5.

Capital Asset Pricing Model

Capital Asset Pricing Model

Applications of CAPM

The Capital Asset Pricing Model

Return models are used to determine expected returns on a security They predict the relationship between the risk of an asset and its expected return
Capital Asset Pricing Model (CAPM) is the most commonly used return model

Key Assumptions of CAPM

Individuals are risk averse, utility maximising, and rational The market is perfectly competitive and investors are price takers All assets are marketable and there are no taxes & transaction costs Individuals plan for a single period horizon Individuals have homogenous expectations (same subjective estimates of means, variances and covariances among returns) Individuals can borrow and lend at risk free rate of return Capital markets are in equilibrium

CAPM Equation
Equation of CAPM E (ri) = rf + i[E(rM) rf] where i = Cov (ri,rM)/Var (rM) = i,M x i/M
i (Beta) represents systematic risk
For the market, beta = 1 Defensive portfolios hold stocks with beta <1 Aggressive portfolios hold stocks with beta>1 Beta of a portfolio p = wi i, weighted average of security betas

CAPM Equation
The CAPM equation can also be written as [E (ri) rf] = i[E(rM) rf]

where [E (ri) rf] is risk premium of the security and [E(rM) rf] is the market risk premium

Note on Systematic Risk

Total risk = Systematic Risk + Unsystematic Risk
Total variance of a security = Variance explained by market + Residual (firm-specific) variance i 2= i2 M2 + 2(ei) If the investor has the choice of holding a well-diversified portfolio, he can diversify the unsystematic risk
hence only the systematic risk component of variance, i2 M2 remains relevant

Security Market Line

The relationship between expected return E(Ri) of a security and its beta is graphically depicted as the Security Market Line (SML)

Security market line relates Expected Return on a security with its Systematic risk (as indicated by beta)
E (ri) = rf + slope x i; slope = (rm-rf) It can be used to identify mispriced securities
If the security lies on the SML, it is correctly priced If the security lies above the SML, it is underpriced If the security lies under the SML, it is overpriced

Security Market Line

SML
C B

Expected Return

E(rM)
A

Slope = E(rM) - rf

rf

1.0 Beta Security Market Line is the graphical representation of the Capital Asset Pricing Model. In the above representation, security A, which lies above the SML has a high level of expected return relative to the risk and is therefore an attractive choice for investment, while security C, which lies below the SML has lower level of expected return relative to risk and could be considered over-valued. Security B may be considered fairly priced.

Capital Market Line & Security Market Line

CML applies to efficient (well-diversified) portfolios The CML is the relationship between required returns of efficient portfolios with their risk, as indicated by p For individual securities and portfolios that are not welldiversified, relevant risk (systematic risk) is lesser than total risk The SML is the relationship between required returns of securities (& portfolios that are not efficient) with their systematic risk, as indicated by i The CML is a special case of SML, because for efficient portfolios total risk is equal to systematic risk.

Security Characteristic Line

Security characteristic line (SCL) relates Excess return on a security with Excess market return
(Ri-Rf) = + slope x (Rm-Rf); slope = i It can be used to estimate excess return earned on a security
If realised returns are equal to required returns, = 0 If realised returns > required returns, > 0 If realised returns < required returns, < 0 (Note: Unlike for CAPM/SML, we have written the equation for SCL ex post facto, hence without E( ) notation)

Security Characteristic Line

Ri - Rf
Excess Security Return

Slope = i

i Rm - Rf

Excess Market Return

i, i can be estimated by regressing (Ri - Rf) against (Rm - Rf) for a security.

CAPM and Investment Decision

Prospectively (ex ante), invest when
The stock is undervalued/underpriced, hence Current price < Fair price (intrinsic value), hence Forecast return (ex ante return) > Required return (based on CAPM), hence Forecast return lies above SML, hence Forecasted is positive

Historically (ex post facto), an investment was a good investment if

Actual return (ex post facto) > Required return (based on CAPM), hence Actual return was higher than SML, hence The of the SCL was positive

Note: To avoid confusion

Think of E(R) that is forecasted as forecast return, and E(R) from the CAPM as required return, i.e. return required to compensate for taking systematic risk. At equilibrium, the forecast and required returns of the market should be equal. However, investors seek mispriced opportunities where forecast return could be higher than required return. Remember that price and return are inversely correlated. Therefore positive excess return indicates underpriced security (good opportunity), while negative excess return indicates overpriced security (not a good opportunity). The terms underpriced and overpriced relate to current market prices (not prices in future) as compared with theoretically correct prices.

Determination of beta
1. Using ex ante forecast returns (subjective returns with probabilities) and the beta formula 2. Using past holding period returns and regression
Based on the regression slope Using coefficient of determination, R-squared

1. Determination of Beta from Forecast Returns

Beta formula
COVi,M r i , M s i bi = = 2 M [9-7] s M

Forecast returns (probability distribution), ex.,

Scenario Optimistic Base Pessimistic Probability 20% 50% 30% Expected Return (Stock) 18% 12% 8% Expected Return (Market) 13% 10% 8%

Determining beta from forecast returns

Expected returns = piri
E(ri) = 12%; E(rM) = 10%

2 = pi[ri-E(ri)]2
2m = 0.0003

Covi,M = pi [ri-E(ri)] [rM-E(rM)]

Covi,M = 0.0006

= 0.0006/0.0003 = 2
Note: where only 2 scenarios are known, beta can also be determined using simultaneous equations of CAPM for both scenarios and subtracting one equation from the other.

2. Determination of Beta by Regression

Regression of realised (historical) security returns with historical market returns ri = intercept + slope x rM + i

Where slope = Alternatively, beta can be determined using coefficient of determination, R2 of regression 2 = R2 x i2 /M2

Beta obtained from historical returns may be adjusted given its tendency to revert towards 1
Ex. Adjusted beta = 2/3 (Historical beta) + 1/3

Applications of CAPM
1. To determine cost of equity (required equity return) of a company
Cost of equity = E(ri), is used in present value based equity valuation methods

2. For stock selection in a portfolio

Using measures based on CML and CAPM

3. To measure portfolio performance

Using measures based on CML and CAPM

1. Application of CAPM in India for cost of equity/ required return: Example

Risk-free rate: 10 year G-Sec rate Market risk premium: Adjusted historical risk premium: 8.75% adjusted to say 7% Beta: Based on regression of monthly returns between stock and index (BSE 30) over 5 years. Historical beta adjusted towards 1 (say in 2/3:1/3 proportion)

Example: Cost of Equity for Tata Steel

Monthly Tata Steel Stock Returns

= 1.61 R = 0.6171 n = 58

50%

40%
30% 20% 10%

0% -15% -10% -5%

-10% -20% -30% -40% Monthly Sensex Returns

-20%

0%

5%

10%

15%

20%

Adjusted beta = 2/3 x 1.61 + 1/3 = 1.41

Example: Cost of Equity for Tata Steel

Rf
Market Risk Premium Adjusted beta Cost of Equity

8.15%
7% 1.41 18.1%

E(ri) = 8.15% + 1.41 x 7% = 18.1% Value of E(ri) using alternative assumptions: Lower value E(ri) = 8% + 1.2 x 7% = 16.4% Higher value E(ri) = 8.4% + 1.6 x 8% = 21.2%

These estimates can be used for building alternative scenarios during valuation

2. Stock selection
Case 1: The investor is analysing many stocks, intending to form a diversified portfolio
Compare securities using Treynor measure = (E(ri)rf)/i

Case 2: The investor intends to choose one stock as the only risky asset to hold
Compare securities using Sharpe ratio = (E(ri)-rf)/i

3. Measurement of portfolio performance

Sharpe ratio, & M2 measure, where the portfolio represents an investors entire risky investment fund Jensens alpha, Treynor measure, where the portfolio is among many combining into the total investment fund Information ratio to measure performance of active management of mutual fund portfolios or hedge funds Note: M2 measure = rp*- rM, where rp* is return of an adjusted portfolio combining portfolio P with risk free asset in such proportions that p* = M Note: Information ratio = (rp-rb)/(p-b), where rb is return on the benchmark index portfolio Note: Jensens alpha and M2 are absolute measures stated as percentages. Sharpe ratio, Treynor measure and information ratio measure excess returns per unit of risk (total risk, systematic risk and active risk respectively) and have no units.

Criticism of CAPM
Restrictive, unrealistic assumptions Does not explain the sources of systematic risk
Empirical tests have challenged its validity

Alternative models
Arbitrage Pricing Theory (APT)
CAPM is based on the dominance argument, implying that all investors hold mean-variance efficient portfolios. If some securities are mispriced, all investors shift their portfolios in favour of underpriced and away from overpriced securities till market equilibrium is restored. APT is based on the no-arbitrage principle that two identical assets cannot sell at different prices and a few investors who identify arbitrage opportunities can restore equilibrium. APT also relies on a proposition that security returns can be described by a model that relates expected returns to multiple risk factors.

Multifactor Models
Macro-economic based risk factor models Firm-specific risk factor models

Reading
BKMM Ch 9, PC Ch 9 Home Work
Use the CAPM to calculate the cost of equity for your stock