Thursday, June 25, 2009

Diversification In the Context of Number of Stocks

There is a general perception that higher number of stocks in a portfolio provides better diversification (i.e. better risk management). I am discussing the relationship between number of stocks and its effect on diversification. I am using very simple probability mathematics for an ideal scenario.

I am using a stock being positive or negative as measure of diversification. In an ideal scenario, at a minimum, one would like to have all stocks to be positive relative to the buy price. For example, if a portfolio has 5 stocks, then one would like to have all positive side. If a portfolio has 10 stocks, one would like to have all in positive side and so on. Please note that I am not discussing the value of individual stock or portfolio. It is likely that a positive value in one stock can offset the negative value of other stock.

My interest, here, is to break it down to the best possible scenario, and that is, stocks being +ve or stocks being –ve. I am calculating the probability of “one stock being +ve” in “one stock portfolio”, in a “two stock portfolio”, in a “five stock portfolio”, in a “10 stock portfolio”, and in a “20 stock portfolio”.

Furthermore, I am extending this probability to “two stocks being positive” or “three stocks being positive”. The results are then plotted for graphical presentation and discussion. The chart below shows the plot of probability of stock being positive vs. number of stocks in a portfolio. This chart is read as follows:

  • The curve, “1 +ve stock” is the curve (topmost curve, in red color) for probability of one stock being positive for a portfolio with 5 stocks, 10 stocks, 15 stocks, and 20 stocks.
  • If a portfolio consists of only one stock, then probability that it will be positive is 0.5 (i.e. 50% chance that it will be positive).
  • If a portfolio has 5 stocks, then the probability that at least one will be positive is 0.9.
  • If a portfolio has 10 stocks, then the probability that at least one will be positive is 0.95.
  • If a portfolio has 20 stocks, then the probability that at least one will be positive is 0.98.
  • The chart shows similar curves for probability of “two positive stocks”, three positive stocks, four positive stocks, and five positive stocks.


This chart shows that there is a significant increase in probability of stocks being positive until 5 or 10 stocks in the portfolio. Beyond that the change in probability of stocks being positive is very very small.

What we really want is to have a higher probability for stock to be positive. Let us say we want 0.8 probability, so we look at this chart horizontally at 0.8. We can have that with 5 stocks, 8 stocks, 10 stocks, and 12 stocks (intersection points between curve and full digit stock). As an individual investor what would you do? If the odds are same wouldn’t you try to use lesser number of stocks?

These simple probability curves show that there is an optimum point beyond which more stocks will not have any diversification benefits.

The best examples are DOW Index (with 30 stocks) and S&P500 index (500 stocks). I have superimposed these two indices on the same chart below. Both these index follow each other. If more stocks are supposed to provide diversification benefits, shouldn’t the S&P500 perform better than DOW?


In my next post, I will continue my discussion on some reason why quality is more important than quantity of stocks for better diversification.


This article was written by Dividend Tree. If you enjoyed this article, please vote for it by clicking the Buzz Up! button below.

1 comment:

  1. I would say that above picture gives you the impression that S&P has performed much better...is the picture not logarithmic on the y-axis?

    Looks to me that there is a big difference between DOW and S&P percentage wise at the end of the X-axis scale - or have I missunderstod something?

    ReplyDelete

Recent Posts From DIV-Net Members