I recently finished reading 'Security Analysis' by Benjamin Graham and David Dodd. It made me fundamentally rethink investing in the stock market. There were two large lessons in this book. One is the distinction between investment and speculation. An investment requires analysis, it should preserve capital, and produce a reasonable rate of return. Anything other than that is speculation. Graham and Dodd think that most so called investors should be called speculators. The other large lesson is the importance of price and what you are getting for a given price. Be sure that you are not overpaying. Graham and Dodd suggest you use analysis to estimate the fundamental price and then add a margin of safety (say 20%). This policy preserves capital as everything you buy should be under-priced and therefore be expected to rise in value.

I was curious how these ideas could be applied to index funds. For example, what if you bought an index fund when the S&P 500 was underpriced and then sold when overpriced? Of course, we need a way to determine price. One way suggested by Graham and Dodd is looking at the average price over the past 5 years. So if the current price is 20% below the 5 year moving average, BUY! And when the current price is 20% above the 5 year moving average, SELL! Classic buy low, sell high.

How well would this simple policy have performed since 2001? I estimate a return of 67% or 6.1% a year. There were only 3 trades Buy at 990 in May 2002, Sell at 1320 in March 2006 (Gained 33%), Buy at 1050 in Sept 2008. Currently the index is at 1260 (Gained 20%). But wait that only looks like 53%. Well there were also dividends paid for the approximately 7 years of the holding. Assume 2% dividends (the current S&P 500 rate) and you get an extra 14%. I also assumed that the dividends were not reinvested.

So how does this compare to buying and holding the S&P 500 for those 11 years? Well you would get 22% in dividends and would have lost 4% of your capital for a net 18% or 1.63% a year. I guess price DOES make a difference!

This strategy does well for being so simple. Of course it is possible to do much better. If you had a time machine and could only buy and sell the S&P 500 just twice you could make 214% or 19.45% annualized! The trades are Buy at 780 in October 2002, Sell at 1580 in September 2007 (Gain 102%), Buy at 680 in April 2009, Sell at 1350 in May 2011 (Gain 98%), plus 7 years of dividends.

The problem with the buy low sell high strategy, as Graham and Dodd point out, is that it is difficult for humans to actually execute. It requires patience and the faith that the depressed prices will return to above the five year average.