Pwc Pres 080206 Bacon e
-
Upload
abhishek210585 -
Category
Documents
-
view
215 -
download
0
Transcript of Pwc Pres 080206 Bacon e
-
7/28/2019 Pwc Pres 080206 Bacon e
1/64
1
RiskRisk--adjusted Performance Measurementadjusted Performance Measurement
Alpha to OmegaAlpha to OmegaDownside to DrawdownDownside to Drawdown
Appraisal to PainAppraisal to Pain
Carl Bacon
Zurich6th February 2008
-
7/28/2019 Pwc Pres 080206 Bacon e
2/64
2 2
Risk ManagementRisk Management
Risk Managers Front Office
Paid to take risk
Risk is good Risk Controllers
Middle office
Paid to monitor/reduce risk Risk is bad
-
7/28/2019 Pwc Pres 080206 Bacon e
3/64
3 3
Risk MeasuresRisk Measures
Ex-Post
Risk after the event
Historical
Ex-Ante
Risk before the event
Prospective
-
7/28/2019 Pwc Pres 080206 Bacon e
4/64
4 4
Simple Risk MeasuresSimple Risk Measures
Absolute Variability, Sharpe Ratio
Relative Tracking Error, Information Ratio
Regression
,
-
7/28/2019 Pwc Pres 080206 Bacon e
5/64
5 5
VariabilityVariability
mean
r
= [ri-mean]2
n
-
7/28/2019 Pwc Pres 080206 Bacon e
6/64
-
7/28/2019 Pwc Pres 080206 Bacon e
7/64
7 7
6.4
6.6
6.8
7.0
7.2
7.4
7.6
7.8
8.0
A B Benchmark
Portfolio ReturnsPortfolio Returns
7.9%
6.9%
7.5%
-
7/28/2019 Pwc Pres 080206 Bacon e
8/64
8 8
Portfolio RiskPortfolio Risk
0.0
1.0
2.0
3.0
4.0
5.0
6.0
A B Benchmark
5.50%
3.20%
4.50%
-
7/28/2019 Pwc Pres 080206 Bacon e
9/64
9 9
Sharpe RatioSharpe Ratio
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Risk
Retur
n
B ABenchmark
-
7/28/2019 Pwc Pres 080206 Bacon e
10/64
10 10
MM22
0%1%
2%
3%
4%5%
6%
7%
8%9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Risk
Return
B
A
Benchmark
Return adjusted
to risk M2 for B
Return adjusted
to risk M2 for A
-
7/28/2019 Pwc Pres 080206 Bacon e
11/64
11 11
Risk Adjusted ReturnsRisk Adjusted Returns
Sharpe
Ratio P
FP rrSR
=
P = P o r t f o l i o R i s k M = M a r k e t R i s k R P = P o r t f o l i o R e t u r nR F = R i s k F r e e R a t e
or
( )pMP SRrM +=
2
( ) FP
MFP rrrM +=
2
-
7/28/2019 Pwc Pres 080206 Bacon e
12/64
12 12
Modigliani Excess ReturnModigliani Excess Return
4.5%
B
Return adjusted for
Risk M2= 8.9%
M2 (Excess) =8.9% - 7.5% = 1.4%
Risk adjusted excess return
-
7/28/2019 Pwc Pres 080206 Bacon e
13/64
13 13
Tracking errorTracking error
Tracking Error
Standard Deviation of excess return
Measures consistency of excess return
Ex-Post or Ex-ante Normally annualised
Information RatioAnnualised Excess Return
Annualised Tracking Error
-
7/28/2019 Pwc Pres 080206 Bacon e
14/64
14 14
Information RatioInformation Ratio
Excess
Return
Tracking Error
-
7/28/2019 Pwc Pres 080206 Bacon e
15/64
15 15
Information RatioInformation Ratio
0
0.2
0.4
0.6
0.8
1
1.2
1.41.6
Q1
97
Q3
97
Q1
98
Q3
98
Q1
99
Q3
99
Information ratio
-
7/28/2019 Pwc Pres 080206 Bacon e
16/64
16 16
Risk Efficiency RatioRisk Efficiency Ratio
Risk Efficiency Ratio
or
Measures the quality of the tracking error forecast
Observed Tracking ErrorForecast Tracking Error
Ex-Post Tracking ErrorEx-Ante Tracking Error
-
7/28/2019 Pwc Pres 080206 Bacon e
17/64
17 17
Regression EquationRegression EquationR
B
=0.1
xx x
xx x
x
x x
xx
x
x x
x
x =1.0
R = + (B-RF) +
-
7/28/2019 Pwc Pres 080206 Bacon e
18/64
18 18
Bull & Bear BetasBull & Bear BetasR
B
xx x
xx
xx
x x
xx
x
xx
xx+ =1.1
=0.9
-
7/28/2019 Pwc Pres 080206 Bacon e
19/64
19 19
Regression EquationRegression Equation
Regression Equation
Beta Gradient of the regression equation
Jensens Alpha Intercept of the Regression Equation
( )
( )
++= )( FMFp rrrr
-
7/28/2019 Pwc Pres 080206 Bacon e
20/64
20 20
Regression StatisticsRegression Statistics
Correlation ( r or ) =
Coefficient of =
Determination (R2)
Systematic RiskPortfolio Risk
Systematic VariancePortfolio Variance
-
7/28/2019 Pwc Pres 080206 Bacon e
21/64
21 21
Treynor RatioTreynor Ratio
0%
1%2%
3%
4%
5%
6%
7%
8%
9%
10%
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Beta
Return
B
A Benchmark
-
7/28/2019 Pwc Pres 080206 Bacon e
22/64
22 22
Risk MeasuresRisk Measures
Treynor Ratio
Systematic Risk
Appraisal Ratio
ms =
FPrr
TR
=
Modified Jensen
=
RiskSpecific
AlphasJensen'
-
7/28/2019 Pwc Pres 080206 Bacon e
23/64
23 23
FamaFama DecompositionDecomposition
Early form of performance attribution Return above risk free rate decomposed into two
components:
Systematic Risk Investors risk
Managers risk
Selectivity ( Jensens alpha) Diversification
Net selectivity
-
7/28/2019 Pwc Pres 080206 Bacon e
24/64
24 24
FamaFama DecompositionDecomposition
Return
Total
Excess
Return
Return
from
Selectivity
Return from
Managers Risk
Return from
Investors risk
Diversification
Net Selectivity
A1
Benchmark
A3
A2
I P F
-
7/28/2019 Pwc Pres 080206 Bacon e
25/64
25 25
FamaFama Decomposition: Systematic riskDecomposition: Systematic risk
Return due to systematic risk
Return due to investors systematic risk
Return due to managers systematic risk
( )FMPp rrr =
( )FMII rrr =
( ) ( )FMIPM rrr =
-
7/28/2019 Pwc Pres 080206 Bacon e
26/64
26 26
FamaFama DecompositionDecomposition
Fama Equivalent F
Diversification (d) (Return required to justified specific risk taken)
Net Selectivity (s)
M
PF
=
ds =
)()(FMP
F rrd =
-
7/28/2019 Pwc Pres 080206 Bacon e
27/64
27 27
GH 1 ( Graham & Harvey)GH 1 ( Graham & Harvey)
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Risk
Retur
C
D
Benchmark
GH1 C
GH1 D
-
7/28/2019 Pwc Pres 080206 Bacon e
28/64
28 28
GH2 ( Graham & Harvey) for MGH2 ( Graham & Harvey) for M22
0%
1%2%
3%
4%
5%
6%
7%
8%9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Risk
Retu
r
C
D
Benchmark
GH2 C
GH2 D
-
7/28/2019 Pwc Pres 080206 Bacon e
29/64
29 29
Hedge FundsHedge Funds
A hedge fund constitutes an investment programwhereby the managers or partners seek absolute
returns by exploiting investment opportunities
while protecting principal from potential financialloss
Ineichen (2003)
-
7/28/2019 Pwc Pres 080206 Bacon e
30/64
30 30
RiskRisk
--adjusted Performance for Hedge Fundsadjusted Performance for Hedge Funds
Investment returns are not normal
Particularly hedge funds
Skewness
Kurtosis
Excess Kurtosis
n
rrS
P
i 13
=
=
n
rrK
p
i 14
31
4
=
n
rrK
p
iE
-
7/28/2019 Pwc Pres 080206 Bacon e
31/64
31 31
KurtosisKurtosis
Kurtosis >3
Kurtosis
-
7/28/2019 Pwc Pres 080206 Bacon e
32/64
32 32
SkewnessSkewness
Positive Skew Negative Skew
-
7/28/2019 Pwc Pres 080206 Bacon e
33/64
33 33
BeraBera--Jarque TestJarque Test
Test for normality
Reject if exceeds 5.99 (95%)
Reject if exceeds 9.21 (99%)
+=
46
22 EKS
nBJ
-
7/28/2019 Pwc Pres 080206 Bacon e
34/64
34 34
Adjusted Sharpe RatioAdjusted Sharpe Ratio
+=
2
24
3
6
1 SRK
SRS
SRASR
-
7/28/2019 Pwc Pres 080206 Bacon e
35/64
35 35
Downside RiskDownside Risk
D=Downside semi-standard deviation
Fewer data points
Less information Does upside dispersion matter?
down
TP rr
=RatioSortino
Tr =Mimimum acceptable (or Target) return
[ ]
=
=
n
i
Ti
D
n
rr
1
20),(min
-
7/28/2019 Pwc Pres 080206 Bacon e
36/64
36 36
Sortino RatioSortino Ratio
0%
1%
2%
3%
4%5%
6%
7%
8%
9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Downside Risk
Retur
B
A
Benchmark
-
7/28/2019 Pwc Pres 080206 Bacon e
37/64
37 37
MM
22 (for downside risk)
0%
1%
2%
3%
4%5%
6%
7%
8%
9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
Downside Risk
Retur
B
A
Benchmark
M2 for downside risk
-
7/28/2019 Pwc Pres 080206 Bacon e
38/64
38 38
Upside Potential Ratio (Discrete)Upside Potential Ratio (Discrete)
UPR= Upside Potential
Downside Risk
( )
D
ni
iTi nrr
UPR
=
= =
10,max
-
7/28/2019 Pwc Pres 080206 Bacon e
39/64
39 39
OmegaOmega
Gain-Loss Ratio ( Bernardo & Ledoit )
Omega-Sharpe Ratio
1=
=
=
=
ni
iiT
TP
rrn
rr
1
)0,max(1
0=Tr
PotentialDownside
PotentialUpside
)0,max(
1
)0,max(1
1 =
=
=
=
iT
ni
i
Ti
rrn
rrn
-
7/28/2019 Pwc Pres 080206 Bacon e
40/64
40 40
Prospect RatioProspect Ratio
Penalises loss greater than gain Based on Prospect Theory
( ) ( )( )
D
T
ni
i
ii rrMinrMaxn
+=
=
=10,25.20,
1
-
7/28/2019 Pwc Pres 080206 Bacon e
41/64
41 41
Value at Risk (VaR)Value at Risk (VaR)
Maximum loss in cash terms over a finite period
given a certain level of confidence (say 95%)
Typically ex-ante
Volatility is an input - not an output Similar to ex-ante tracking error
-
7/28/2019 Pwc Pres 080206 Bacon e
42/64
42 42
VaRVaR
Historical Simulation Re-organises actual historical returns
Monte Carlo Simulation
Multiple hypothetical trials Variance- Co-variance
Assumes normality
-
7/28/2019 Pwc Pres 080206 Bacon e
43/64
43 43
Return on VaRReturn on VaR
Return on VaR
(%)VaR
rrR FPVaR
=
-
7/28/2019 Pwc Pres 080206 Bacon e
44/64
44 44
Return on VaRReturn on VaR
0%
1%
2%
3%
4%
5%
6%
7%
8%9%
10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0%
VaR
Retur
BA
Benchmark
-
7/28/2019 Pwc Pres 080206 Bacon e
45/64
45 45
Conditional VaR (Conditional VaR (E x p e c t e d S h o r t f a l l E x p e c t e d S h o r t f a l l ))
Conditional
VaR
-
7/28/2019 Pwc Pres 080206 Bacon e
46/64
46 46
Conditional Sharpe RatioConditional Sharpe Ratio
Conditional Sharpe RatioCVaR
FP rr =
-
7/28/2019 Pwc Pres 080206 Bacon e
47/64
47 47
Modified VaModified VaRR
VaR modified for Kurtosis & Skewness
Cornish-Fisher expansion
+
++=
2332
36
52
24
3
6
1MVa S
zzK
zzS
zzrR
ccE
ccccp
=cz -1.96 with 95% confidence=cz -2.33 with 99% confidence
-
7/28/2019 Pwc Pres 080206 Bacon e
48/64
48 48
Modified Sharpe RatioModified Sharpe Ratio
Modified Sharpe RatioMVaR
FP rr =
-
7/28/2019 Pwc Pres 080206 Bacon e
49/64
49 49
DrawdownDrawdown
MaximumDrawdown
Time
D1
D2
D3
Recovery time
Cumulativereturn
St li C l & B k R tiSt li C l & B k R ti
-
7/28/2019 Pwc Pres 080206 Bacon e
50/64
50 50
Sterling, Calmar & Burke RatiosSterling, Calmar & Burke Ratios
Sterling Ratio
Calmar Ratio
Burke ratio
Drawdown = losing period
Max
D
F
r
rr=
=2
Dj
F
r
rr
=D
R
=DjR
=Max
DR
Average Drawdown
Maximum Drawdown
Drawdown j
D
F
r
rr=
-
7/28/2019 Pwc Pres 080206 Bacon e
51/64
51 51
Versions of the Sterling RatioVersions of the Sterling Ratio
Multiple Variations
Original (Deanne Sterling Jones)
Average or Largest Individual Drawdown (say 3
or 5)
Average annual Maximum Drawdown
%10+
D
p
r
r
-
7/28/2019 Pwc Pres 080206 Bacon e
52/64
52 52
SterlingSterling--Calmar RatioCalmar Ratio
Max
D
Fp
r
rr Sterling-Calmar Ratio =
Combines concepts in Calmar & Sterling Ratios
-
7/28/2019 Pwc Pres 080206 Bacon e
53/64
53 53
Pain & Ulcer IndexesPain & Ulcer Indexes
Reflect pain & worry since high water mark
Pain Index
Ulcer Index =
=
!=
ni
i
i
n
DUI
1
2
=
=
!=
ni
i
i
n
DPI
1
iperiodinpeakpervioussincedrawdown=!i
D
-
7/28/2019 Pwc Pres 080206 Bacon e
54/64
54 54
Pain & Martin RatiosPain & Martin Ratios
Sharpe type ratios
Pain Ratio
Martin Ratio(or Ulcer Performance Ratio)
=
=
!
=
ni
i
i
FP
n
D
rrMR
1
2
=
=
!
=mi
i
i
FP
n
D
rrPR
1
-
7/28/2019 Pwc Pres 080206 Bacon e
55/64
55 55
DurationDuration
Duration is a Fixed Income Risk Measure Macaulay Duration
Macaulay-Weil Duration
Modified Duration
Effective Duration
It measures price sensitivity to changes in interest
rates
B
P
DDDuration =
-
7/28/2019 Pwc Pres 080206 Bacon e
56/64
56 56
Hurst IndexHurst Index
Where:
Between 0.5 and 1 - Persistent
Around 0.5 - Totally Random
Between 0 and 0.5 - Anti-Persistent
( )
( )n
mH
log
log=
( ) ( )
P
ii PPm
minmax =
-
7/28/2019 Pwc Pres 080206 Bacon e
57/64
57 57
Risk Adjusted Attribution
( )FMF RRRR +=
Ri k Adj t d Att ib ti
-
7/28/2019 Pwc Pres 080206 Bacon e
58/64
58 58
Risk Adjusted Attribution
Portfolio Benchmark Portfolio Index Weight Weight Return Return
UK 40% 40% 20.0 10.0 1.3
Japan 30% 20% -5.0 -4.0 1.0
US 30% 40% 6.0 8.0 0.8
Total 100% 100% 8.3 6.4 1.1
Total Excess Return =1.79
Risk Adjusted Attribution (Asset Allocation)
-
7/28/2019 Pwc Pres 080206 Bacon e
59/64
59 59
Risk Adjusted Attribution (Asset Allocation)
Portfolio Benchmark Portfolio Index Semi-
Weight Weight Return Return Notional
UK 40% 40% 20.0 10.0 10.0
Japan 30% 20% -5.0 -4.0 -4.0US 30% 40% 6.0 8.0 8.0
Total 100% 100% 8.3 6.4 5.2
Total Excess Return -1= 1.79
Asset (or country) Allocation
1.083
1.064
[ ] 01064.1
10.1%40%40 =
[ ] 97.01064.1
96.0%20%30 =
[ ] 15.01064.1
08.1%40%30 =
13.11064.1
052.1=
UK
JAPAN
US
TOTAL 0 - 0.97 - 0.15 = -1.13 or alternatively
S t ti Ri kS t ti Ri k
-
7/28/2019 Pwc Pres 080206 Bacon e
60/64
60 60
Portfolio Benchmark Portfolio Index Risk-
Weight Weight Return Return adjusted
UK 40% 40% 20.0 10.0 12.7
Japan 30% 20% -5.0 -4.0 -4.0US 30% 40% 6.0 8.0 6.4
Total 100% 100% 8.3 6.4 5.8
Systematic Risk
03.1052.1
10.1110.1
127.1%]40[ =
0.0052.1
96.01
96.0
96.0%]30[ =
44.0052.1
08.11
08.1
0644.1%]30[ =
58.01052.1
0581.1=
UK
JAPAN
US
TOTAL 1.03 +0.0 - 0.44 = 0.58
Systematic RiskSystematic Risk
or alternatively
S l ti itS l ti it
-
7/28/2019 Pwc Pres 080206 Bacon e
61/64
61 61
Portfolio Benchmark Portfolio Index Risk-
Weight Weight Return Return adjusted
UK 40% 40% 20.0 10.0 12.7
Japan 30% 20% -5.0 -4.0 -4.0US 30% 40% 6.0 8.0 6.4
Total 100% 100% 8.3 6.4 5.8
76.20581.1
127.11127.1
20.1%]40[ =
28.00581.1
96.01
96.0
95.0%]30[ =
12.00581.1
0644.11
0644.1
06.1%]30[ =
35.210581.1
083.1=
UK
JAPAN
US
TOTAL 2.76 - 0.28 - 0.12 = 2.35
SelectivitySelectivity (or Adjusted Stock Selection)(or Adjusted Stock Selection)
or alternatively
C l di R kC l di R k
-
7/28/2019 Pwc Pres 080206 Bacon e
62/64
62 62
Concluding RemarksConcluding Remarks
Risk like beauty is in the eye of the
beholder The basic structure of all composite risk measures
is the same:
Get it wrong and youre in trouble
RiskReward
U ilUnile er M illMerrill
-
7/28/2019 Pwc Pres 080206 Bacon e
63/64
63 63
UnileverUnilever -- MerrillMerrill
Financial Times
1st October 2001
AppendixAppendix Excel FunctionsExcel Functions
-
7/28/2019 Pwc Pres 080206 Bacon e
64/64
64 64
AppendixAppendix -- Excel FunctionsExcel Functions
Sample excess Kurtosis
Sample Skewness
( )
( ) ( ) ( )
( )
( ) ( )32
13
321
12
4
+
=
nn
n
nnn
nnrr
p
i
( ) ( )21
3
= nn nrr P
i