Investment Analysis and Portfolio Management: Ch 9

You are given the following six assumptions:
(1) Capital markets are perfectly competitive.
(2) Quadratic utility function.
(3) Investors prefer more wealth to less wealth with certainty.
(4) Normally distributed security returns.
(5) Representation as a K factor model.
(6) A market portfolio that is mean-variance efficient.
Which ones of these assumptions are used in the APT? The others are used in what pricing model?
1, 3, and 5
CAPM
Discuss the main similarities and differences between the CAPM and the APT.
– See previous question for different assumptions
– CAPM has one pre-determined risk factor (the market excess return), the APT has an unknown number of risk factors (none of which are pre-determined)
– Both are difficult to test:
CAPM requires a market proxy
APT requires identification of factors
– Both are linear models that assume that unique risks are diversified away
Discuss the Roll critique of the CAPM. (slide 9.3)
The Roll critique of the CAPM is that the only testable implication from the CAPM is that the market is efficient. However, the market portfolio is not observable.
Discuss the Shanken critique of the APT. (slide 9.15)
The APT does NOT identify the factors. Thus, if a set of assets is not priced with a given set of factors, it is not evidence against the APT, but if the factors do price the assets it is seen as support for the APT.
A portfolio that meets the following 3 conditions in an arbitrage portfolio:
The initial cost of the portfolio = 0
No risk
Positive Profit