Utility and Risk Aversion: (Asset Pricing and Portfolio Theory)
Utility and Risk Aversion: (Asset Pricing and Portfolio Theory)
Indifference curve
Asset Q
ERp
Asset P
sp s
Risk and Return (US Assets
: 1926 – 1998) (% p.a.)
20
18
Small Company Stocks
16
14
12
Large Company Stocks
10
ER Long Term
8
T-bonds
6
Medium Term T-bonds
4
2
Treasury Bills
0
0 5 10 15 20 25 30 35 40 45
Standard deviation
Expected Utility
E[U(W)] = SpiU(Wi)
Example : Alternative
Investments
Investment A Investment B Investment C
Outcome Prob Outcome Prob Oucome Prob
20 3/15 19 1/5 18 ¼
18 5/15 10 2/5 16 ¼
14 4/15 5 2/5 12 ¼
10 2/15 8 ¼
6 1/15
Example : Alternative
Investments (Cont.)
Assume following utility function :
U(W) = 4W – (1/10) W2
– Expected Utility
Investment A : E(UA) = … = 36.3
Investment B : E(UB) = … = 26.98
Investment C :
E(UC) = 39.6(1/4) + 38.4(1/4) + 33.6(1/4) + 25.6(1/4) = 34.3
Utility Function :
U(W) = 4W – (1/10)W2
45
40
35
30
25
U(W)
20
15
10
0
0 5 10 15 20 25
W
Example : Alternative
Investments (Cont.)
Ranking of investments remains
unchanged if
– a constant is added to the utility function
– the utility function is scaled by a constant
Example :
a + bU(W) gives the same ranking as U(W)
Fair Lottery
A fair lottery is defined as one that has expected value of
zero.
Risk aversion applies that an individual would not accept
a ‘fair lottery’.
Concave utility function over wealth
Example :
– tossing a coin with $1 for WIN (heads) and -$1 for LOSS (tails).
x = k1 with probability p
x = k2 with probability 1-p
E(x) = pk1 + (1-p)k2 = 0
k1/k2 = -(1-p)/p or p = -k2/(k1-k2)
– Tossing a coin : p = ½ and k1 = -k2 = $ 1.-
Utility : The Basics
Utility Functions
More is preferred to less :
U’(W) = ∂U(W)/∂W > 0
Example :
– Tossing a (fair) coin (i.e. p = 0.5 for head)
– gamble of receiving £ 16 for a ‘head’ and £ 4 for
‘tails’
– EW = 0.5 (£ 16) + 0.5 (£ 4) = £ 10
– If costs of ‘gamble’ = £ 10.- EW – c = 0
– How much is an individual willing to pay for playing
the game ?
Utility Functions (Cont.)
U(W) = W1/2
U(4) = 2
0
4 EW=10 16 Wealth
(W–p) = 9
Degree of Risk Aversion
Assumptions
– Investor has wealth W and security with outcome
represented by the random variable Z
– Let Z be a fair game E(Z) = 0 and E[Z–E(Z)]2 = sz2
– Investor is indifferent between choice A and B
Choice A Choice B
W+Z Wc
E[(U(W + Z)] = EU(Wc) = U(Wc)
Utility Theory (Cont.)
Define p = W – Wc is the max. investor is willing to pay
to avoid gamble. Measurement of investor’s absolute
risk aversion.
Choice A Choice B
WZ = Wc
U(16)
U(10)
4 10 16 Wealth
Indifference Curves in
Risk – Return Space
Risk Averter
Expected Return
Risk Lover
Risk Neutral
Risk, s
Examples of Utility
Functions
Utility Function : Power
RA(W) = g/W
RR(W) = g (a constant)
Utility Function :
Logarithmic
As g 1, logarithmic utility is a limiting case
of power utility
U(W) = ln(W)
U’(W) = 1/W
U’’(W) = -1/W2
RA(W) = 1/W
RR(W) = 1
Utility Function :
Quadratic
U(W) = W – (b/2)W2 b>0
U’(W) = 1 – bW
U’’ = -b
RA(W) = b/(1-bW)
RR(W) = bW / (1-bW)
RA(W) = c
RR(W) = cW
Empirical Evidence
How does it Work in the
‘Real World’ ?
To investigate whether (specific) utility
functions represent behaviour of
economic agents :
– Experimental evidence from simple choice
situations
– Survey data on investor’s asset choices
Empirical Studies
Blume and Friend (1975)
– Data : Federal Reserve Board survey of financial
characteristics of consumers
– Findings : Percentage invested in risky asset
unchanged for investors with different wealth
Cohn et al (1978)
– Data : Survey data from questionnaires (brokers
and its customers)
– Findings : Investors exhibit decreasing relative
RA and decreasing absolute RA
Coefficient of Relative
Risk Aversion (g)
From experiments on gambles coefficient of
relative risk aversion (g) is expected to be in
the range of 3–10.
S&P500
Average real return (since WW II) 9% p.a. with SD
16% p.a.
C-CAPM suggests coefficient of relative risk aversion
(g) of 50.
Equity Premium Puzzle
Is g = 50 Acceptable ?
Based on the C-CAPM
For g = 50, risk free rate must be 49%
Cochrane (2001) presents a nice example
– Annual earnings $ 50,000
– Annual expenditure on holidays (5%) is $ 2,500
– Rft ≈ (52,500/47,500)50 – 1 = 14,800% p.a.
– Interpretation : Would skip holidays this year
only if the risk free rate is 14,800% !
How Risk Averse are You ?
W = W0(1 + Rp)
U(W) = U[W0(1 + Rp)]
Mean-Variance Model and
Utility Functions (Cont.)
Expanding U(Rp) in Taylor series around mean of Rp (=mp)
U(Rp) = U(mp) + (Rp – mp) U’(mp)
+ (1/2)(Rp – mp)2 U’’(mp)
+ higher order terms
Taking expectations
E[U(Rp)] = U(mp) + (1/2) s2p U’’(mp) +E(higher-terms)