Sharpe’s wrote his thesis on CAPM showing how to go from both individual’s stock and market risks to the portfolio weights which optimize the risk/return trade-off.
Basically, when the market has reached equilibrium and only one factor market is driving returns, the set of efficient portfolios is the set of diversified portfolios. If a stock offers a lower expected return than required by its market sensitivity, its price would fall down and its expected return would rise. In the opposite case, the price would rise and the expected return would go down.
Relying on the single factor intuition, Lintner’s version may appear more general than Sharpe’s but on the other hand, Sharpe's work justified the empirical market model suggested by Benjamin King.
King’s idea was that stock prices change because of new information: some affecting the entire market, some affects only certain industries, the last ones specific to an individual firm. He found a common factor called the market factor which is responsible for at least half of stock price fluctuation.
A for Neo.
ReplyDeleteKing found that the market factor explained half of stock movements. This is amazing when you consider the context - no one doing investing had ever measured the extent to which everything moved together (although practitioners admitted to familiarity with the pattern).
Sharpe was able to tell them why: because all that mattered in a diversified portfolio was covariances. If you have all the covariances ... you have the market portfolio. So, the biggest factor pushing stock prices was the covariance of all the stocks with each other.
The CRSP data mentioned here is also huge. Everyone who does serious investments research still uses it. Up until that time, even though huge volumes of stock data were generated every day, no one had ever collected it and looked at it. The first thing they did when they got that data was to establish a baseline for the return on stock investments - in a bad period that included the Great Depression - of 9%. We think that's low now, and something like 12% seems more realistic.