Finding the Best Funds for Diversification

If you've done any reading or watched any TV on the topic of investing, you've undoubtedly run into the topic of diversification, and heard how important it is. It's often the only real form of portfolio risk management that is generally recommended in the popular press, but it seems to often that folks write about it without really comprehending what makes it work.

You get advice like:

1. Put 75% of your money in stocks, and 25% in bonds
2. Buy at least 8 stocks, none in the same industry
3. Put 5% of your money in precious metals
4. Put 20% of your money in international stocks

and so forth. Now as it turns out, there is some truth behind all these, but as stand alone advice they give you an incomplete picture.

Why Standard Deviation and Who Cares?

To give a reasonable idea of how diversification works, we will need to talk about portfolio risk as measured by its standard deviation of returns. (I know I said we would use drawdown, but this topic really requires that we use SD). What does that mean? I will not grind through the math, but (making some gross and marginally incorrect simplifications) if you take the average return of an investment, and then create a window of +/- 1 standard deviation, you would bracket the returns of that investment about 2/3 of the time.

For example, if an investment returned an average of 10%, with a standard deviation of 5%, then 2/3 of the time the return would be between 10 +/- 5%, or between 5 and 15%.

So, what does the historical return of the market and it's risk look like?

For the S&P 500, it turns out that over the years 1970 to 1998, the average annual return was about 14%, with a standard deviation of about 15%. For small stocks the standard deviation was over 20%.

Even the bond market (as represented by 20 year treasuries) had a standard deviation of over 11%.

You can find the standard deviation of returns for many funds at Morningstar, listed under the "Risk Measures" tab. You need to be aware that it lists the standard deviation for the trailing 3 years (they calculate it using monthly returns). For a string of years like 2003-2006, that's a little misleading, as the market has not had a significant correction in that time.

There is a lot more information about this topic in the book "The Intelligent Asset Allocator" by William Bernstein.

You can buy it at Amazon at

http://fundztrader.com/intelligent

If you want to get a quick introduction to the topic the first two chapters are online at his website at:

http://www.efficientfrontier.com/BOOK/chapter1.htm

and

http://www.efficientfrontier.com/BOOK/chapter2.htm

If that's too much math for you to understand, he has written another book on the topic, "The Four Pillars of Investing", he characterizes it as being "for liberal arts majors", i.e. no math.

Either of these books is quite readable, as I understand it when he wrote these books he was a practicing neurosurgeon, so he's not out to impress you with some leftover MBA thesis material, but he's sharing some information he picked up from the school of hard knocks.

So, standard deviation of returns is simply one measure of risk.

Next time: Next we look at what diversification does to reduce risk, and in doing so we find some good investment choices to help control risk.