FIN 412: Topic 9

CAPM

Capital Asset Pricing Model, a linear model that relates risk and return: Re = Rf + Beta(Rm-Rf).

what is the most commonly tested implication of the CAPM?

It should work for every asset j. It should work for the entire cross section of assets.

the independant variable

B

Test procedure

1. set up the sample data
2. estimate the SCL in first-pass regression
3. estimate the SML in second-pass regression

Most tests reject the single factor model. why?

Intercept is wrong, estimated SML is too flat and slope is waaay wrong

what do these tests tell us?

that the model doesn't fit the data very well

Richard Roll's Critique

CAPM uses the market portfolio as the only source of risk premiums
- true market portfolio is unobservable. Tests usually use market portfolio as a proxy.

Is CAPM wrong?

simple CAPM is a single period model
do betas vary over time? probably

what are the models that account for time variation?

ARCH and GARCH

what are some factors other then the market portfolio?

1. Size (SML)
2. Value (HML)
3. Momentum
4. liquidity
5. volatility (eg. VIX)
6. quality
7. low volatility (Betting against B or BAB)

Fama-French Model

They examined:
- beta
no significant related to returns
- firm size
returns inverse
- firm valuation (B/M)
returns vary directly

Fama French 3 Factor Model

Factors SMB and HML

- 0 net investment portfolios
- desingned to capture risk coming from differences in sizes and B/M
- spreads and only capture that type of risk
SMB: return spread of smallest to biggest firms
HML: return spread of highest B/M to lowest B/M

what factors of the fama and french is consistent with APT?

size and B/M ratios
- describe returns

results show that there is an irrational preference for...

large firm size and low b/m

why are there these apparent facors?

-behavioral reasons
- rational, risk-based reasons
- institutional reasons

Equity Premium Puzzle

historical excess returns on risky assets in the US are too large to be consistent with economic theory and reasonable levels of risk aversion.
- risk premium is too high

what could be the reason for equity premium puzzle?

1. unrepresentative sample: perhaps the US is unlike other countries
2. perhaps markets are riskier then we assume, our current models are garbage
3. Perhaps we are not measuring the risk premium appropriately

the best we can get is an appealing theory that has some reasonable empirical support and has these warnings:

- empirical support is sample specific
- estimated parameters are likely to vary over time
- no theory will apply in all seasons, even those with simple requirements

Buffets Alpha

- highest sharpe ratio
- significant alpha to traditional risk factors
- returns can be explained by use of leverage with a focus on safe, cheap and quality stocks

what are other ways to proxy for the market portfolio?

1. Market capitalization indexation (MC)
2. equally-weighted portfolios (EQ)
3. risk efficient indexation (Min Var)

EAFE Index

The EAFE Index stands for Europe, Australasia, and the Far East. It consists of companies of developed countries in these areas - so these are all companies outside of North America. Investing in an EAFE ETF would give the customer international exposure.