Etbics and Professional Standards

CFA ProfessionaI Conduct Program

1he rules of Lhe program are wrlLLen ln Lhe ClA lnsLlLuLe 8ylaws and Lhe 8ules of roceedlng 8elaLed Lo
rofesslonal ConducL 1he program ls based on Lhe ldea of falrness and confldenLlallLy 1he ulsclpllnary
8evlew CommlLLee of Lhe ClA 8oard of Covernors ls responslble for Lhe program 1he ClA lnsLlLuLe
ueslgnaLed Cfflcer Lhrough Lhe rofesslonal ConducL sLaff conducLs Lhe lnqulrles lf you do someLhlng
LhaL's lllegal or uneLhlcal and a rofesslonal ConducL sLaff member of Lhe ClA lnsLlLuLe flnds ouL Lhen
Lhey may mall you a leLLer asklng an explanaLlon lnLervlew you lnLervlew oLhers relaLed Lo Lhe lncldenL
collecL documenLs and records relevanL Lo Lhe lncldenL 1he ueslgnaLed Cfflcer would Lhen declde Lo 1)
Lake no dlsclpllnary acLlon 2) wrlLe you a sLrongly worded leLLer or 3) dlsclpllne you 1he dlsclpllnary
acLlon may be condemnaLlon" by your peers or suspenslon of your parLlclpaLlon ln Lhe ClA program lf
Lhey choose Lo dlsclpllne you Lhen you may elLher accepL or re[ecL Lhelr accusaLlon lf you re[ecL lL Lhen
lL wlll be referred Lo a panel of ClA members
Code of Ethics
Act with integrity, competence, diligence, respect,
and in an ethical manner with the public, clients,
prospective clients, employers, employees,
colleagues in the investment profession, and other
participants in the global capital markets
Be a good person to everyone you work with
Place the integrity of the investment profession and
the interests of the clients above their own
personal interests
Give the investment profession a good name and
place your clients interests over your own
Use reasonable care and exercise independent
professional judgment when conducting investment
analysis, making investment recommendations,
taking investment actions, and engaging in other
professional activities
Be prudent in investment analysis,
recommendation and action
Practice and encourage others to practice in a
professional and ethical manner that will reflect
credit on themselves and the profession
Be professional and ethical and influence others to
be as well
Promote the integrity of, and uphold the rules
governing, capital markets
Promote the integrity of capital markets
Maintain and improve their professional
competence and strive to maintain and improve the
competence of other investment professionals
Be competent and influence others to be as well
Standards of ProfessionaI Conduct

Professionalism, Integrity of Capital Markets,
Duties to Clients, Duties to Employers, Investment
Analysis, Recommendations, and Actions,
Conflicts of nterest, Responsibilities as a CFA
nstitute Member or CFA Candidate
Plan on Interesting Destinations, Develop
Interests, Cultivate, Relationships

Professionalism, Knowledge of the Law,
Independence and Objectivity, Misrepresentation,
Misconduct
Pre-eminent Knowledge of Investing; More Money
Integrity of Capital Markets, Material Non-Public
Manipulation, Market Manipulation
Interested in More Money?
Duties to Clients, Loyalty, Prudence & Care, Fair
Dealing, Suitability, Performance Presentation,
Preservation of Confidentiality
Don't Lose Faith, Save Plenty of Patience
Duties to Employers, Loyalty, Additional
Compensation, Responsibilities of Supervisors
Do Like Answering Right
Investment Analysis, Recommendations, and
Actions, Diligence and Reasonable Basis,
Communication with Clients and Prospective
Clients, Record Retention
Interest Does Create Retention
Conflicts of nterest, Disclosure of Conflicts,
Priority of Transactions, Referral Fees
Create a Deeply Personal Recalls
Responsibilities as a CFA nstitute Member or
Candidate, Conduct as Members and Candidates
in the CFA Program, Reference to CFA nstitute,
the CFA Designation, and the CFA Program
Ross Cheairs Rules

lobal Investment Performance Standards {IPS]
rite out the standards

aantitative Metbods
Time Valae of Money
Discoanted Casb Flow Applications
O 8ank ulscounL ?leld (8u?) ulscounL / lace value * 360 / 1lme ln uays u/l*360/1
O oldlng erlod ?leld (?) Lndlng value / 8eglnnlng value
O LffecLlve Annual ?leld (LA?) (1 + ?)363/1
O ,oney ,arkeL ?leld (,,?) aka Cu LqulvalenL ?leld ? x (360/L)
O 1o geL from Lhe 8u? Lo Lhe ,,? converL Lhe 8u? Lo ? and Lhen ? Lo ,,?
4 1o converL Lhe 8u? Lo ? Lake Llme ouL of Lhe equaLlon by mulLlplylng 8u? Llmes 1/360
Lo geL Lhe dlscounL over Lhe face value 1o go from Lhe dlscounL Lo Lhe beglnnlng
value you mulLlply Lhe endlng value by Lhe dlscounL 1o geL Lhe ? you dlvlde Lhe
endlng value by Lhe beglnnlng value
4 1o geL Lhe ,,? you mulLlply Lhe ? by 360/1
O 8ond LqulvalenL ?leld (8L?) equals Lwlce Lhe semlannual raLe 8ond coupon paymenLs are
LradlLlonally pald Lwlce a year so descrlblng Lhelr value as Lwo Llmes Lhe coupon raLe has
become Lhe convenLlon
O 1o geL from ? Lo 8L? Lake Lhe ? and ralse lL Lo Lhe number of perlods you wanL dlvlded by
Lhe number of perlods Lhe ? represenLs (Lx WhaL ls bond equlvalenL yleld of a 3 monLh loan
wlLh a ? of 2? We wanL Lhe 6 monLh equlvalenL of Lhe bond's 3 monLh yleld so we ralse Lhe
2 by (2 perlods we wanL / Lhe 1 perlod Lhe bond represenLs) or 102(2/1)10404 and Lhen we
mulLlply Lhls semlannual yleld by 2 Lo geL Lhe bond equlvalenL yleld or 404 * 2 808
O 1o go from LA? Lo 8L8 you ralse Lhe LA? by (1/2) and Lhen double lL
O 8ond LqulvalenL ?leld ls also deflned as Lhe oldlng erlod 8eLurn (8) mulLlplled by Lhe
number of days ln a year dlvlded by Lhe days Lo maLurlLy whlch ls essenLlally Lhe same as Lhe
oLher deflnlLlon 8L? 8 * (363/uays Lo ,aLurlLy) (Lx A 3 mo bond has a 2 reLurn 1o geL
Lhe 8L? we mulLlply Lhe 2 reLurn by 363 days dlvlded by Lhe days Lo maLurlLy or 102 x (363 /
(3 x 363 / 12)) 102 x 4 108 whlch ls abouL equlvalenL Lo Lhe 808 we calculaLed Lhrough
Lhe alLernaLlve deflnlLlon
Statistical Concepts and Market Retarns
O uescrlpLlve sLaLlsLlcs uLlllze populaLlons whlle lnferenLlal sLaLlsLlcs uLlllze a sample as a means of
maklng lnferences abouL Lhe populaLlon lL represenLs
O 1here are dlfferenL meLhods of measurlng lnvesLmenLs uslng a nomlnal scale means LhaL we
aren'L really maklng any quanLlLaLlve assessmenL of Lhe lLems we're measurlng buL are [usL
asslgnlng Lhem values (eg bonds 1 sLocks 2 real esLaLe 3 eLc) uslng an ordlnal scale
means we are maklng one quanLlLaLlve valuaLlon of Lhe seL such as asslgnlng a group of 100
sLocks lnLo 1 of 10 caLegorles 1 belng besL and 10 belng worsL based upon ?1u reLurn We can
say LhaL Lhose wlLh a ranklng of 1 dld quanLlLaLlvely beLLer Lhan Lhose wlLh a ranklng of 10 8uL
we cannoL say by how much a sLock ln group dld beLLer Lhan one ln group 10 1emperaLure ls an
lnLerval measure however because 0 degrees does noL communlcaLe a lack of heaL we can'L say
LhaL 30 degrees ls Lhree Llmes 10 degrees lf 0 degrees represenLed absoluLe zero Lhen we
would call Lhls measuremenL scale a raLlo scale and ln LhaL case 10 degrees would be 1/3
rd
as
warm as 30 degrees ,oney ls a raLlo scale lf you have zero dollars Lhen you have no money lf
you have 1 dollar Lhen you have half as much as havlng 2 dollars 2 dollars ls more Lhan 1 dollar
?ou can order make lnLervals and use raLlos ln a meanlngful way wlLh a raLlo scale
O A parameLer ls a characLerlsLlc of a populaLlon whereas a sample sLaLlsLlc ls a characLerlsLlc of a
sample
O 8elaLlve frequency ls Lhe frequency of Lhe lnLerval dlvlded by Lhe LoLal number of observaLlons
CumulaLlve frequency and cumulaLlve relaLlve frequency are as Lhey sound
O armonlc ,ean n/(1/x
1
+1/x
2
1/x
n
)
O armonlcCeomeLrlc,ean unless Lhe observaLlon values have no varlaLlon
O ;uanLlles and measures of cenLral Lendency are known as measures of locaLlon
O When calculaLlng quarLlles remember you know how Lo calculaLe percenLlles lf you have 11
observaLlons and wanL Lhe second quarLlle Lhen you mulLlply 11 + 1 by 030 and you'll flnd LhaL
6 ls Lhe second quarLlle or 30
Lh
percenLlle meanlng half of all observaLlons fall below LhaL
number lf lL doesn'L make sense Lo you Lhen wrlLe ouL a number llne up Lo 3 and counL Lhe
observaLlons above and below Lhe calculaLed number and you'll flnd Lhey are Lhe same
O 1he range of any daLa seL ls Lhe largesL value mlnus Lhe smallesL
O 1he mean absoluLe devlaLlon ls Lhe sum of Lhe absoluLe value of Lhe mean subLracLed from each
value dlvlded by Lhe number of observaLlons or ,Au |x
l
meanofx|/n
O opulaLlon varlance ls (x
l
)
2
/n
O Landard devlaLlon ls square rooL of varlance
O Chebyschev's Inequa||ty ls useful because lL Lells you Lhere aL leasL x of observaLlons LhaL lle
wlLhln a cerLaln number of devlaLlons from Lhe mean regardless of lf we're Lalklng abouL a
populaLlon or sample and regardless of Lhe shape of Lhe dlsLrlbuLlon 1he formula ls 1 1/k
2

when k1 ,eanlng lf l wanL Lo know how many observaLlons lle below four sLandard devlaLlons
of Lhe mean Lhen 1 1/4
2
1 1/16 093730 9373
O lf l wanL Lo know Lhe relaLlve dlsperslon of a seL of values Lhen l can dlvlde Lhe sLandard
devlaLlon of Lhe values by Lhe average value whlch ls also known as Lhe coeff|c|ent of var|at|on
(CV) lL can be LhoughL of as varlaLlon per unlL of reLurn
O 1he harpe kat|o ls lmporLanL or aL leasL ofLen referred Lo lL's a measure of Lhe raLlo of excess
reLurn of a porLfollo over Lhe varlablllLy of Lhe porLfollo or (orLfollo 8eLurn 8lskfree 8eLurn) /
Ld uev of Lhe orLfollo ?ou would be sharp Lo remember LhaL lnvesLors llke Lo know how much
excess reLurn for how much rlsk
O ln negat|ve|y skewed d|str|but|ons you have Lhe ouLllers on Lhe lefL pulllng Lhe mean medlan
and mode wlLh lL whlch sLreLches ouL Lhe lefL of Lhe dlsLrlbuLlon curve ln a pos|t|ve|y skewed
d|str|but|on you have Lhe ouLllers on Lhe rlghL pulllng Lhe mean medlan and mode Lo Lhe rlghL
1he mean ls always pulled flrsL Lhen ls Lhe medlan Lhen Lhe mode 1hls has Lo do wlLh Lhe exLenL
Lo whlch Lhese measuremenLs welghL Lhe value of Lhe observaLlons ,ode does noL aL all and so
lL ls leasL senslLlve Lo ouLllers
O urLosls measures Lhe degree Lo whlch a dlsLrlbuLlon ls more or less peaked Lhan a normal
dlsLrlbuLlon LepLokurLlc refers Lo a dlsLrlbuLlon LhaL ls more peaked Lhan normal laLykurLlc
refers Lo one LhaL ls less peaked A dlsLrlbuLlon LhaL shares Lhe same kurLosls as a normal
dlsLrlbuLlon ls mesokurLlc ?ou can remember Lhls because Lhe kurLlc LhaL leaps ls golng Lo be
Lhe hlghesL (lepLokurLlc) whlle Lhe one LhaL looks llke a plaLypus ls golng Lo be Lhe lowesL
(plaLykurLlc) and Lhe one LhaL looks sLandard ls meh (mesokurLlc)
O WheLher a dlsLrlbuLlon ls plaLy or lepLo lL exhlblLs excess kurLosls A normal dlsLrlbuLlon has a
kurLosls of 3 or zero excess kurLosls A lepLokurLlc dlsLrlbuLlon has poslLlve excess kurLosls and a
plaLykurLlc dlsLrlbuLlon has negaLlve excess kurLosls
O ln order Lo undersLand Lhe exLenL Lo whlch a seL of daLa ls skewed ln one dlrecLlon or anoLher
you can Lake Lhe sum of Lhe cubed devlaLlons from Lhe mean dlvlded by Lhe cubed sLandard
devlaLlon and by Lhe number of observaLlons or
k
1/n * (x
l
meanofx)
3
/s
3
1hls ls called
samp|e skewness
O lf you wanL Lo calculaLe kurLosls Lhen Lake Lhe sample skewness calculaLlon and lnsLead of
ralslng Lhe sum of Lhe devlaLlons from Lhe mean and Lhe sLandard devlaLlon by 3 you ralse lL by
4
O When conslderlng wheLher lL's approprlaLe Lo use Lhe arlLhmeLlc mean or Lhe geomeLrlc mean
ask yourself wheLher compoundlng lf a facLor ln Lhe case of mulLlyear reLurns compoundlng
probably ls a facLor and Lhe geomeLrlc mean ls approprlaLe ?eL when you conslder Lhe
slLuaLlon where you have an lnvesLmenL wlLh a 30 chance of a 100 reLurn and a 30 percenL
chance of a 30 reLurn Lhe arlLhmeLlc mean ls 23 and Lhe geomeLrlc mean ls 0 ?eL lf l was
Lo sLarL ouL wlLh $1k and wanL Lo know whaL you wlll llkely make off of Lhls over lnvesLmenL
over Lhe course of 2 years you would do Lhe probablllLy Lree and see LhaL you have a 23 of
maklng $4k 30 chance of maklng $1k and 23 chance of maklng $230 1he expecLed reLurn ls
$4k*23+$1k*30+$230*23$136230 whlch ls a 23 reLurn compounded over 2 years 1he
aplan book says LhaL Lhe CeomeLrlc mean ls useful for forecasLlng fuLure compound reLurns
over mulLlple perlods ow does LhaL guldance apply Lo Lhls slLuaLlon?
O robablllLles are deflned by Lhe facL LhaL Lhe acLual probablllLy of any evenL ls beLween 1 and 0
and for an evenL wlLh a seL of muLually excluslve and exhausLlve ouLcomes Lhe sum of Lhe
probablllLles equals 1
O 1here are ob[ecLlve probablllLles and sub[ecLlve probablllLles Cb[ecLlve probablllLles are elLher
emplrlcal probablllLles or LheoreLlcal (a prlorl) probablllLles An emplrlcal probablllLy ls one
whlch ls based upon pasL observaLlons A LheoreLlcal one ls based upon reasonlng lf you were
Lo fllp a coln and calculaLe Lhe probablllLy of heads based upon Lhe observed resulLs Lhen LhaL
would be uslng an emplrlcal approach lf you were Lo say LhaL Lhe coln has Lwo sldes each of
whlch ls equally llkely as Lhe oLher Lherefore Lhere ls a 30 chance of lL landlng on one or Lhe
oLher Lhen LhaL would belng uslng a LheoreLlcal or a prlorl approach
O lf you wanL Lo know Lhe probablllLy of Lwo evenLs boLh happenlng and Lhey are muLually
excluslve Lhen you would mulLlply Lhe probablllLy of Lhe one evenL (A) by Lhe probablllLy of
Lhe oLher evenL (8) lf one ls dependenL on Lhe oLher Lhen you would calculaLe Lhe probablllLy
of evenL A glven Lhe evenL 8 happened by Lhe probablllLy of evenL 8 happenlng l Lhlnk Lhere ls a
023 chance LhaL l'll sllp lf lL's ralnlng ouL and Lhere ls a 13 chance lL wlll raln Lomorrow
1herefore Lhe chance LhaL l'll sllp ls 023 * 13 or 0038 1hls ls Lhe mu|t|p||cat|on ru|e of
probab|||ty
O lf l wanLed Lo deLermlne Lhe probablllLy LhaL one of Lwo Lhlngs happens or Lhey boLh happen
Lhen l add Lhe probablllLy of one evenL Lo Lhe probablllLy of anoLher and subLracL Lhe probablllLy
of boLh of Lhem happenlng lor example lL may raln Lomorrow (13 and lL may be wlndy
Lomorrow (10) and lL may boLh be wlndy and ralny Lomorrow (2) Lherefore Lhe chance lL wlll
elLher raln or be wlndy ls 13 + 2 1 or 16 1hls ls Lermed Lhe add|t|on ru|e of probab|||ty
O lf l wanL Lo know Lhe probablllLy of an evenL A Lhen l can mulLlply Lhe probablllLy of evenL A
glven evenL 8
1
Llmes Lhe probablllLy of evenL 8
1
and add LhaL Lo Lhe (A|8
2
)*(8
2
) and so on for
each of Lhe muLually excluslve and exhausLlve ouLcomes 1hls ls called Lhe tota| probab|||ty ru|e
8emember Lhe reLurns glven economlc slLuaLlon example
O uo some problems where you calculaLe Lhe mean varlance and sLd dev of a seL of probablllsLlc
reLurns
O lf you wanL Lo know how Lwo asseLs move LogeLher Lhen lf you were Lo Lake a sample of Lhe
dally reLurns of each asseL over Lhe pasL 363 days and calculaLe Lhe mean reLurn of each
lndlvldual asseL and subLracL each asseL's mean from each asseL's one days reLurn and sum up
all Lhe values Lo come Lo Lhe covar|ance CCv(A8) ((A
l
meanofA)* (8
l
meanof8))
O lf you wanL Lo puL Lhls covarlance ln Lerms of per unlL of volaLlllLy Lhen by dlvldlng Lhe
covarlance by Lhe producL of Lhe sLandard devlaLlons of each varlable you come Lo Lhe
correlaLlon Corr(8
l
8
[
)Cov(8l 8
[
)/(o(8
l
) *o(8
l
))
O 1hls means LhaL you can also deflne Lhe Cov(8l 8
[
) as Corr(8
l
8
[
)*(o(8
l
) *o(8
l
))