; however, its
implied rating, when not accounting for its size, is not as strong as that for France, Germany,
or Holland but is more similar to the UK’s. Were the US sovereign CDS to move toward those
of the UK, the 5s30s spread could flatten to -60bp to -65bp from the current level of -43bp.
Negative implications for bank Such a flattening should also be supported by the effect on bank CDS spreads and flight to
credit ratings and inflows into quality flows. Credit ratings of large banks factor in an implicit government support, which if
front to intermediate Treasuries were to weaken, would have negative implications for bank ratings as well. Recent data
from other government debt show that the spread of financial to non-financial credit spreads have been correlated with
should further flatten the swap sovereign CDS spreads. Higher CDS spreads should translate into higher Libor settings; this
spread curve is supported by recent Libor submissions; banks with higher CDS spreads have submitted an
above average Libor rate. In addition, as government bonds in other countries become less
attractive amid rating concerns, there may be flight-to-quality inflows into the US in the
front to intermediate sector, further supporting a flatter spread curve. The flatness of the US
swap spread curve is, therefore, not an anomaly but a manifestation of the sovereign risk
that can be observed in other countries. The persistence of deficits and related sovereign
risk, argue for tighter long-end spreads, both outright and particularly relative to front-end.
Figure 9: Spread curve priced for lower deficits Figure 10: Sovereign CDS and swap spread curve
50 5s30s Spread Curve, bp

-35
40
-40 Holland
30
US
20 -45 Germany
10
-50
0
-10 -55
France
-20 -60 Belguim Spain
UK
-30
-65
-40 Italy
2s 5s 10s 30s -70
0 25 50 75 100 125
Market CBO-Alternative-1y fwd
10y Sovereign CDS Spread, bp
CBO-Baseline-1yfwd CBO-Baseline-Spot
8 January 2010 70
Deviating from fair value

While the above factors affect the trend level of spreads, convexity, swapped issuance and
ultra-long bond contract-related flows may result in a deviation from it. Convexity paying as
Treasury rates move higher and mortgage rates even higher (our strategists expect
mortgage spreads to normalize) should lead spreads to stay above our fair value estimate.
In addition, financial issuance should slow, putting widening pressure on spreads on the
margin. The ultra-long contract could temporarily widen 30y swap spreads. Swap spreads,
particularly 10y and 30y, could, therefore, be pushed above our fair value estimate in H1 10.
Convexity paying
Convexity paying as As mortgage rates move higher (/lower), the duration of mortgage-backed securities
mortgage rates rise should increases (/decreases), forcing active hedgers to rebalance their duration exposure. Such
push 5y and 10y spreads flows typically happen in swaps, causing swap spreads to widen (/tighten), resulting in rate
above our fair value estimate directionality between swap spreads and rates. Historical sensitivities suggest that 10y spreads
in the early part of the year widen 7-8bp for a 1-year extension in the fixed rate mortgage universe. However, the beta
could be higher if such a move were to occur over a short period. Figure 11 shows that 10y
spreads widened 35bp in the late May/early June sell-off for a 1.5-year extension in the
mortgage index. The effect, however, was temporary; the entire excess move was reversed in
the subsequent month, even though the duration of the fixed rate universe declined only
partially. Over the following months, the duration stayed rangebound but 10y spreads
tightened all the way back to the pre-sell-off levels. The same can be observed in prior
episodes of convexity paying. Figure 12 plots the excess move in 10y spreads in convexity
episodes against the move in the subsequent month adjusted for that implied by the change
in duration over that month. In all episodes, the excess move is completely reversed and then
some more. Any widening of spreads due to convexity flows should, therefore, be temporary.
Given our expectation of higher mortgage rates and the current convexity of the mortgage
universe, the duration of the fixed rate mortgage universe (ed purchases) should increase
0.5 years each in Q1 and Q2, 0.2 years in Q3, and stay flat in Q4. Figure 13 tabulates how
wide swap spreads should trade versus their respective fair values in each quarter in that
scenario. The effect should be the highest for the 10y and lowest for the 30y (in line with
the distribution of the KRDs). While spreads may initially widen by more than what Figure
13 suggests, we expect that move to reverse and would recommend using it is an
opportunity to initiate tighteners. As the mortgage duration stabilizes toward the end of
Figure 11: Convexity effect short lived Figure 12: Spreads correct in subsequent month
50 4.5 40
45 4.0 30
40 3.5 20
35 3.0
30 10
2.5
25 0
2.0
20 -10
15 1.5
1.0 -20
10
5 0.5 -30
0 0.0 Dec-01 Jul-03 Apr-04 Mar-08 Oct-08 May- Dec-09
Jan-09 Mar-09 May-09 Jul-09 Sep-09 Nov-09 Jan-10 10y Spread Widening, bp 09
Fixed Rate MBS Durn, yrs, RHS Excess Spread Widening, bp
10y Swap Spreads, bp, LHS Move over the subsequent month
8 January 2010 71
2010, the convexity effect should fade, bringing spreads more in line with fair values
discussed in Figure 9.
Financial issuance
Slowing financial issuance Fixed financial issuance is typically swapped back into floating, putting temporary tightening
should have a marginal pressure on swap spreads. Financial issuance increased in 2009, led by government-
widening effect on spreads guaranteed debt, which formed the bulk of issuance in the first half of the year. Figure 14
shows that it accounted for almost all of issuance in the first three months of 2009. As the
year progressed, banks moved away from it as a show of strength of being able to pay back
TARP money. In the past few months, it has been nil. The key difference from the perspective
of swap spreads is which sector is affected; while guaranteed issuance has been
concentrated in the front end, non-guaranteed issuance is usually concentrated in the 5y and
10y part of the curve.
There will likely be little in terms of guaranteed issuance, and non-guaranteed issuance
should also decline. Our credit strategists expect financial issuance to fall to $190bn in 2010
from $312bn in 2009 (and from $207bn, excluding guaranteed issuance). The effect on
spreads should, however, be limited. Guaranteed issuance has already slowed to 0 in the
past few months, so there should not be any effect going into 2010. Similarly, the decline in
the monthly run-rate of non-guaranteed should be marginal: from $20-25bn in the past few
months to $16bn in 2010. We estimate that a $10bn deviation from the recent trend is
worth only 2bp in 10y spreads. However, seasonals in financial issuance may result in
monthly fluctuations. Figure 14 shows that financial issuance has historically been high in
the first half of the year, particularly in January and May.
Ultra-long bond contract

Introduction of the ultra-long The introduction of the ultra-long contract next week may result in temporary widening of
contract can temporarily widen 30y spreads as long-end investors switch from receiving in swaps to the contract or other
30y spreads investors position for such a switch. However, as discussed above, the negative level of 30y
spreads is not due to technical factors but fundamental ones such as persistent deficits and
sovereign risk. Our fair value model indicates that 30y spreads should trade at -10bp for
current CBO baseline projections, compared with the current -14p. Hence, a 5bp initial
widening cannot be discounted, but as the year progresses and the CBO’s deficit projections
get more realistic, we would expect 30y spreads to tighten back.
Figure 13: Convexity flows adjustments to fair value, bp Figure 14: Financial issuance less of a factor
7 60 4%
6 3%
50
5 2%
40
1%
4
30 0%
3
-1%
20
2 -2%
10
1 -3%
0 0 -4%
q409 q1-10 q2-10 q3-10 q4-10 1 2 3 4 5 6 7 8 9 10 11 12
Non-Gtd Gtd Issuance
Convexity Impact on 10y on 5y
on 2y on 30y 2004-08 Seasonals, RHS Average 2010 Issuance
8 January 2010 72
Trade recommendations
Macro
Beta-weighted 10y spread tighteners
We recommend beta-weighted 10y spreads should tighten over the course of the year as the market re-prices long-term
10y spread tighteners as a core Treasury supply expectations with a non-trivial possibility of negative 10y spreads (Figure
position for 2010 but wait for a 15). However, we are wary of initiating the trade right away. A sharp pickup in financial
better entry point issuance is partly responsible for the recent tightening in 10y spreads, which could reverse
over the coming weeks. Spreads have been directional with rates, and we recommend beta-
weighting (using a 15% beta) the trade to reduce mark-to-market volatility. We would add
aggressively on any violent reaction to convexity paying.
5s30s spread curve flattener

The spread curve should flatten The spread curve should be flatter for similar reasons: Treasury supply expectations
through 2010; we would use any embedded in longer-dated spreads are optimistic, and related sovereign risk is understated.
ULB-related widening of long- A revision to both should lead to tighter long-end spreads. We expect 30y spreads to
end spreads to initiate 5s-30s gradually tighten to -35bp. The latter should also boost front-end spreads and flatten the
spread curve flatteners spread curve further through flight-to-quality flows from other governments’ debt into front
to intermediate US Treasuries and the negative effect on bank credit ratings. Convexity-
paying related flows also support a flatter spread curve, as they should be concentrated in
the 5-10y part of the curve. We expect the 5s30s spread curve to flatten from -43bp to -
60bp over the course of the year. We would wait for any potential bounce in 30y spreads in
the coming weeks helped by the ultra-long contract before initiating the trade.
Micro
2s-3s spread curve flattener
3y spreads still look wide 3y spreads continue to look wide relative to 2y and 5y spreads, due to the richness of 3y
to 2y and 5y spreads Treasuries. The 3y OIS-Tsy basis is trading at 13bp, compared with 7bp in the 2y and 5y
sectors. Recent auction statistics highlight that demand for Treasuries is smoothing out in
the front to intermediate sector. We prefer a 2s3s spread curve flattener over an outright 3y
spread tightener, as the former should also act as a hedge for a risk flare. Investors could
also consider a 3s5s spread curve steepener to position for a convexity-paying event.
Figure 15: Spread curve should flatten, led by the tightening of long-end swap spreads
Swap Spreads Market Q1 10 Q2 10 Q3 10 Q4 10
2s 26 28 30 30 30
5s 29 30 30 28 25
10s 10 10 5 0 -5
30s -12 -15 -20 -28 -35
5s-30s -41 -45 -50 -56 -60
Note: Data: matched maturity. Source: Barclays Capital
8 January 2010 73
VOLATILITY
Vol: Entering a quiet period

Piyush Goyal The first half of 2010 should be characterized by a Fed on hold, an economic recovery
+1 212 412 6793 that puts paid to fears of a double dip, and increased volatility supply driven by higher
piyush.goyal@barcap.com yields. In this environment, gamma should drop across tails. However, longer expiries
could stay well bid, given demand from the MBS universe and very little supply.
Across the volatility surface, the supply of optionality should more than match demand
in 2010. The primary drivers of supply should be structured and callable note issuance,
and selling programs from fast money and real money accounts. The primary demand
drivers will probably be MSR hedgers, MBS portfolios, and annuity hedgers.
Supply should be concentrated in short and intermediate expiries, while demand will
likely be spread out more evenly, including in longer expiries.
− As a result, short and intermediate options should cheapen out across tails in 2010.
But we caution against selling longer expiries, even at what may seem historically
rich levels.
We recommend the following trades, some immediately and some at specific price levels:
− In the immediate future, we like selling 2y*10y straddles. For investors worried
about a rate back-up after the Fed finishes QE, we like buying 3m*10y payer
spreads to hedge this event risk. We like selling 3m*10y straddles, but only when
implied vol touches 8.5 bp/day.
− The payer skew should cheapen as annuity hedgers bid low-strike options and
realized inflation stays low, as we expect. Buy 5y*5y ATM payer versus sell 200bp
out of money 5y*5y payer. Or buy 2x7 digital cap versus 2x7 linear cap.
− Finally, we recommend buying 1x3 caps funded with 1y2y payer swaption. The
uncertainty about short rates should pick up as the economy recovers and as Fed
rhetoric starts to point to monetary tightening.
Across the volatility surface: Supply…

There is enough supply to meet The overall supply-demand picture for options during 2010 is more balanced than it was
demand during 2009 (Figure 1). In this section, we go through these dynamics for short-dated and
intermediate expiries. It is far more difficult to quantify the supply-demand drivers for very
long-dated expiries, so we tackle that issue separately.
The major sources of supply should be structured note issuance (mainly fixed rate step-ups
and zero coupon callables), and callable issuance by the GSEs. Option selling programs by
hedge funds and real money investors could also play a role, as in 2009, but those numbers
are more difficult to estimate. Key demand drivers for short and intermediate expiries will likely
be MSR hedgers and other MBS portfolios. Consider the supply factors first:
8 January 2010 74
Figure 1: Supply-demand dynamics in volatility in 2010 Figure 2: Increased issuance in fixed rate step-ups
Net Issuance ($bn)

MBS
Fixed-rate 12
Portfolio
Step Ups
Managers
10
Agency 8
OPTION Mortgage
Callable
DEALER Servicers
Notes 6
4
Annuity
Zero Hedgers,
2
Coupon Other
Callables Portfolio
Managers 0
Q1 08 Q2 08 Q3 08 Q4 08 Q1 09 Q2 09 Q3 09 Q4 09
Source: Barclays Capital Note: As of December 21, 2009. Source: mtn-i
A) Structured note issuance: From the standpoint of volatility supply, there were two types
of notes issued in size in 2009 – 1) fixed rate step-ups, and 2) zero coupon callables. There
was around $27.5bn in fixed rate step-up note issuance (net) in 2009. This was nearly six
times the gross issuance in 2008, Figure 2. A typical fixed rate step-up note from 2009 has a
5-year maturity, and is callable every quarter after the first six months. So, this issuance led
to supply of short-dated options on short tails.
Fixed-rate step up issuance Given this surge in issuance, our estimate is that the options market was able to source
should provide $1.5-$2 mm around $15mn log vega in 2009 from this segment alone. For 2010, the vega supply from
vega/month this source should be even larger, for two reasons. First, investor appetite for fixed-rate step-
ups seems to be on the rise, Figure 2. Also, issuers are likely to enthusiastically lock in longer
funding in 2010 before rates rise much. This should generate $1.5-2mn log vega per month,
in our opinion.
B) Zero coupon callables: 2009 also had $13.8bn in zero coupon callable issuance (net),
which is twice the gross issuance from last year (Figure 3). Most of these securities mature
in 30 years, and are not callable for the first two years. Therefore, issuance typically leads to
a supply of 30y NC 18m Bermudan options. Hence, the $13.8bn issued in 2009 probably led
to supply of $25-35mn log vega.
Zero-Cpn callable issuance If the experience of 2009 is repeated, as seems likely, we are looking at new supply of $2.5-
should provide $2.5-3 mm 3mn log vega per month in 2010. In addition, if rates rise in 2010 as per our forecasts, the
vega/month existing notes should extend, causing additional supply of vega.
8 January 2010 75
Figure 3: Robust issuance in zero-coupon callables Figure 4: Net agency callable issuance has been positive
Net Issuance ($bn) Net issuance ($bn) 5yr swap rate

30 5.0
5
avg =
15 4.5
4 3.5bn/qtr
0 4.0
3
avg = -15 3.5
2 1.7bn/qtr
-30 3.0
1 -45 2.5
0 -60 2.0
Q1 08 Q2 08 Q3 08 Q4 08 Q1 09 Q2 09 Q3 09 Q4 09 Jan-08 May-08 Sep-08 Jan-09 May-09 Sep-09
Net Issuance (LHS) 5yr swap rate (RHS)
Source: mtn-i Source: Barclays Capital
C) Agency callable notes: Gross issuance was $40-50bn every month in 2009, but the
amount getting called decreased considerably as rates stopped rallying to the extent seen in
2008 (see Figure 4). With callable funding levels currently rich to bullets’ on an LOAS basis,
we expect Fannie Mae/Freddie Mac, in particular, to continue emphasizing callables within
their funding mix. We look for callable supply to be concentrated in 1H10, as the Fed should
be inactive, keeping the curve steep and stimulating demand for discount note alternatives.
In addition, as rates rise across 2010, existing bonds should extend, also pushing up net
issuance. We do not go through all the numbers here, but our analysis points to callable
issuance of $5-8bn a month in 2010, which implies supply of $2-3mn log vega per month.
Agency Callable issuance should When these three sources of volatility are added up, we end up with $6-8mn log vega per
provide $2-3 mm vega/month month that needs to be absorbed by fixed income investors in 2010. In addition, hedge
funds and other real money investors will probably continue to sell short-dated options
while rates remain low and spreads tight. Many investors do this as an alternative to selling
vol in MBS because they do not want to be long MBS at current tight spreads. However, this
source of supply might be temporary and dissipate as spreads widen once the Fed stops
buying MBS after March 2010.
…and demand
The MBS universe always plays a big role in the demand for volatility, and 2010 is unlikely to
be an exception. The three big mortgage companies tend to be MSR hedgers, leveraged
MBS portfolios, and mortgage originators. Originators are unlikely to be a significant bid for
volatility in 2010, with rates rising. Demand should, thus, be concentrated in MSR hedgers
and MBS portfolios.
A) MSR hedgers: The mortgage servicing asset (MSR portfolio) is significantly negatively
convex (Figure 5). By our estimate, if all the gamma risk in the MSR is hedged using
3m*10y straddles, this would lead to a demand for $100mn log vega. But much of the
negative convexity is not hedged using options; MSR portfolios tend to be very active
delta hedgers in the agency MBS markets. In addition, we believe that model convexity
risk might be overstated. Figure 6 looks at the refinancing behavior of agency MBS in
different periods, and with the same level of in-the-moneyness. Consider the 100bp
point. Agency MBS that were 100bp in the money prepaid at around 55 CPR (55%)
8 January 2010 76
Figure 5: MSRs have significant model convexity risk Figure 6: MBS seem to have lower interest rate sensitivity
10yr equivalents ($bn) Prepay speed

250 60
200 50
150 40
100
30
50
20
0
10
-50
0
-100
0 25 50 75 100
-400 -200 0 200 400
Interest rate incentive to refi (bp)
Change in rate (bp)
Oct-09 Mar 09- May 09 2003
Overall MSR Index
Note: As of December 21, 2009. Source: Barclays Capital Source: Barclays Capital
across the year 2003. Similar loans, with the same level of in-the-moneyness, are
prepaying at only 20 CPR (20%) right now.
Through 2010, servicers should The reason for this is that underwriting standards have tightened sharply in agency MBS
buy ~ $30mm vega; portfolio across the past year. So even if rates are low, the ability of all but the most pristine agency
managers should buy ~ $40mm MBS borrowers to access those rates is limited. Consequently, the MSR has less optionality
vega (and, thus, less negative convexity) than models may suggest. We assume that hedgers will
choose to hedge a generous 30% of the asset convexity using options, after factoring in
delta hedging and the lower need to hedge due to tighter credit conditions. This leads to
demand of $30mn log vega from the MSR community.
B) MBS portfolios: Mortgage servicers are not the only hedgers from the MBS universe.
While many MBS portfolios do not hedge convexity risk (such as”real money”
investors), leveraged portfolios such as the GSEs do hedge a portion of their risk.
Figure 7 shows the convexity profile of the MBS index. But we use the index net of the
Fed’s MBS portfolio. This is important since the Fed is a non-hedger, owns MBS that are
very likely to extend, and owns $1.25trn of them. Thus, the Fed portfolio serves as a
dampener on the option bid from MBS portfolios.
Our models estimate total convexity exposure equal to $700mn log vega. But only 20%
of that is held by the hedgers (leveraged investors). And even these investors hedge
only around 30% of their risk. Hence, we expect MBS portfolios to end up being a big
support for this sector of about $42mm logs vega in 2010.
8 January 2010 77
Figure 7: Convexity profile of the non-Fed MBS index Figure 8: Trust preferred issuance has dropped recently
10yr equivalents ($bn) $bn

1,000 60 1,500
500 50 1,400
0 40 1,300
-500 30 1,200
-1,000 20 1,100
-1,500 10 1,000
-2,000
0 900
-400 -200 0 200 400
1999 2001 2003 2005 2007 2009
Change in rate (bp)
Overall MBS Index Trust Preferred Issuance ($bn, LHS) SPX (RHS)
Note: As of December 21, 2009. Source: Barclays Capital Source: SNL, Barclays Capital
In summary, the demand for short and intermediate expiries should be around $70-80mn
log vega over the year. This is on the lower end of the range we expect for supply, which
should be $75-100mn over the year. But our supply projections do not include the selling
programs from investors, especially in the first quarter while MBS remain tight. This extra
supply, coupled with a reduction in fears about the economy, should drive shorter expiries
lower across tails at the start of 2010.
Vega: Demand should outstrip supply

The picture seems quite different on the long-dated front. One big reason for this is the
absence of significant trust preferred issuance, which has been a big supply of long-dated
expiries in past years. Figure 8 shows trust-preferred issuance over the past 10 years.
Issuance was almost non-existent during 2009, a by-product of the credit crisis. This lack of
issuance should continue in 2010, for two reasons. First, banks have limited capacity to
issue trust-preferred capital. These securities count as Tier 1 capital only as long as the
restricted core capital (which includes trust-preferred) is less than 25% of the total Tier 1
capital. Currently, banks have approximately $30bn in excess capacity to issue trust-
preferred next year. After sieving through individual financial institutions, we believe $7bn is
a realistic upper limit of how much may be issued in 2010 as banks focus on their tangible
common equity before they attempt to improve their Tier 1 ratios.
There will be little issuance in Second, trust-preferred capital is attractive only if banks have a high return on assets and
Trust Preferred securities are allowed significant leverage. With low rates and tight credit spreads, banks should be
challenged to find high yielding instruments. This, along with lower leverage, should keep
issuers away from the trust-preferred market, at least in 2010. And finally, regulators and
investors have started to emphasize tangible common equity capital. As a result, financial
institutions are unlikely to issue much in the form of trust preferred capital, capping the
supply of very long-dated volatility.
Meanwhile, there are two separate sources of demand for vega – annuity hedgers and
highly leveraged MBS portfolios. Consider the annuity hedgers first. Variable annuity (VA)
sales have bottomed out recently after falling steadily across 2008 and early 2009
(Figure 9). Moreover, many such investors unwound large portions of their portfolios earlier
in 2009. For instance, Figure 10 shows the amount of floors outstanding for two VA
8 January 2010 78
Figure 9: Variable annuity sales seem to have bottomed Figure 10: VA hedgers unwound a large amount of floors
$bn Floors outstanding (notional, $bn)

50 1,550 80
70
45 1,400
60
40 1,250 50
40
35 1,100 30
20
30 950
10
25 800 -
Q1 05 Q1 06 Q1 07 Q1 08 Q1 09 Q4 05 Q2 06 Q4 06 Q2 07 Q4 07 Q2 08 Q4 08 Q2 09
VA Sales ($bn, LHS) S&P (RHS)

Metlife AXA
Source: LIMRA, Bloomberg Source: Barclays Capital Inc.
hedgers who have traditionally maintained large option portfolios to manage their annuity
rate/vol risks. Finally, as Figure 11 shows, VA issuance should continue rising in 2010, as
per the Life Insurance Marketing and Research Association’s (LIMRA) forecast.
There should be notable demand The other source of vega demand should come from very highly leveraged portfolios, and
from annuity hedgers and MBS especially the GSEs. Most MBS portfolios prefer to hedge convexity risk. The vega exposure
vega hedgers of a MBS is 20-25bp for every $100, which seems insignificant. But when it is leveraged up
25-30 times, it becomes a much bigger risk, which is why the GSEs hedge vega. Figure 12
shows the vega sensitivity of different coupons in the agency MBS index today. Currently,
the average dollar price of the MBS universe is around $105, very similar to the price of the
5.5% coupon. So, it is reasonable to assume that the vega of the agency MBS universe is
roughly the same as that of the 5.5s, or around $0.08. If rates rise 100bp, the average dollar
price should drop to that of the 4.5% coupon (since 4.5s are 100bp apart from 5.5s).
Admittedly, model vegas are tricky, and generate several questions. Does one shift log vol
when rates rise or keep it constant? How about keeping bp vol constant and then re-
calibrating to a new lognormal volatility? Is clean vega or “dirty” vega (which includes the
impact of a volatility change on the mortgage rate) a better measure? All of these complex
questions are why we hesitate to quantify the demand for vega from leveraged MBS
portfolios. But the direction is clear enough. If rates rise, the demand for vega from
leveraged hedgers should go up significantly.
8 January 2010 79
Figure 11: Variable annuity sales are forecasted to rise Figure 12: MBS vega exposure should rise with rates
$bn Vega (cts)

220 207 0
198
200 187 -5
175
180
-10
156
160
135 -15
140
120 -20
100 -25
2008 2009 2010 2011 2012 2013 4.0 4.5 5.0 5.5 6.0 6.5
Cpn
LIMRA Forecast of VA Sales ($bn)
Source: LIMRA Source: Barclays Capital
On the face of it, longer expiries do look historically rich, given the level of gamma. But with
no supply to speak of, and demand from annuity hedgers and MBS portfolios, long-dated
expiries should stay well-supported in 2010. That is why we caution against shorting this
sector despite our overall bearishness on volatility.
Trades for 2010

Any view on volatility is only useful if it can then be executed efficiently in the options
market. We like the following trades for 2010.
Sell 2y*10y straddles
Sell 3m*10y un-delta-hedged when implied vol is at-least 8.5bp/year
Sell payer skew: Buy 5y*5y ATM versus 200bp high-strike payer, buy digital caps versus
sell linear caps
Expect rise in short rates: Buy 1x3 caps versus 1y*2y payer
A) Sell intermediate expiry options

The overall supply-demand picture for options during 2010 is a lot more balanced than it
was during 2009. So, any spike in vols should prove temporary. For now, intermediate
expiry options are still expensive even as short-dated options have cheapened. Figure 13
shows the implied vol ratio of 3m*10y and 2y*10y. This vol ratio is at the lowest levels of
the past eight years. Given the supply-demand backdrop and current levels, we like selling
2yr*10y straddles.
8 January 2010 80
Figure 13: Short-dated options have cheapened notably relative to longer-dated options
1.65
1.55
1.45
1.35
1.25
1.15
1.05
0.95
0.85
0.75
Jun-02 Jun-03 Jun-04 Jun-05 Jun-06 Jun-07 Jun-08 Jun-09
3m*10yr to 2yr*10yr vol ratio

Note: Data Period: June 2002 – December 2009. Source: Barclays Capital
B) The gamma strategy: Sell 3m*10y @ 8.5-9bp/day

3m*10y should trade within 6-10bp/day for 2010, in our view. We estimate the range using
historical experience. Figure 14 shows a 15-year history of 60d realized vol of 10y swaps. The
lower bound of 6bp/day is a reflection of the average vol prior to the period of great
moderation (June 2004 – June 2007). There are only two periods when realized vol exceeded
10bp/day. The first was after the Lehman Brother’s bankruptcy in Q4 08. The other was the
convexity episode during June 2009, which was more violent than usual as most hedgers were
caught off-guard, given their anticipation the Fed would not let longer rates rise too much.
Figure 14: 3m*10y swaps should price 6-10bp/day during 2010
16
Convexity episode
14
Escalation
12
10
Avg ~ 6bp/day Onset of crisis
8
6
4
2
0
Jan-95 Jan-97 Jan-99 Jan-01 Jan-03 Jan-05 Jan-07 Jan-09
Both events are unlikely, in our opinion, so 10y swaps are unlikely to realize more than
10bp/day in volatility over a 3m period.
Given the supply of options and the possibility of smaller demand from mortgage hedgers,
the upper bound may be much less than 10bp/day. Accordingly, we recommend selling
straddles when priced vol expectations are 8.5-9bp/day.
Also, a un-delta hedged portfolio should fare better, in our view, because with the Fed
guarding against deflation and a large duration supply mismatch, the 10y rates would
8 January 2010 81
unlikely decline below 3% during 2010. Furthermore, without the Fed losing inflation
expectations (not our base case), the 10y will stay below 4.5% for most of the next several
months. Thus, a portfolio of short 3m*10y option straddles struck within 3.5-4%, with
option premium enough to cover a 50bp move each way, should do well, in our view. Note
that 8.5-9bp/day priced vol corresponds to a 54-57bp move in the rate to breakeven on a
long straddle.
C) Short tails versus long tails

Experience suggests that options on short rates suffer the most during Fed-on-hold periods.
However, the Fed has been on hold for more than a year now. The real driver of tails could,
thus, be how the Fed chooses to drain liquidity from the system. Will reverse repos be the
only instrument of choice? Would they suffice to drain liquidity? How will the market react if
the Fed approaches Congress to get the authority to pay interest on reserves to the two
GSEs? There are several uncertainties which should cause larger swings in short rates than
in 2009.
At the least, short tails should richen relative to longer tails. However, we believe a long-
short relative value trade is not the best way to express this view. We suggest positioning
against the rise in uncertainty of short rates by buying 1x3 caps funded with 1y2y payer
swaption. This is a limited loss trade that benefits if 2y rates rise 50-100bp over the next six
months. The primary reason for this trade is that an investor does not have to be correct
about the timing of any rate move. Also, this is one of the few strategies that do not carry
negatively, the way most of the others (such as put spreads) do.
D) Monetize the payer skew

The payer skew, defined as the implied vol spread between high- and low-strike options, is
currently very large. Fundamentally, a high payer skew implies a large move lower in rates is
less likely than a large move higher. This could hold if rates are very low, say 1%. However,
the US rates are much higher. Yet, the skew exists. The reason is the one-way flow option
dealers that have been subject to in 2009. Hedgers and speculators alike had a bid for high-
strike options. So, option desks were net short high-strike options. Thus, vols rallied as rates
rose, and vice-versa for most of 2009.
We think the skew will cheapen during 2010 for reasons already discussed. One, as rates
rise, the existing option supply from zero coupon callable will keep vol from rising too much.
Two, annuity hedgers will have a bid for low-strike options after longer rates have risen.
And three, without the Fed losing inflation fighting credibility, nominal rates may not push
higher fast enough to increase the appetite of the speculative investor in the out-of-money
payer swaption. We suggest a long position in 5y*5y ATM payer swaption versus a short
position in 200bp high-strike 5y*5y swaption, with a delta hedge.
At the very least, investors should monetize the payer skew via 1y and 2y expiry options: sell
1y*10y 100bp high-strike payer versus 100bp low-strike receiver, with a delta-hedge.
However, we prefer the following trade, as it combines our rate and vol views:
Buy $400mn 2y x 7y digital cap at 7%
Sell $100mn 2y x 7y linear cap at 7%
Premium neutral
8 January 2010 82
The digital cap is currently very cheap compared with the linear cap, due to the payer skew
and supply of digital options from structured note issuance. Figure 15 shows the price of
the digital and linear cap over the past several years. Due to such valuations, one can buy
four digital caps for the cost of one linear cap.
We feel positive about this trade recommendation as it not only exploits the payer skew, but
also allows us to benefit from a rise in 5y rates over the next few quarters. As Figure 16
shows, the trade should do well unless 5y rates rise more than 700bp in two years. That
seems very unlikely, especially with the Fed staying on hold for at least the first half of 2010.
Accordingly, the risk-reward seems very attractive to us.
Figure 15: Digital caps are cheap to linear caps Figure 16: The long digital vs. short linear cap will likely do
well
Premium (cts) p&l (cts)

250 400
Curr 5yr = 2.65
300
200
200
150 100
0
100
-100
50 -200
-300
0
-400
Feb-05 Nov-05 Aug-06 May-07 Feb-08 Nov-08 Aug-09
-3 0 3 6 9 12
2x7 Linear @ 7% 2x7 Digital @ 7% 5yr swap rate, 2yr later
8 January 2010 83
SOVEREIGNS
How risk-free are US Treasuries?

Rajiv Setia The rise in sovereign risk in other developed economies is a near-term positive for US
+1 212 412 5507 rates. However, over the medium term, if the CBO’s deficit projections are realized in
rajiv.setia@barcap.com conjunction with even an innocuous rise in rates, the US AAA rating could suffer.
Against this backdrop, continued diversification away from USD by global central banks
Anshul Pradhan could precipitate negative ratings action, though not in 2010, in our view.
+1 212 412 3681
Last July 18, we introduced a simple “implied ratings” framework for assessing sovereign
anshul.pradhan@barcap.com
credit. At the time, our goal was to build a consistent framework for discerning relative value
in the various government-guaranteed programs that were in full bloom. We also wanted to
Amrut Nashikkar
better understand why, for instance, the UK was being subjected to several negative rating
+1 212 412 1848
agency comments, while the US (which, in many respects, has a similar financial profile)
remained AAA.
Our key finding was that sovereign ratings, very similar to corporates, place a significant
weight on size – in this case, size of the economy. In particular, we found that the US AAA
status was very closely tied to the dollar’s status as a reserve currency, which insulated its
rating from any marked deterioration in financial metrics. However, if the proportion of
world reserves held in dollars fell below 55%, standalone risk indicators such as debt/GDP,
became increasingly relevant. We concluded that the US remained AAA near term, but there
were plenty of reasons to worry in the medium term, given growing speculation about
central bank diversification away from USD and expectations for sustained deficits over the
next decade.
Sovereign credit concerns have With worries about sovereign credit clearly back in the forefront (Figure 1), we now build on
re-emerged over the past few our prior work and examine the potential implications of another flare in sovereign risk on
weeks the US rates market. We also supplement our prior analysis with a bottoms-up look at the
medium-term threats to the US AAA sovereign rating, and the implications for US rates.
Our key findings:
The sovereign CDS market leads negative ratings news. For developed economies, we
estimate that CDS widen by an average of 25bp per rating notch. In this context, the
widening in UK CDS suggests that the market has priced in a one-notch downgrade. In
contrast, the widening in US CDS has been marginal, suggesting market expectations of
a stable rating (Figure 2).
In the near term, negative news in developed sovereign credits is a positive for
Treasuries, as it leads to risk aversion flows into the US. By our estimates, a two-notch
downgrade in other developed economies could lead to a temporary rally of around
25bp in 10y rates.
While prospects for a near-term US ratings downgrade are remote, the AAA rating could
come under threat over a three- to four-year horizon.
Two key metrics watched by the ratings agencies are a) the debt/GDP ratio, and b) the
ratio of interest costs to government revenues. Making realistic adjustments to the CBO’s
baseline deficit projections results in a debt/GDP ratio that rises steadily to 100% over the
18
“Safe for Now: Deconstructing Sovereign Ratings” in Market Strategy Americas, July 16, 2009
8 January 2010 84
next decade. Given this backdrop, even a relatively innocuous rise in interest rates could
lead both metrics to breach the comfort zone necessary to maintain its AAA rating.
Even through the financial crisis, foreign central banks were diversifying away from USD
assets, although the extent of this was somewhat masked by dollar strength. Continued
diversification, coupled with a deteriorating fiscal profile, may precipitate negative
ratings action.
The oft-expressed view that the US can simply inflate away its debt is not credible. Not
only do government revenues and nominal GDP rise at the rate of inflation, so do
government outlays. On the contrary, the rise in nominal rates, in an environment in
which the Fed loses its credibility, could actually worsen the fiscal metrics outlined
earlier, accelerating the potential for a ratings downgrade.
Market Implications
The AAA near-term rating, coupled with the possibility of negative action in the medium
term, argues for the US yield curve to stay steeper than forwards. In our view, investors
currently do not factor in any sovereign risk premium when thinking about longer
maturity US Treasury bonds, but one may become increasingly warranted.
A sovereign downgrade could also filter through and hurt credit ratings for large leveraged
financial institutions whose ratings benefit from the presumption of sovereign support.
Upward pressure on interbank rates, together with an increased perception of risk in
longer-term government bonds, could lead to a further inversion of the swap spread curve.
Debt backed by solidly AAA-rated sovereigns, such as France (SFEFR) and Germany
(KfW), offer value. Recent events, such as the Greece downgrade, may have helped
cheapen foreign government-guaranteed bonds, and this dynamic could continue in
H1 10. We would aggressively add to select names on any market overreaction.
Sovereign CDS – A useful leading indicator

In recent months, many developed economies such as the UK, Spain, Italy, Portugal, and Greece
have been subject to negative ratings-related commentary. While many may cast doubt on the
depth and veracity of the sovereign CDS market, it appears that it is a useful signaling
Figure 1: Widening CDS spreads indicate increased concerns Figure 2: US CDS spreads have widened marginally
around sovereign credit risk compared with UK CDS spreads
5y CDS (bp) 50
90
45
80
70 40
60 35
50
40 30
30 25
20
20
10
0 15
Aug-09 Sep-09 Oct-09 Nov-09 Dec-09 Sep-09 Sep-09 Oct-09 Nov-09 Dec-09 Jan-10
UK US Germany France Japan UK-US 5y CDS (bp)
Source: Bloomberg, Barclays Capital Source: Bloomberg, Barclays Capital
8 January 2010 85
mechanism, as a sustained widening in sovereign spreads often precedes ratings actions

(defined as downgrades, changes in the outlook, or negative statements by ratings agencies).
Figure 4 shows an event study centered on negative announcements made by any one of
the three major ratings agencies over the past year regarding specific sovereigns. In
constructing this figure, we plot the change in the 5y sovereign CDS spread for a country
around the ratings announcement, after removing common variation across countries.
Sovereign CDS spreads are a As background, many studies 19 have shown that there is a common “level” factor across
leading indicator of negative sovereign CDS spreads that reflects global risk appetite. Because we are interested in
ratings news analyzing only country-specific changes in the sovereign CDS, we need to control for this
factor. We remove this common factor from the CDS spread for each country. What is left is
the country-specific component. We plot this component around dates on which a
significant ratings announcement was made by one of the three major ratings agencies.
Figure 4 shows that the sovereign CDS market prices negative ratings news.
By our estimates, a one-notch decrease in the rating (ie, from AAA to AA+) corresponds to a
25bp widening in sovereign credit spreads. In this context, the 30bp widening in UK CDS
spreads over the recent past bears watching. Our sensitivity analysis suggests that the CDS
market has already priced in a one-notch downgrade for the UK. In contrast, the widening in
US CDS over the same period has been marginal (8bp), which we view as appropriate, given
the remote potential for a near-term US downgrade. To us, the move in UK CDS is not
altogether surprising, in view of the results of our earlier analysis, which suggested that the
UK’s implied rating was AA+ 20.
For rates investors, the key takeaway is to monitor CDS levels of G8 economies for sustained
country-specific spread widening, as negative rating actions in developed economies could
trigger flight-to-quality flows into the US, a topic we delve into next.
Near term, increased G8 sovereign risk is a positive for US rates

Because of its unique status as a global safe haven, concerns about the sovereign credit risk
of other economies are a positive for US rates because they lead to risk aversion flows into
the US. A plot of the US 10y yield against the average 5y CDS spread in other developed
Figure 3: Within AAA countries, the sovereign CDS spread is Figure 4: CDS spread widening precedes ratings actions
related to the implied rating
90 AA+ 40
80
30
70
60 AAA 20
50 10
AA+
40 AAA
AAA 0
30 AAA AAA
20 -10
10 -20
0 -19 -11 -2 6 15
USA France GermanyCanada Sweden U.K. Australia Ireland (Nov 5) Japan (Nov 6)
UK (Nov 10) UK (May 26)
Ireland (Jun 8) Spain (Dec 9)
CDS spreads (bp) Greece (Dec 7)
19
Default and recovery implicit in the term structure of sovereign CDS spreads, Pan and Singleton (2007)
20
See “Safe for Now: Deconstructing Sovereign Ratings” in Market Strategy Americas, July 16, 2009
8 January 2010 86
economies (Western Europe + Japan) shows a negative relationship between sovereign risk in
other countries and US rates (Figure 5).
In the near term, increased In Figure 6, we show the results of a regression of changes in the US 10y yield versus changes
G8 risk may be lead to in the average developed economy 5y CDS spread (Western Europe + Japan) and changes in
a rally in US rates the US 5y sovereign CDS spread itself. The latter variable was included to segregate the impact
of developed economy CDS (ex-US) on 10y yields 21. The coefficient on developed economy
CDS spreads is negative and statistically significant. A 1bp widening in sovereign risk is
associated with a 0.5bp rally in US 10y rates. Taken together with our prior estimates, this
would imply that a two-notch downgrade in other developed economies could translate to a
rally of 25bp in US rates. This underscores the special status of US debt as a safe haven. We
have analyzed the effect at weekly frequencies, as well as by looking at only large changes in
sovereign credit risk, and the results hold true in all these cases 22.
An analysis of large (more than one standard deviation) changes in other developed
economy sovereign CDS spreads is revealing. In weeks in which there was a one standard
deviation worsening in the sovereign CDS of other developed economies, US rates rallied
71% of the time. In weeks in which there was a one standard deviation tightening in other
country sovereign CDS, US rates sold off 71% of the time.
A possible reason why rising sovereign risk could benefit US rates is offered by foreign
reserve holdings. After Japan’s downgrades by Moody’s and Fitch in 1998, the proportion of
reserve assets held in JPY fell over the next few years from over 6% to 3%. The other reserve
currencies were beneficiaries of this decline. Negative news regarding the sovereign rating
of other developed economies may thus lead to allocation away from assets held in their
currencies. Some of these assets could be allocated into USD debt, slowing the recent trend
of diversification away from the US.
We believe that flight-to-quality flows should benefit front-end rates more than 10y rates.
Quantifying this statistically is tougher, as 2y yields are much more anchored to monetary
policy expectations than 10y. Needless to say, if there are concerns about developed
Figure 5: US 10y rates have rallied when concerns about Figure 6: Regression analysis of changes in US rates and
sovereign risk have flared sovereign risk in other countries
140 5.5 Δ(US

Δ(US 10y Δ(US 10y 10y rate,
120 5.0 rate, bp) rate, bp) weekly)
100 4.5 Δ(Other CDS, bp) -0.52 -0.57 -0.62
80 4.0 t-stat -2.76** -3.46** -2.91**
Δ(US CDS, bp) -0.10 - -
60 3.5
t-stat -0.5 - -
40 3.0
Intercept 0.06 -0.10 -0.35
20 2.5
t-stat 0.14 -0.29 -0.29
0 2.0 Observations 629 629 135
Aug-07 Feb-08 Aug-08 Feb-09 Aug-09
US 10y yield (%, R) non-US G8 CDS (bp, L)
21
The regression is over the period starting from August 2007, when the credit crisis started.
22
Several studies in the Sovereign CDS market find global sovereign risk to be a common risk factor. It is not surprising
that increases in this risk factor are associated with risk aversion rallies in US rates.
8 January 2010 87
economy sovereigns, we would expect US short-end rates to benefit more than 10y rates.
We explain our reasoning in the next section.
Medium term, concerns about US sovereign risk are set to rise

Fiscal metrics may deteriorate While prospects for a near-term US ratings downgrade are remote, the AAA rating could
considerably if projected come under threat over a three- to four-year horizon. Consider the two key financial
deficits are realized metrics 23 tracked by the ratings agencies:
1. Debt/GDP ratio. Making realistic adjustments to the CBO’s baseline deficit projections
results in a debt/GDP ratio that rises steadily to 100% over the next decade.
2. Interest coverage ratio. This is the ratio of interest cost to the government’s revenues.
This captures the overall size of public debt, the cost of issuance, and the revenue
generating capacity of the government. Historically, the highest this ratio has been for
the US federal government is about 18%. The typical number for Aaa countries,
according to Moody’s, is in the single digits. We expect this ratio to breach 18% in 2014
if adjusted CBO deficit projections are realized.
Further, given the backdrop of persistent deficits, even an innocuous rise in interest rates
could lead both key metrics to breach the comfort zone necessary for a AAA rating even
sooner. We elaborate below.
Projecting the interest coverage ratio

We have used a two-step process.
Estimation of primary deficit

The first step is the estimation of the primary deficit, i.e., total deficits before interest
payments, for which we have augmented the CBO’s baseline deficits with its own estimates
of the impact of future policy changes (Figure 7).
As background, the CBO’s baseline forecast uses current laws and policies and does not
account for possible changes. For instance, in the baseline scenario, the CBO assumes that
various tax cuts (individual and corporate) put in place would expire by the end of 2010,
thus boosting revenues and reducing deficits (Figure 7, line a). In the CBO’s estimate of the
president’s budget, published in June 2009, it assumed that most of the tax cuts would be
extended, except for taxpayers in the high income bracket (line b). This would, therefore,
result in higher deficit than assumed in the baseline. In addition, indexation of AMT to
inflation would also reduce revenues, as it would capture a smaller share of taxpayers (line
c). On the outlays side, the baseline deficit assumes discretionary outlays grow with
inflation; historically, they have moved in line with nominal GDP growth, thereby
understating outlays and deficits in the baseline (line e).
23
Moody’s Aaa sovereign monitor, December 2009
8 January 2010 88
Figure 7: Projecting the path of key fiscal metrics

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
CBO baseline primary deficit, Aug 09, $bn -1,184 -681 -293 -165 -82 -16 -29 15 53 0
a. EGTRRA/JGTRRA tax cut extension 0 -102 -185 -199 -210 -221 -230 -240 -250 -260
b. Other tax Provisions -12 -67 -151 -151 -160 -165 -173 -183 -188 -194
c. Indexing the AMT to inflation -7 -69 -31 -34 -37 -41 -46 -53 -60 -70
d. Interaction of a and c 0 -18 -39 -43 -46 -49 -52 -54 -57 -60
e. Appropriations grow @ nominal GDP -8 -33 -74 -120 -159 -194 -228 -261 -294 -328
Total primary deficit -1,210 -970 -773 -711 -695 -686 -758 -777 -796 -913
Interest cost, (along the forward path) 179 267 364 447 530 621 703 793 878 975
Total deficits -1,389 -1,237 -1,137 -1,158 -1,225 -1,306 -1,461 -1,570 -1,674 -1,887
Key fiscal metrics
Total deficit/GDP -10% -8% -7% -7% -7% -7% -8% -8% -8% -9%
Debt/GDP 59% 64% 68% 71% 75% 79% 84% 89% 94% 99%
Interest cost/revenue 8% 11% 13% 15% 17% 19% 21% 23% 24% 25%
Source: CBO, Barclays Capital
Adjusting for the above items with the CBO’s own estimates for each of these categories
results in a higher but more realistic trajectory for primary deficits. Figure 7 highlights that
the primary deficit alone would trough at $700bn or 4% of GDP.
Calculation of interest cost

Next, we estimate interest costs in each period. We make two key assumptions:
3. Forward rates are realized. For instance, 2y rates reach 3.25% and 4.5% in 2012 and
2015. respectively. Similarly, 10y rates reach 4.5% and 5% in 2012 and 2015
respectively.
4. Future issuance patterns evolve such that Treasury gradually achieves its objective of
terming out liabilities to its stated target of six to seven years.
Figure 8: Interest coverage ratio set to breach historical high of 18% by 2012-14, unless
rates stay low forever
Interest Cost/Revenues, %
50%
Rates 200bp above forwards
45%
Rates 100bp above forwards
40%
Rates along the forward path
35%
2012
30%
25%
20% 18%
15%
10% 2014-15
2013
5%
55% 65% 75% 85% 95% 105% 115%
Debt/GDP, %
8 January 2010 89
Debt at the end of each period is starting debt plus the primary deficit and interest cost in that
period. Total deficits/GDP trough at around 7%, while the marketable debt-to-GDP ratio rises
to 100% over the next 10 years. Significantly, the interest cost to revenue ratio increases to
18% by 2014-2015, the highest ratio ever for the US government, and keeps on rising.
Concerns about the rating may If rates were to move higher than forwards, an 18% interest coverage ratio could be
rise further if interest rates are breached even sooner. In Figure 8, we forecast the combination of the debt/GDP ratio and
higher than expected the interest coverage ratio in three scenarios: (a) along forwards, (b) higher than forwards
by 100bp by the end of 2013, and (c) higher than forwards by 200bp by 2013. Interest
coverage rises above 18% by 2013 and 2012 in scenarios (b) and (c), respectively. Needless
to say, given the current low level of rates, the scenarios we have considered in our analysis
are very innocuous, and remain highly plausible.
Also, note that we have not even factored in non-marketable government debt in our
analysis. Total public debt outstanding, as of end of FY09, was $11.8trn, not just $7trn.
State and local government debt would add another $2.3trn. Taking both into account, the
ratio of total federal, state, and local government debt to the GDP is already close to 100%.
Last, but not least, we have not included Fannie Mae (FNM)/Freddie Mac (FRE) debt
outstanding in government debt. Just because the government refuses to put it on the
balance sheet does not mean the taxpayer does not backstop the liabilities 24. Recent events,
including the Christmas Eve announcement by Treasury providing virtually unlimited capital
support to FNM/FRE, confirm that the government does not intend to put FNM/FRE
formally on the public balance sheet. However, from an economic perspective. removing the
capital backstop limits is no different than the US government assuming all GSE liabilities.
Including FNM/FRE debt outstanding in public finances would add another $2trn to the
outstanding public debt.
The key takeaway of our bottoms-up analysis is that a US ratings downgrade is not simply a
long-term concern. Rather, depending on the resolve of Congress to reduce deficits, the
future path of interest rates, a negative outlook warning, or even an actual downgrade
could occur over a three- to four-year horizon.
Diversification of reserves away from dollar assets may add fuel to the fire
While deficits can be lower than expected because of better economic growth or policy changes
on the tax front that lower debt funding needs, we believe that two of the widely discussed non-
fiscal solutions – namely a weakening of the dollar or higher inflation – are untenable.
Because a large proportion of US government debt is held by foreign investors, a conceptual

case is sometimes made that a weaker dollar reduces the value of foreign holdings, thereby
improving the ability of the US to repay its debt. We believe the opposite to be true, since
persistent deficits imply large future funding needs for the US government, which would
require continued participation by foreign investors.
In a prior publication 25, we suggested that that as long as the dollar’s status as a reserve
asset is stable, the US government can afford to finance large quantities of debt in the
international market at relatively low interest rates and avoid the prospect of a downgrade,
even at high debt/GDP ratios. We suggested a 50-55% level (from current levels close to
62%) for the dollar as percentage of world reserves that would begin to raise concerns
about its special status (Figure 10).
24
For details, see GSE 2010 Outlook,”The long and winding road”, as well as GSEs: Back to the Future, December 11,
2009
25
“Safe for Now: Deconstructing Sovereign Ratings” in Market Strategy Americas, July 16, 2009
8 January 2010 90
Currency composition of foreign exchange reserves (COFER) statistics published by the IMF
suggests a decline in dollar reserves relative to total reserves over much of the past decade
(from 72% to around 63%, Figure 7). However, the reported data are not adjusted for
currency fluctuations. On every reporting date, the amounts of reserve assets held in other
currencies are converted to an equivalent dollar amount using the prevailing exchange rate.
Thus, fluctuations in the value of the dollar relative to other currencies have an effect on the
total amount of reported reserves, as well as the percentage held in dollar assets.
If the pace of diversification For instance, if the dollar had depreciated since the last reporting date, reserves held in
away from the USD continues, it other currencies would get a boost in dollar terms, which would increase the overall level of
would reduce the ability of the reported reserves, as well as lower the ratio of reserves held in dollars.
US government to finance large
Figure 9 shows the reported proportion of dollars assets as a percentage of world reserves,
future deficits and increase
together with how the proportions would have looked if the current exchange rate was used
concerns about US sovereign risk
at every point in the past (thus taking out the effect of currency fluctuations). For the most
of the past decade, the adjusted proportion was rangebound – the decline in the reported
numbers is almost purely because of the depreciation of the dollar against the other major
reserve currencies over this period. Notably, the decline since the beginning of 2008 is due
to the diversification of central banks away from dollar assets, even though the extent was
masked by the strength of the USD.
Reaching the 50-55% level that we regard as a danger mark from current levels would
require either a further 30% decline in the value of the dollar relative to other reserve
currencies, or a 10% actual diversification by central banks away from the dollar, or a more
modest combination of both factors
From our vantage point, the decline in dollar reserves in currency-adjusted terms (2-3%
over the past year) is a source of concern. If diversification continues at the current pace,
the dollar’s share of global reserves may dip below 55% precisely when other financial
metrics, such as the debt/GDP ratio and interest coverage ratios, reach dangerous territory
three to four years hence.
Figure 9: Diversification away from the dollar as a reserve Figure 10: US AAA rating more sensitive to fiscal ratios as the
currency dollar loses reserve status
74%
300%
72%
250%
70%
200%
68%
66% 150%
64% 100%
62% 50%
60% 0%
Sep-99 Sep-01 Sep-03 Sep-05 Sep-07 Sep-09 Debt/GDP Contingent liabilities/GDP
% Reserves Held in $, Reported
World USD reserves: 50% 55% 60% 65%
Currency Adjusted
Source: IMF, Barclays Capital Source: Barclays Capital
8 January 2010 91
Inflation cannot be used as a way out of fiscal problems

The oft-expressed view that the US can simply inflate away its debt is also not credible. Not
only do government revenues and nominal GDP rise at the rate of inflation, so do outlays.
Thus, if the government plans to run large and sustained primary deficits, their size will also
increase in nominal terms at the same rate as inflation. In this environment, the Fed would
also lose its inflation credibility, thus increasing interest rates. In turn, this would increase
the cost of funding those future deficits as well as debt that needs to be rolled over. The rise
in interest cost could actually worsen the fiscal problem, accelerating the potential for a
ratings downgrade. To gauge if this is a likely outcome, the key relationship to explore is the
one between government spending and inflation.
What does the CBO assume about the impact of inflation on spending?
Inflation is unlikely to be effective Does higher inflation lead to higher government spending? In building its baseline projections
in reducing the debt burden if for outlays, the CBO differentiates between mandatory outlays and discretionary outlays.
the deficits are large and
The three largest components of mandatory outlays are social security, Medicare, and Medicaid.
persistent; it may even make
All three are likely to increase in a higher inflation environment. There are explicit cost of living
things worse
adjustments built into social security outlays. Further, rising inflation would also translate into
higher health care costs, which would increase outlays for Medicare and Medicaid.
As far as discretionary outlays are concerned, there is some question about how higher
inflation should affect them. In preparing its forecasts, one of the CBO’s guiding rules is that
discretionary spending grows at the same rate as inflation.
We have performed a scenario analysis of CBO baseline projections under different inflation
scenarios using these assumptions. The results for debt/GDP under these scenarios relative
to CBO’s baseline projections are shown in Figure 12. In these scenarios, inflation is
assumed to be unchanged until fiscal-year 2012 and subsequently, 2% and 4% higher than
baseline assumptions. The figure clearly shows that higher inflation does not reduce the
terminal debt/GDP but actually worsens it somewhat.
Figure 11: Statistical results for revenue, spending, inflation Figure 12: Effect of inflation on debt/GDP relative to CBO
and real growth baseline
Revenue Spending 2.0
growth (%) growth (%) Changes in Debt/GDP from CBO baseline (%)
1.8
Inflation 1.28 0.87 1.6
Std error 0.31 0.35 1.4
Real GDP growth 1.12 0.16 1.2
Std error 0.4 0.4 1.0
Intercept -0.01 0.04 0.8
Std error 0.02 0.02 0.6
R-squared 0.51 0.32 0.4
0.2
F-stat >
Granger causality critical value Conclusion 0.0
2008 2010 2012 2015 2017 2019
Inflation -> Spending 11.8 > 4.01 Yes
2% higher inflation 4% higher inflation
Spending-> Inflation 10.9 > 4.01 Yes
8 January 2010 92
Exploring the historical relationship between inflation, revenues, and spending

Away from CBO assumptions, let us look at how inflation has affected government
revenues in the past. A regression of changes in federal government revenues and outlays
over the past 50 years to inflation and real GDP growth (Figure 11) shows that government
revenues have historically increased with real GDP growth and inflation in a 1:1 ratio (and
the relationship is statistically significant).
Government spending, however, has a poor relationship with real GDP growth. This is
because spending tends to be policy driven and countercyclical. However, spending has a
nearly 1:1 relation with inflation. In sum, the effects of inflation on revenues and outlays
mostly offset each other. 26
Taken together, the common notion of “inflating your way out” is not as clear cut a solution
as it is made out to be; it can increase the size of deficits in nominal terms unless the
government finds a way to keep spending growth below inflation.
Trade ideas for the rates market

Forward steepeners to remain attractive
We expect increased concerns In the near term, we do not believe that a US rating downgrade is likely. Rather, sovereign risk
about sustainability to keep the is more of a concern for other developed economies such as the UK and Japan. Negative rating
curve steeper than forward rates news in these economies may lead to risk aversion led demand for US assets, especially at the
front end, which is least likely to be affected by concerns about US sovereign risk.
However, as outlined, the US sovereign rating is at risk over a three- to four-year horizon if
the path of deficits outlined by the CBO is realized. There are negative feedback loops
between interest rates and the sovereign rating – relatively innocuous changes in
assumptions about the future path of interest rates leave the US on a path that is no longer
consistent with a AAA rating. This in itself could lead to higher-than-expected interest rates.
Further diversification away from the dollar by global central banks may add fuel to the fire
by lowering the threshold necessary to trigger a ratings downgrade.
The solidly AAA rating near term, coupled with the possibility of negative ratings action in
the medium term, argues for the US yield curve to stay steeper than forwards. In our view,
investors currently do not factor in any sovereign risk premium when thinking about longer
maturity US Treasury yields, but one may be warranted.
Synthesizing our views leads us to favor forward starting 2s-10s steepeners. This view also
dovetails with our belief 27 that supply of US duration should outstrip demand over the next
few years and result in a steeper curve than is being priced by forwards.
Sovereign credit concerns could lead to a flattening of the swap spread curve
The swap curve may continue to The spread curve in countries with wider CDS spreads tends to be flatter on average (Figure
be inverted as long as fiscal 13). The US is in the top left hand corner of the chart. However, if concerns about US
concerns remain sovereign credit rise, it could move down the chart toward the bottom right. The US 5s30s
spread curve could further flatten as a result. There are two reasons for this:
26
A key issue here is causality. It is widely understood that high levels of government spending can be inflationary
because they lead to increased demand for goods. As a result, the significance of the coefficient between inflation and
government spending could be a result of increases in spending causing inflation, rather than the other way round.
Figure 8 also reports the results of granger causality tests for annual changes in spending and inflation over 1950-
2008; these suggests that high inflation both causes and is caused by a rise in government spending in a statistically
significant way. Statistically, the issue is difficult to resolve, and we do not attempt to do so here.
27
See ”Supply Deluge,” which appears in this publication.
8 January 2010 93
Figure 13: Spread curve tends to be flatter in countries with Figure 14: Bank credit spreads against sovereign credit
higher sovereign risk spreads
5s30s Spread Curve, bp 180 120

-35
160
Holland 100
-40 140
US
120 80
-45 Germany 100
60
-50 80
-55 60 40
France
40
-60 Belguim Spain 20
UK 20
-65 0 0
Italy
Jan-09 Mar-09 May-09 Jul-09 Sep-09 Nov-09
-70
0 25 50 75 100 125 US AAA financial - industrial spread (bp)
10y Sovereign CDS Spread, bp US 5y CDS (RHS, bp)
1. An increase in sovereign risk should cheapen long-term Treasuries relative to matched

maturity swap rates. While swap rates, irrespective of their tenor, are exposed to risk
just in 3m Libor, a 30y Treasury is exposed to sovereign risk over a 30y period. A
number of studies also highlight that long-term swap spreads tighten on rising budget
deficits, which also corresponds to increasing sovereign risk.
2. Concerns about sovereign credit also affect inter-bank rates, as the health of the country
will reflect on the health of the banking system (Figure 14). Our credit strategists regard
sovereign credit as an implicit floor for bank credit spreads, since financial institutions in
particular rely on direct and indirect state support. 28 This point is further underscored by
the way in which the credit rating agencies factor in state support in their judgments on
bank credit.
For instance, S&P uses a two-tiered system to rate banks. It provides a standalone credit
rating, which incorporates ongoing systemic support from the state. However, the final
credit rating is higher than the standalone rating because it additionally incorporates
extraordinary support from the government 29 Because of this, a sovereign downgrade
could directly flow through and impact bank credit ratings. In turn, higher bank credit
risk would lead to wider Libor-OIS spreads, leading to front-end swap spread widening if
a US rating downgrade becomes a realistic possibility.
Select sovereign and supranational dollar denominated debt offers value

Debt backed by more solidly A negative ratings action on a sovereign affects all debt backed by the government. For
AAA governments may offer instance, UK-guaranteed dollar denominated debt trades about 30bp cheap to US agency
greater value debt on average, and the spread has widened over the past few weeks (Figure 15). Events
such as the Greece downgrade have also cheapened foreign GGBs, as well as supranationals
such as KFW and EIB versus US agencies.
While this dynamic could persist in H1 10, over a longer horizon, governments such as
France, Germany, the Netherlands, and Australia have fewer challenges than the US and UK.
In particular, the rating agencies have categorized Germany and France as resistant to a
28
See “Credit gets caught up in sovereign risk”, in European Credit Alpha, 18 December 2009
29
See “Potential Legislation could constrain government support for financial institutions and affect ratings on
financial institutions”, S&P, December 16, 2009
8 January 2010 94
downgrade, while the US and UK are merely deemed resilient (if only because the former
names do not depend on any special status for their currencies and are less reliant on
foreign creditors).
While relative valuations are not as compelling as they were in mid-2009, we believe that
select foreign GGBs are cheap versus US agencies and TLG paper (Figure 16). In addition, on
a sovereign-risk adjusted basis, EIB/KFW should trade through FNM/FRE debt, especially in
the case of KFW, as it is owned by the German government. While this is unlikely in the near
term, given lingering worries about European sovereign risk, we recommend aggressive
buying of these names in the event of any inordinate cheapening in H1 10.
Figure 15: UK-backed paper has cheapened over the past Figure 16: Foreign GGB levels are cheap
few weeks
Spread to Libor, bp GGB ASW Sovereign CDS

45 bp 3y 3y 5y
40
US -16 (FRE) 33 37
35
-3 (TLGP)
30
UK 25 51 78
25
France 3 23 31
20
15 Germany -6 (KfW) 19 26
10 Netherlands 18 21 30
5 Sweden 20 38 52
0 Australia 21 33 39
6-Jul-09 20-Aug-09 4-Oct-09 18-Nov-09 2-Jan-10
RBS 2.625% May 12 (L)
8 January 2010 95
AGENCY MBS
Mortgage basis outlook: Life after Fed

This article is a reprint from the US Securitized Products Outlook 2010: Back from the brink.
Kumar Velayudham The Fed’s $1.25trn purchase program had an outsized effect on the agency MBS basis in
+1 212 412 2099 2009. Here, we examine how the mortgage market will fill the void when the purchases
savelayu@barcap.com stop after March.
We expect spreads to widen 30bp from current levels when the Fed purchases stop but
Nicholas Strand then to be met by strong demand. Fed purchases have had other positive effects, such
+1 212 412 2057 as richer rolls and lower mortgage volatility. In the absence of the Fed backstop and
nicholas.strand@barcap.com related benefits, agency MBS look rich. This is despite low supply - new issuance should
stay muted at $350-400bn for 2010, but pay-downs from Fed and GSE portfolios should
Matthew Seltzer also add to supply.
+1 212 412 1537
matthew.seltzer@barcap.com We think that banks, money managers, and foreign investors should all be buyers of MBS
in 2010, albeit at wider spreads. A host of factors – muted loan demand, low appetite for
Philip Ling credit risk, a steep yield curve, large cash holdings, and shortage of other spread products –
+1 212 412 3202 point to increased MBS demand from banks. International demand could also rise as risk
philip.ling@barcap.com aversion abates and reserve growth picks up.
Despite strong demand, MBS look rich to Treasuries, agency debt, and corporates on
both an absolute and a historical basis. We believe active money manager
participation will be needed to backstop spreads. Our analysis suggests that money
managers should provide a backstop when spreads are 30bp wider, which is where we
expect them to stabilize.
The risks and returns for the MBS markets seem well balanced in 2H10. The enormous
Treasury supply and the shortage of spread assets should help the basis. But potential
liquidity withdrawal could increase volatility and drag down performance.
Figure 1: Fed purchases of agency MBS ($bn) Figure 2: Net issuance ($bn)
35 1,200
$bn 600
30 1,000
500
25 41
800 178
400
20 176
600
15 300
135 213
400
10 200
200 272 91
5 100 206
0 0 80
0
Jan-09 Apr-09 Jul-09 Oct-09
2007 2008 2009 YTD
Weekly Net Purchase Cum Net Purchases, RHS FN FH GN
Source: Federal Reserve, Barclays Capital Source: Fannie Mae, Freddie Mac, Ginnie Mae, Barclays Capital
8 January 2010 96
Figure 3: Mortgage current coupon OAS (bp) Figure 4: Sector excess performance versus Treasuries (bp)
200 2006 2007 2008 2009 YTD

US Agg 85 -206 -710 713
150 MBS 122 -177 -232 532
Agencies 75 -52 -110 226
100 ABS 87 -634 -2223 2357
CMBS 137 -532 -3274 2710
50 Corporate 126 -523 -1988 2072
-50
Dec-99 Dec-01 Dec-03 Dec-05 Dec-07
Libor OAS TSY OAS
Source: Barclays Capital Note: 2009 values are through November 2009. Source: Barclays Capital
Review of 2009: Basis tightening and index outperformance

As we began to develop our longer-term outlook over the past few months, 30 it was clear
that the Fed’s role in the market was overwhelming in 2009 and that its eventual exit could
be similarly important. The Fed has thus far purchased $1.086trn in agency MBS and is set
to add approximately $165bn more through Q1 10 (Figure 1).
While the Fed’s emergence as a buyer of mortgages was meant to drive rates lower and
foster new purchase activity, net issuance of agency MBS was actually lower than in 2007-
2008 (Figure 2). Many competing forces, such as tighter underwriting standards and a
weak housing market, contributed to lower net issuance. We direct readers to “Agency
underwriting: When will the squeeze ease?” on page 97 for a more complete discussion.
As a consequence of this supply-demand imbalance, the mortgage basis tightened

considerably throughout 2009 (Figure 3). After widening to over L+100 in late 2008, the
current coupon has now rallied to levels that appear rich from a historical perspective. But
how did agency MBS fare against other spread product classes? Interestingly, agency MBS
Despite the strong performance,
lagged the performance of mortgage credit products (Figure 4). For example, while agency
agency MBS lagged the
MBS outperformed Treasuries by more than 500bp on a duration-adjusted basis, this was
performance of mortgage
less than one-fourth of the excess performance of the ABS and CMBS sectors.
credit products
Significant secondary effects of Fed purchases

The Fed’s actions have had significant effects on MBS beyond the basis tightening. These
spill-over effects have influenced day-to-day trading, hedging activity, and funding levels for
agency MBS.
Tradable float has been significantly reduced

One by-product of purchasing $1.25trn in agency MBS out of a $5.2trn market is that
liquidity in many sub-sectors can be greatly reduced. Figure 5 shows the outstanding
balances of conventional (Fannie and Freddie) 30y fixed rate MBS by coupon. From these
balances, we remove bonds purchased by the Fed and pools pledged to CMO trusts. This
leaves only 50% of the pools available to be traded (Figure 5). And if we consider the GSEs’
retained portfolios or other MBS owned in held-to-maturity (HTM) accounts, the tradable
float starts to look even thinner.
30
Please see The day the Fed stood still, October 9, 2009.
8 January 2010 97
Figure 5: Conventional 30y float stands at only $1.6trn Figure 6: Implied financing rates have dropped sharply (%)
1000 2.50
2.00
800 1.50
1.00
600
0.50
400 0.00
-0.50
200 -1.00
-1.50
0 Nov-08 Feb-09 May-09 Aug-09 Nov-09
<4 4 4.5 5 5.5 6 6.5 7 >7
1m Libor Implied Roll Funding
Float CMO Fed
Source: Federal Reserve, Fannie Mae, Freddie Mac, Barclays Capital Note: Implied funding calculated on FNCL 5.0s. Source: Barclays Capital
Implied funding for mortgages is at historical lows

Dollar rolls and implied financing rates go hand in hand with the reduced float. At the end of
2008, funding for agency MBS was noticeably more expensive than 1m Libor (Figure 6). A
by-product of the Fed’s relentless buying (especially during the first few months of the
program) was that implied financing rates in the dollar roll market declined sharply. This
effect was so powerful that rolls consistently traded to fail, with negative implied financing
rates, for much of the summer and fall. As a result, toward the end of 2009 the Fed sold
rolls in 5s and 5.5s to keep a technical situation from developing.
Mortgages traded to short empirical durations in 2009

Empirical durations were With Fed holding mortgages prices relatively steady, empirical durations were much shorter
shorter than model, despite than model effective durations during the year (Figure 7). As a result, MBS seemed to
higher callability of models outperform in a sell-off and underperform in a rally when hedged to model durations. Some
of the key benefits of shorter empirical durations have been better hedged carry and lower
mortgage volatility. Figure 8 shows that without the secondary effects of Fed purchases
(richer rolls and lower empirical durations), hedge-adjusted mortgage carry would have
been significantly lower.
Figure 7: Mortgages have traded to short empirical durations Figure 8: Limited carry without Fed support (32s/month)
10 4.5s 5.0s 5.5s 6.0s

TBA Carry 11 11.5 10.75 10
8 10Y Carry 14 14 14 14
Empirical HR 0.47 0.35 0.24 0.17
6 Model HR 0.67 0.52 0.39 0.34
4 Empirical Hedged Carry 4.4 6.6 7.4 7.6

Model Hedged Carry 1.7 4.2 5.4 5.2
2
Roll Richness 1.7 2.1 2.5 1.8

0
Dec-06 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09
Carry Without Fed Impact 0.0 2.1 2.9 3.4
Empirical Duration Model Duration
8 January 2010 98
The Fed and agency MBS in 2010

Now that we have considered the Fed’s role in agency MBS in 2009, we turn our attention
to 2010. We believe it is highly unlikely that the Fed will increase the size of the purchase
program beyond $1.25trn or extend the timeline for purchases. Aside from further reducing
the float and liquidity of the agency MBS market, additional purchases could face
congressional opposition. Moreover, assuming that the economy grows at 5% in Q1 10, as
forecast by our economics team, further quantitative easing (QE) is unlikely.
In the second half of the year, the Fed will probably begin to drain liquidity from the system.
We believe reverse repos are the likely tool of choice. Meanwhile, direct asset sales from the
Fed’s portfolio seem extremely unlikely, at least in 2010. Given the amount of MBS the Fed
owns, selling bonds outright could pressure the basis considerably and de-value the rest of
its agency MBS holdings.
Supply: Few changes from 2009

We expect about $10bn in With the Fed out of the picture, we examine the overall supply and demand landscape for
monthly run-off each from the agency MBS in 2010. Net supply should remain at $30-35bn per month, consistent with the
Fed and GSE portfolios run-rate of the past two years (Figure 9). Net supply will likely be driven primarily by new
and distressed home sales. Adding to this is the incremental run-off from the Fed and GSEs’
portfolios. Although we expect a subdued prepayment environment, the Fed also owns
higher-coupon MBS that could experience meaningful speeds. Figure 10 shows the
distribution of the Fed’s MBS holdings – note the sizeable amount in >=5% coupons. We
expect about $10bn in monthly run-off from the Fed, or $125bn for the year.
The GSEs face a mandate to reduce their retained portfolios by 10% per year. They will likely
find it difficult to sell their non-agency MBS or whole loan holdings, which may put more
pressure on their agency MBS portfolios (Figure 11). Overall, we expect the GSEs to
contribute an additional $10-12bn of supply per month, or $150bn over the course of 2010.
Admittedly, the 10% rule can be changed by Treasury without congressional approval
which could ease GSE selling pressure, but this is not our base case expectation. On the
other hand, GSE buyouts of delinquent loans could pick up in early 2010 (Figure 12). In this
case, the need to make room for delinquent loans in their retained portfolio could cause the
GSEs to sell MBS, which could exacerbate basis widening and volatility.
Figure 9: Agency net monthly supply over the past two Figure 10: Fed portfolio coupon distribution
years ($bn)
Net issuance ($bn) 450

100 400
80 350
60 300
250
40
200
20
150
0
100
-20
50
-40 0
Jan-08 Jul-08 Jan-09 Jul-09 3.5 4 4.5 5 5.5 6 6.5
FHLMC FNMA GNMA Fed Purchases ($bn)
Source: Fannie Mae, Freddie Mac, Ginnie Mae, Barclays Capital Source: Federal Reserve
8 January 2010 99
Figure 11: Composition of GSE retained portfolio ($bn) Figure 12: GSE delinquency pipeline (%)
800 Serious delinquency rate

5%
600
4%
3%
400
2%
200
1%
0 0%
FN FH Jan-06 Apr-07 Jul-08 Oct-09
Agency MBS Non-agency MBS Whole loans FNMA FHLMC
Note: As of October 2009. Source: FN/FH Monthly Summary Source: FN/FH Monthly Summary
Finally, we expect $350bn in net issuance, mainly from home sales. Add the amounts coming
from the GSEs and the Fed, and we end up with roughly $625bn (350 + 125 + 150) for the
market to absorb next year. What are the risks to this forecast? First, changes in underwriting
standards (as discussed in “Agency underwriting: When will the squeeze ease?” page 97)
could affect issuance. For example, tighter FHA standards could limit GNMA supply, while
lower standards from the GSEs could shift more production into FN/FR securities. Second,
higher rates could reduce affordability, pushing supply projections downward.
Demand side: The usual suspects

Over the last five recessions, the We now examine the demand side to see whether private demand can accommodate the
banking system increased its $625bn in agency MBS supply in 2010. The biggest sources of demand should be the usual
security holdings by 15% in the suspects – banks, money managers, and foreign investors.
year following the recession
Banks: A steep curve and a cash surplus should keep demand strong
Banks face an important decision in 2010 – should they expand their loan books or their
securities portfolios? To answer this question, we look at some of the factors that drive the
choice:
A) Risk appetite and loan demand: During periods of heightened risk aversion, banks
generally prefer to hold securities with limited credit risk. Underwriting tends to tighten,
causing loan demand to drop off.
B) Attractiveness of securities versus loans: When the yield curve is steep, banks move
away from loans to securities, as they can meet their NIM targets without taking credit
risk (Figure 14). And when the front end is low, current yields on C&I loans are limited,
since they are mostly floating-rate loans.
These factors have meant that over the past five recessions, the banking system increased
its security holdings by around 15% in the year following the recession (Figure 13). On a
balance sheet level, banks have typically increased security holdings by 2-3% of total
balance sheet in the first year after a recession. With their current security holdings close to
$2trn and a balance sheet of around $13.5trn, we estimate that banks could add $300-
400bn in securities during 2010.
8 January 2010 100

Figure 13: Bank securities versus loans, 1y change (%) Figure 14: Bank MBS holding versus curve
25% 3 60,000
20% 2.5 50,000
15% 2 40,000
10% 1.5 30,000
5% 1 20,000
0% 0.5 10,000
-5% 0 -
-10% -0.5 (10,000)

1970 Q1 1977 Q1 1984 Q1 1991 Q1 1998 Q1 2005 Q1 Mar-95 Mar-98 Mar-01 Mar-04 Mar-07
Recession Securities Loans Agency MBS Addition ($mn, RHS) 1s5s slope (%)
Source: Federal Reserve H8 Data, Barclays Capital Source: FDIC, Barclays Capital
Shortage of other spread What percentage of this security buying will be in agency MBS? Historically, banks have held
products argues for a higher close to 50% of securities in agency MBS pass-throughs and CMOs, equating to $150-
allocation to agency MBS 200bn in bank demand for MBS in 2010 (Figure 15). The shortage of other spread products
(non-agencies, CMBS, ABS, etc.), however, may argue for a higher allocation to agency MBS
during this cycle. Also, the banking system is currently flush with cash (Figure 16). This
should also encourage the move into agency MBS. We believe that a good portion of this
buying will be in short-duration assets (hybrids, short CMOs, and 15y), given concerns
about extension risk.
Foreign demand: Risk reversal

During the credit crisis of 2008-09, foreign investors allocated away from riskier US assets
(agency debt, MBS, and corporates) and into Treasuries (Figure 17). Specifically, foreign
holdings of bills surged by $400bn. This risk aversion seems to have normalized over recent
months – MBS holdings have stopped falling, and foreign investors have continued to
reinvest paydowns. We expect a continued, yet moderate, risk reversal. We expect foreign
reserves to grow by $1trn in 2010, of which $400-500bn should find its way back to US
Figure 15: MBS+CMO % of securities, bank portfolios Figure 16: Cash % of total assets, bank portfolios
60% 12%
50% 10%
40% 8%
30% 6%
20% 4%
2%
10%
0%
0%
1992-05 1995-11 1999-05 2002-11 2006-05 2009-11
Q1/1994
Q3/1996
Q1/1999
Q3/2001
Q1/2004
Q3/2006
Q1/2009
Cash as % of assets
Source: FDIC, Barclays Capital Source: FDIC, Barclays Capital
8 January 2010 101

securities. In addition, $1.2trn of short-term securities are expected to roll off during the
year. If these two cash flows are reallocated in line with current holdings, foreign investors
could add around $100bn in agency MBS in 2010.
Figure 17: Foreign holdings of MBS have stabilized, but look to pick up in 2010 ($bn)
Type of security Jun-05 Jun-06 Jun-07 Jun-08 Dec-08 Mar-09 Oct-09
Long-term securities 6,262 7,162 9,136 9,463 8,792 8,212 9,528

Equities 2,144 2,430 3,130 2,969 2,101 1,867 2,528
Debt 4,118 4,733 6,007 6,494 6,691 6,345 7,000
US Treasury 1,599 1,727 1,965 2,211 2,634 2,439 2,733
Agency MBS 264 386 570 773 645 599 600
Agency debt 527 599 735 691 651 640 588
Corporate 1,729 2,021 2,738 2,820 2,761 2,667 3,079
Short-term debt 602 615 635 858 1,158 1,239 1,223
Total 6,864 7,778 9,772 10,322 9,950 9,451 10,751
Source: TIC, Barclays Capital
Money managers: Entry at wider spreads

Money manager universe is now When the Fed purchases end, monthly supply could spike to $50-60bn, helped by Fed and
significantly underweight agency GSE pay-downs. Strong demand from banks and foreign investors may not be able to
MBS, due to tight spreads absorb this supply; however, we expect that money managers will also add MBS in size and
help keep spreads contained. As stated earlier, we believe that the money manager universe
is now significantly underweight agency MBS, due to tight spreads. Hence, money
managers are unlikely to buy until spreads widen. The question becomes, “at what spread
level do MBS look attractive to money managers?”
Figure 18: Supply vs demand, post-Fed purchases (monthly) Figure 19: Money manager MBS allocation
Supply % in Agency MBS

Net
Net Issuance $30-35Bn Assets
Fund Name ($bn) 12/17/09 12/31/08 Change
Fed PayDowns $10Bn
PIMCO Total Return Fund $185.6 22% 62% -40%
GSE Paydowns $10-15Bn Vanguard Total bond Market $67.7 36% 37% 0%
Total $50-60Bn Index
The Bond Fund of America $37.9 17% 15% 2%
Demand Dodge & Cox Income $17.9 41% 47% -5%
Banks $15-20Bn Fidelity Advisor Total Bond $12.0 12% 14% -1%
Foreign $10Bn Western Asset Core Plus Bond $7.7 40% 60% -20%
Openheimer Strategic Income $8.1 9% 12% -3%
Money Managers ???
Fund
Lord Abbett Bond Debenture $7.1 2% 14% -12%
Combined $344.0 -23%
Source: Barclays Capital Note: We use the latest holdings available for the largest eight funds indexed to
the Barclays US Aggregate Index. Source: Fund Quarterly Statements, Barclays
Capital
8 January 2010 102

Figure 20: MBS spread to agency debt (agency OAS, bp) Figure 21: MBS and high quality credit spreads (bp)
80 400
40 300
200
0
100
-40
0
-80
-100
Oct-99 Oct-01 Oct-03 Oct-05 Oct-07 Oct-09
-120
Nov-02 Nov-04 Nov-06 Nov-08 CC OAS (bp) Credit Spread(bp)
MBS looks rich to Treasuries, When we compare agency MBS with agency debt, the spread (adjusted for optionality) has
corporates and agency debt, historically been close to zero (Figure 20), but it has tightened since the start of the Fed
both on an absolute and purchases. But even though we have seen some moderation, MBS still look 40bp tight to
historical basis agency debt. Mortgage Treasury option-adjusted spreads also look significantly rich
compared with historical levels (Figure 3). We also compare MBS with corporate debt and
look at the spread to Treasuries for the top 20% industrial names (to limit the effect of
credit risk) (Figure 21). These names have historically traded 30bp back of agency MBS,
compensating for lower liquidity. Currently, the spread is close to 100bp, although some of
this due to the credit cycle. In summary, we believe that MBS spreads need to widen 30-
40bp before money managers add MBS in size.
Basis call: A stable basis at wider levels

We anticipate that spreads could We expect spreads to remain choppy at the start of the new year, with the Fed still a force to
widen 30-40bp before money reckon with. As we get closer to 2Q10 and the Fed exit nears, spreads should widen steadily.
managers come back and Bank and international demand may not be strong enough to hold the basis at these rich
stabilize the basis levels. These investors might also be wary of the impact of the Fed’s exit and might wait for
wider spreads before adding MBS in size. Consequently, we anticipate that spreads could
widen 30-40bp before money managers, who are currently underweight the index, come
back and stabilize the basis.
We do not expect spreads to “blow out” much beyond 30-40bp for a number of reasons:
1) money managers are currently underweight and should add at wider spreads; 2) the Fed
could step in if mortgage rates rise too fast; 3) the enormous liquidity in the system should
keep spread assets well bid; and 4) the GSEs could provide a local backstop if MBS cheapen
significantly.
Things to watch out for

While our base case is for relatively stable spread widening, convexity flows could change that
story. If rates rise sharply, MBS could extend and trade to much longer durations. Tight
underwriting and low HPA also add some extension risk. In a sell-off, convexity flows from
servicers could exaggerate spread widening. Our rates forecast, however, calls for a slow rise
in rates in the first half of the year, so a big convexity-shedding episode is not our base case.
8 January 2010 103

On the other hand, there is technical risk to our short basis call. There is a lot of short
interest, both implicit and explicit, in the agency basis now. Much of this has come from
crossover investors in equities and other parts of fixed income. A sharp rally could force
these shorts to cover, further richening MBS. Also, although it is unlikely, in our view, there
is an outside chance that the Fed program could be extended. Both are risks to keep in
mind, but we do not expect either to play out.
For H2 10, we think that On the GSE front, we believe that FAS 166/167 should have limited effect on buyouts (for
risks and rewards are details, please see “Less than meets the eye,” Securitized Products Weekly, November 20,
well balanced 2009 But if the GSEs decide to buy out delinquent loans aggressively, the coupon stack will
obviously compress quite a bit. Another factor to keep in mind is the cheapening of rolls. We
see the removal of excess reserves by the Fed (in 2H10) as the primary driver of roll
cheapening, not the end of Fed purchases, so this is not a story for the next few months.
In fact, most of this article has focused on H1 10. Looking into H2 10, we think that risks
and rewards are well balanced. On the one hand, the enormous Treasury supply and the
shortage of spread assets should cause risky assets to outperform. On the other hand, if the
Fed starts to remove liquidity, then volatility could pick up and mortgages could
underperform. Either way, it promises to be an exciting year for agency MBS.
8 January 2010 104

US FIXED INCOME RESEARCH
Ajay Rajadhyaksha
Head of US Fixed Income Strategy Securitised
Products Strategy
+1 212 412 7669
ajay.rajadhyaksha@barcap.com
Joseph Abate Piyush Goyal James Ma

US Fixed Income Research Fixed Income Strategy Fixed Income Strategy
+1 212 412 6810 +1 212 412 6793 +1 212 412 2563
joseph.abate@barcap.com piyush.goyal@barcap.com james.ma@barcap.com
Chirag Mirani Amrut Nashikkar Michael Pond

Fixed Income Strategy Fixed Income Strategy Treasury and Inflation-linked Strategy
+1 212 412 6819 +1 212 412 1848 +1 212 412 5051
chirag.mirani@barcap.com amrut.nashikkar@barcap.com michael.pond@barcap.com
Anshul Pradhan Rajiv Setia

Treasury and Inflation-linked Strategy Fixed Income Strategy
+1 212 412 3681 +1 212 412 5507
anshul.pradhan@barcap.com rajiv.setia@barcap.com
8 January 2010 105

Analyst Certification(s)
We, Ajay Rajadhyaksha, Joseph Abate, Piyush Goyal, James Ma, Chirag Mirani, Amrut Nashikkar, Michael Pond, Anshul Pradhan, Rajiv Setia, Kumar
Velayudham, Nicholas Strand, Matthew Seltzer, and Philip Ling, hereby certify (1) that the views expressed in this research report accurately reflect our
personal views about any or all of the subject securities or issuers referred to in this research report and (2) no part of our compensation was, is or will be
directly or indirectly related to the specific recommendations or views expressed in this research report.
Important Disclosures
For current important disclosures regarding companies that are the subject of this research report, please send a written request to: Barclays Capital
Research Compliance, 745 Seventh Avenue, 17th Floor, New York, NY 10019 or refer to https://ecommerce.barcap.com/research/cgi-
bin/all/disclosuresSearch.pl or call 212-526-1072.
Barclays Capital does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that Barclays
Capital may have a conflict of interest that could affect the objectivity of this report. Any reference to Barclays Capital includes its affiliates. Barclays
Capital and/or an affiliate thereof (the "firm") regularly trades, generally deals as principal and generally provides liquidity (as market maker or otherwise)
in the debt securities that are the subject of this research report (and related derivatives thereof). The firm's proprietary trading accounts may have either
a long and / or short position in such securities and / or derivative instruments, which may pose a conflict with the interests of investing customers.
Where permitted and subject to appropriate information barrier restrictions, the firm's fixed income research analysts regularly interact with its trading
desk personnel to determine current prices of fixed income securities. The firm's fixed income research analyst(s) receive compensation based on various
factors including, but not limited to, the quality of their work, the overall performance of the firm (including the profitability of the investment banking
department), the profitability and revenues of the Fixed Income Division and the outstanding principal amount and trading value of, the profitability of,
and the potential interest of the firms investing clients in research with respect to, the asset class covered by the analyst. To the extent that any historical
pricing information was obtained from Barclays Capital trading desks, the firm makes no representation that it is accurate or complete. All levels, prices
and spreads are historical and do not represent current market levels, prices or spreads, some or all of which may have changed since the publication of
this document. Barclays Capital produces a variety of research products including, but not limited to, fundamental analysis, equity-linked analysis,
quantitative analysis, and trade ideas. Recommendations contained in one type of research product may differ from recommendations contained in
other types of research products, whether as a result of differing time horizons, methodologies, or otherwise.
This publication has been prepared by Barclays Capital, the investment banking division of Barclays Bank PLC, and/or one or more of its affiliates as
provided below. This publication is provided to you for information purposes only. Prices shown in this publication are indicative and Barclays Capital is
not offering to buy or sell or soliciting offers to buy or sell any financial instrument. Other than disclosures relating to Barclays Capital, the information
contained in this publication has been obtained from sources that Barclays Capital believes to be reliable, but Barclays Capital does not represent or
warrant that it is accurate or complete. The views in this publication are those of Barclays Capital and are subject to change, and Barclays Capital has no
obligation to update its opinions or the information in this publication. Barclays Capital and its affiliates and their respective officers, directors, partners
and employees, including persons involved in the preparation or issuance of this document, may from time to time act as manager, co-manager or
underwriter of a public offering or otherwise, in the capacity of principal or agent, deal in, hold or act as market-makers or advisors, brokers or
commercial and/or investment bankers in relation to the securities or related derivatives which are the subject of this publication.
The analyst recommendations in this report reflect solely and exclusively those of the author(s), and such opinions were prepared independently of any
other interests, including those of Barclays Capital and/or its affiliates.
Neither Barclays Capital, nor any affiliate, nor any of their respective officers, directors, partners, or employees accepts any liability whatsoever for any
direct or consequential loss arising from any use of this publication or its contents. The securities discussed in this publication may not be suitable for all
investors. Barclays Capital recommends that investors independently evaluate each issuer, security or instrument discussed in this publication and
consult any independent advisors they believe necessary. The value of and income from any investment may fluctuate from day to day as a result of
changes in relevant economic markets (including changes in market liquidity). The information in this publication is not intended to predict actual results,
which may differ substantially from those reflected. Past performance is not necessarily indicative of future results.
This communication is being made available in the UK and Europe to persons who are investment professionals as that term is defined in Article 19 of the
Financial Services and Markets Act 2000 (Financial Promotion Order) 2005. It is directed at, and therefore should only be relied upon by, persons who
have professional experience in matters relating to investments. The investments to which it relates are available only to such persons and will be entered
into only with such persons. Barclays Capital is authorized and regulated by the Financial Services Authority ('FSA') and member of the London Stock
Exchange.
Barclays Capital Inc., US registered broker/dealer and member of FINRA (www.finra.org), is distributing this material in the United States and, in
connection therewith accepts responsibility for its contents. Any U.S. person wishing to effect a transaction in any security discussed herein should do so
only by contacting a representative of Barclays Capital Inc. in the U.S. at 745 Seventh Avenue, New York, New York 10019.
Subject to the conditions of this publication as set out above, ABSA CAPITAL, the Investment Banking Division of ABSA Bank Limited, an authorised
financial services provider (Registration No.: 1986/004794/06), is distributing this material in South Africa. Any South African person or entity wishing to
effect a transaction in any security discussed herein should do so only by contacting a representative of ABSA Capital in South Africa, 15 ALICE LANE,
SANDTON, JOHANNESBURG, GAUTENG, 2196. ABSA CAPITAL IS AN AFFILIATE OF BARCLAYS CAPITAL.
Non-U.S. persons should contact and execute transactions through a Barclays Bank PLC branch or affiliate in their home jurisdiction unless local
regulations permit otherwise.
In Japan, foreign exchange research reports are prepared and distributed by Barclays Bank PLC Tokyo Branch. Other research reports are distributed to
institutional investors in Japan by Barclays Capital Japan Limited. Barclays Capital Japan Limited is a joint-stock company incorporated in Japan with
registered office of 2-2-2, Otemachi, Chiyoda-ku, Tokyo 100-0004, Japan. It is a subsidiary of Barclays Bank PLC and a registered financial instruments
firm regulated by the Financial Services Agency of Japan. Registered Number: Kanto Zaimukyokucho (kinsho) No. 143.
Barclays Bank PLC Frankfurt Branch is distributing this material in Germany under the supervision of Bundesanstalt fuer Finanzdienstleistungsaufsicht
(BaFin). This material is distributed in Malaysia by Barclays Capital Markets Malaysia Sdn Bhd.
IRS Circular 230 Prepared Materials Disclaimer: Barclays Capital and its affiliates do not provide tax advice and nothing contained herein should be
construed to be tax advice. Please be advised that any discussion of U.S. tax matters contained herein (including any attachments) (i) is not intended or
written to be used, and cannot be used, by you for the purpose of avoiding U.S. tax-related penalties; and (ii) was written to support the promotion or
marketing of the transactions or other matters addressed herein. Accordingly, you should seek advice based on your particular circumstances from an
independent tax advisor.
© Copyright Barclays Bank PLC (2010). All rights reserved. No part of this publication may be reproduced in any manner without the prior written
permission of Barclays Capital or any of its affiliates. Barclays Bank PLC is registered in England No. 1026167. Registered office 1 Churchill Place, London,
E14 5HP. Additional information regarding this publication will be furnished upon request.
8 January 2010 LAST PAGE

Positioning For A Liquidity Drain: U.S. Interest Rates Outlook 2010

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Positioning For A Liquidity Drain: U.S. Interest Rates Outlook 2010

Uploaded by

Copyright:

Available Formats

INTEREST RATES RESEARCH

U.S. INTEREST RATES OUTLOOK 2010

Money markets: Short rates – Choppy seas ahead 23

Treasuries: Pick your poison 32

TIPS: Still cheap insurance 42

Agencies: The long and winding road 48

INTEREST RATE DERIVATIVES

Swap spreads: Weighed down by supply 65

Vol: Entering a quiet period 74

Mortgage basis outlook: Life after Fed 96

US rates markets: A new, new world

US fixed income markets: Life without the Fed

Figure 1: Absence of Fed to be felt in 2010

Treasury, ex-bills 191 101 332 1,556 1,256 1,840 1,840

Ex-Ante Demand Total Fed Funds 3m Libor 2y 5y 10y 30y 10y RY

Source: US Treasury, Flow of Funds, Barclays Capital Source: Barclays Capital

Rates products: Swaps, volatility and the TIPS markets

Agency debt and the GSEs: Don’t hold your breath

US securitized markets: Stepping back from the limelight

EOY 2008 9-Jun Diff 60+ (%

Home prices and bank losses: Tail risk has declined

Sq ft (millions) YOY % Chg in NFP Homes (000s)

300 forecast 4% 3,000

Source: BLS, Barclays Capital Source: Loan Performance, Barclays Capital

Figure 8: Bank losses over the next few years ($bn)

1st lien mortgages 1741 11% 200 75 63 63

Regulatory changes and other wild cards

US FIXED INCOME SUPPLY/DEMAND

Figure 1: Absence of Fed to be felt in 2010

Treasury, ex-bills 191 101 332 1,556 1,256 1,840 1,840

Treasury supply: As far as the eye can see

Figure 2: 2010 Treasury issuance calendar

January 44 40 43 33 21 13 10 204 53 151 128

Spread products: A sharp swing

3,000 40% 2,000 10y Eq

 Securitized: Securitized issuance was dominated by agency MBS, as net issuance of

Private demand: Not enough to fill the void

Foreign demand: FX reserve growth aided by terming out

Figure 7: Diversification under way Figure 8: $400-450bn to be invested in USD

74% Pace of diversification

35% Other Treasuries

Source: Treasury Source: FDIC

Banks: Safety of liquid securities

Recessions Loans / Bank Credit, % FOF yearly savings Rate, % Estimated

Source: Federal Reserve Source: Federal Reserve, Barclays Capital

Individuals: High savings aided by terming out

Asset composition of investment

30% Household investment 2009 up to Q309

25% a. Net savings 436

yoy growth, HH MMF shares, % Retirement A/C - 53 53

Source: Federal Reserve, Barclays Capital Source: Federal Reserve

Short term 150 65%

Defined benefit pensions: Increasing allocation to fixed income

Conclusion: Steeper curve/Tighter spreads

Figure 18: Supply-demand imbalance of term-fixed income securities

Ex-ante demand Total Tsy Spread pdts. Source

Overseas +600 +600 0 Reserve growth/terming out

Source: Milliman Source: Barclays Capital

Short rates: Choppy seas ahead

How much to drain?

Figure 1: Bank reserves ($bn) Figure 2: Bank daylight overdraft ($bn)

1,600 300 Fed begins paying

Securitized: Securitized issuance was dominated by agency MBS, as net issuance of

Heading into midterm elections, incumbent Democrats in Congress may be unwilling to

Endorsement of alternatives to traditional mortgage finance, including covered bonds,