To: Wyätt Gwyön who wrote (2134 ) 8/20/2001 2:08:40 AM From: Dan Duchardt Read Replies (1) | Respond to Mucho,if you create a portfolio that deliberately mixes asset classes that have different degrees of correlation with each other (covariances), your portolio's return will (obviously) be the weighted average return of its component asset classes, BUT (not so obviously, and quite interestingly), the standard deviation of the portfolio as a whole will be less than the average of the SDs of the component assets more precise as the size of the sample grows larger. If the sample is not random, there is a theorem called the Central Limit Theorem that says that if you take several samples, even if each is itself not random, the distribution of the means of the samples will become random (or normal) as the number of samples grows, the mean of the sample means will approach the mean of the universe, and the standard deviation of the group of samples will be less than the individual standard deviations.must be selective and attempt to avoid the elements of the universe that are below average, keeping the elements that are above average. In other words, you must try to deliberately skew your sample toward the better performing stocks or sectors. If you are right, your sample average is better than that of the universe, probably with higher volatility. If you are wrong, you may do worse, but that is the risk you must accept for trying to do better. If you don't try to do better in order to protect against doing worse, that is exactly what you will get- no better and no worse.
