Professional Documents
Culture Documents
Fall 2013
Topics covered
by the module.
Topics covered
by the textbook
Because of this reason, class attendance is crucial. Reading through the textbook
will not replace missed classes!!
2
Announcement
If you find any errors, typos, grammar or spelling mistakes, etc., please let me know.
The lectures are recorded for your convenience. (We do not take responsibility for any
failure in recording due to system errors, so please try to attend the class as much as
possible.)
There will be an assignment (a collection of small questions) which counts 20% of the final
mark. The questions will be different between BEE2028 students and BEE3043 students.
The details will be posted later.
The final exam will be a closed-book/note exam. Only calculators will be permitted.
Math foundations
Summation notation ()
http://www.youtube.com/watch?v=8i9-9zHbW6g
http://www.youtube.com/watch?v=U74zP3BsAbM
Matrix multiplication
http://www.youtube.com/watch?v=kuixY2bCc_0
http://www.youtube.com/watch?v=OvTEhNL96v0
http://www.youtube.com/watch?v=35NWFr53cgA
http://www.youtube.com/watch?v=52RiUx5B-ms
http://www.youtube.com/watch?v=5h6YzRRxzO4
Differentiation (Derivatives)
http://www.youtube.com/watch?v=54KiyZy145Y
http://www.youtube.com/watch?v=-EnecLe_0B4
Topics (plan)
Credit risk
Interest risk
Collateral risk
Liquidity risk
Belief update
Hidden type
Adverse selection
Adverse selection in banking and use of collateral
General lesson: How to separate types?
Hidden action
Moral hazard in banking and use of collateral
Moral hazard in banking and use of capital
Off-balance-sheet banking
Options
No-arbitrage pricing
Loan commitment
Revision: Return
date-(t+1) price
Pt +1 + Dt +1
1
Rt +1 =
Pt
Return on portfolio
net
date-t price
Stock B
Stock C
Total
Invested amount
50
25
25
100
Realized return
5%
20%
-10%
5%
52.5
30
22.5
105
Final value
Revision: Short-selling
Short-selling
A principle of capital gain
Expectation of price increase
Buy it, and sell it after the price increases.
Expectation of price decrease
Borrow it and sell it. After the price decreases, buy and return
it. (= short-selling)
Investors can sell a thing even if they dont have it. (In practice,
there are regulations and limitations for short-selling.)
State 1:
Hot
0.5
9K
8K
State 2:
Mild
0.3
5K
6K
State 3:
Cold
0.2
3K
4K
Weighted
Average
6.6 6.6
K
K
means
11
Prob.
X X
(X X)2
State 1:
Hot
0.5
9K
8K
2.4 K
5.76 (K)2
State 2:
Mild
0.3
5K
6K
-1.6 K
2.56 (K)2
State 3:
Cold
0.2
3K
4K
-3.6 K
12.96 (K)2
6.24 (K)2
Weighted
Average
6.6 6.6
K
K
means
14
Prob.
X X
(X X)2
Y - Y
(Y Y)2
State 1:
Hot
0.5
9K
8K
2.4 K
5.76 (K)2
1.4 K
1.96 (K)2
State 2:
Mild
0.3
5K
6K
-1.6 K
2.56 (K)2
-0.6 K
0.36 (K)2
State 3:
Cold
0.2
3K
4K
-3.6 K
12.96 (K)2
-2.6 K
6.76 (K)2
(K)2
variance
0
2.44 (K)2
Weighted
Average
6.6 6.6
K
K
means
6.24
Prob.
X X
(X X)2
Y - Y
(Y Y)2
State 1:
Hot
0.5
9K
8K
2.4 K
5.76 (K)2
1.4 K
1.96 (K)2
State 2:
Mild
0.3
5K
6K
-1.6 K
2.56 (K)2
-0.6 K
0.36 (K)2
State 3:
Cold
0.2
3K
4K
-3.6 K
12.96 (K)2
-2.6 K
6.76 (K)2
(K)2
variance
0
2.44 (K)2
Weighted
Average
6.6 6.6
K
K
means
6.24
X = 6.24
= 2.50 K Standard
deviation
Y = 2.44
= 1.56 K
16
Prob.
X X
(X X)2
Y - Y
(Y Y)2
(X - X)(Y - Y)
State 1:
Hot
0.5
9K
8K
2.4 K
5.76 (K)2
1.4 K
1.96 (K)2
3.36 (K)2
State 2:
Mild
0.3
5K
6K
-1.6 K
2.56 (K)2
-0.6 K
0.36 (K)2
0.96 (K)2
State 3:
Cold
0.2
3K
4K
-3.6 K
12.96 (K)2
6.76 (K)2
9.36 (K)2
6.24 (K)2
-2.6 K
variance
0
2.44 (K)2
3.84 (K)2
covariance
Weighted
Average
6.6 6.6
K
K
means
X = 6.24
= 2.50 K Standard
deviation
Correlation Coefficient
X,Y = Cov( X , Y ) =
XY
Y = 2.44
= 1.56 K
3.84( K ) 2
= 0.984
2.50 K 1.56 K
18
0
No
correlation
1 Perfect
correlation
Most studies report standard deviations and correlation coefficients, but not
variances or covariances.
If you need variances and covariances, you need to recover them from st.
deviations and correlation coefficients.
Std. deviations equivalent variances
covariance
Correlation coef.
19
Tutorial questions
State 1:
Hot
0.5
8K
2K
State 2:
Mild
0.3
7K
6K
State 3:
Cold
0.2
4K
9K
Weighted
Average
X X
(X X)2
Y - Y
(Y Y)2
(X - X)(Y - Y)
Y =
X,Y =
_
X =
Exercise2.
What is the expectation and standard deviation of the number you get from rolling a die?
20
British Gas
Capita
Diageo
Easyjet
1.10%
0.90%
1.05%
1.15%
1.20%
British Gas
Capita
Diageo
Easyjet
290 %2
131 %2
58 %2
24 %2
113 %2
290 %2
96 %2
34 %2
31 %2
53 %2
BG
96 %2
131 %2
46 %2
27 %2
45 %2
Capita
34 %2
46 %2
58 %2
15 %2
24 %2
Diageo
31 %2
27 %2
15 %2
24 %2
22 %2
Easyjet
53 %2
45 %2
24 %2
22 %2
113 %2
21
Aviva
British Gas
Capita
Diageo
Easyjet
17%
11%
8%
5%
11%
Correlation matrix
Aviva
BG
Capita
Diageo
Easyjet
Aviva
1.000
0.495
0.263
0.368
0.294
BG
0.495
1.000
0.529
0.482
0.368
Capita
0.263
0.529
1.000
0.392
0.301
Diageo
0.368
0.482
0.392
1.000
0.417
Easyjet
0.294
0.368
0.301
0.417
1.000
22
E[ w1 X 1 + w2 X 2 + + wn X n ] = w1 E[ X 1 ] + w2 E[ X 2 ] + + wn E[ X n ]
EX) E[3X1+0.5X2-10X3+8X4] = 3 E[X1] + 0.5 E[X2] -10 E[X3] + 8 E[X4]
w1
If we denote the
w2
expectation vector by ( 1 , 2 , , n ), and the coefficients by w
,
w
n
w.
then the expectation of a linear combination is simply
w = 1w1 + 2 w2 + + n wn .
23
Aviva
British Gas
Capita
Diageo
Easyjet
1.10%
0.90%
1.05%
1.15%
1.20%
Variance formulas
Var[ X + Y ] = Var[ X ] + 2Cov( X , Y ) + Var[Y ]
EX) If Var[X]=15, Var[Y]=15 and Cov(X,Y)= 10,
then Var[X+Y]=10.
MNEMONICS
( x + y ) 2 = x 2 + 2 xy + y 2
(ax + by ) 2 = a 2 x 2 + 2abxy + b 2 y 2
Var[aX ] = a 2Var[ X ]
25
Variance formulas
26
Variance formulas
More generally,
Var[ w1 X 1 + + wn X n ] =
2
w
i Var[ X i ] + 2wi w j Cov( X i , X j ).
i =1,..., n
i< j
All combinations of i
and j such that i < j
With this formula, as long as you know all the variances and
covariances of Xs, you can compute the variance of any linear
combination given ws.
With a little knowledge of matrix algebra, this computation
becomes a lot clearer.
28
1xn
nxn nx1
w1
Var[ w1 X 1 + + wn X n ] = ( w1 wn )V
w
n
1x1
where
w1
w2
w
w
n
= wVw
.
Matrix addition and transpose
http://www.youtube.com/watch?v=U74zP3BsAbM
Matrix multiplication
http://www.youtube.com/watch?v=kuixY2bCc_0
29
matrix of X 1 , X 2 , X 3 is given by V = 46 58 15 .
27 15 24
= (0.2 0.5 0.3) 46 58 15 0.5
27 15 24 0.3
131 0.2 + 46 0.5 + 27 0.3
57.3
30
Exercise 5. There are three random variables X1, X2, and X3.
Their means are given by = (3.5, 4.5, 7.5).
Their standard deviations are given by = (2, 4, 8)
The correlation matrix is given by 1 0.53 0.48 .
0.53 1 0.39
0.48 0.39 1
(1) Find the variance-covariance matrix.
(2) What is the expectation and variance of 0.2X1 + 0.5X2 + 0.3X3? What is the
standard deviation?
31
( 1 , 2 , , n )
11 1n
nn
n1
w ( w1 , w2 , , wn )
subject to
w
i =1
=1
33
Set the minimal acceptable expected return on the portfolio and minimise its
variance:
min wVw
s.t.
w 10%
n
i =1
2.
A number you
can choose.
Set the maximal acceptable variance on the portfolio return and maximise its
expected return:
wVw (30%) 2
max w
w
3.
=1
s.t.
w
i =1
=1
can choose.
34
36
X ++ X n
the same mean and, its std. is
.
X 1
n
SMALL!!
Proof) In our variance formula, all the covariance terms are zero due
to independence. Therefore,
2
X 1 + + X n
Var
.
=
n
EX) The average of 100 dice. (For one die, E[X1]=3.5 and =1.71)
37
Insurance
What is risk of an asset or a project?
1. variance or standard deviation of its return
2. how its return is correlated with your own income
3. the probability of very bad events
38
Aggregate risk/shock
The common risk that affect many people altogether.
Examples are natural disasters, wars, pandemic disease,
economic recessions, financial crises, etc.
Such risks are hard to diversify out and insurance does not work
well.