November 29, 2012 at 10:05 am #55902
If a positive or negative alpha exists for the shares of the company of the financial manager, and the market is at least semi-strong form efficient, the alpha would be expected to move to zero as the company’s shares price changes due to arbitrage profit taking . For example in theory a company with a positive alpha would expect relatively high demand for its shares, increasing share price and thereby decreasing return until the alpha is zero.
(b) (i) CAPM tends to overstate the required return of high beta securities and to understate the required return of low beta securities. The returns of small companies, returns on certain days of the week or months of the year are observed to differ from those expected from CAPM.
(iii) Other factors in addition to systematic risk might influence required return. The arbitrage pricing theory (APT) suggests that a multi-factor model is necessary.
Can you please elaborate what is meant by the bold sentences.
Thanks so much
John.November 29, 2012 at 8:45 pm #109162
If CAPM worked perfectly, then the return actually given by a share would equal the return required for the beta (from the beta formula).
However the stock exchange is not perfect and some days a share gives a higher return than it should according to CAPM (the extra being a positive alpha) and some days a share gives a lower return that it should according to CAPM (the difference being a negative alpha).
However overall CAPM does seem to work and so what happens is that if there is a positive alpha it means that the share is giving a higher return that it should be giving. So people buy the share (while it is cheap) which pushes up the market value until it is giving the ‘correct’ CAPM return.
(b)(i) is really just stating something that appears to occur in practice. CAPM is not perfect and it does appear that the further away from 1 the beta is then the less ‘accurate’ it becomes.
What (b)(ii) is saying is that beta is measuring the risk due to general economic factors (such as inflation, changing exchange rates etc..). However APT is saying that the effect of each of these factors should perhaps be measured separately and instead of just having one beta to deal with everything, perhaps there should be several betas – for example a beta for the effect of inflation, a beta for the effect of exchange rates, etc.. The idea is very sensible but it has never been properly developed and so you cannot be expected to do anything in the exam except be aware of the idea.
You must be logged in to reply to this topic.