Monday, November 29, 2010

The Average Depth of a Lake

The other day, I found myself writing a response to an op-ed in the Wall Street Journal written by Burton Malkiel from November 18th. If you don’t know, Malkiel is one of the most stalwart of the buy and hold crowd, and his article, called “Buy and Hold is Still a Winner,” pointed out the usual arguments for buy and hold investing. At one point in the article he observes that if you bought and held the S&P 500 Index from 1995 through 2009 you could have earned an average return of 8%, but if you missed the best thirty days of returns through poor market timing your return for the period would have been negative. The corollary to this statement is that if you measure a slightly longer period, from 1990 to 2008, and you miss the 30 worst days of performance, your return increases from 5.06% annually if you bought and held, to 14.5% annually! For the average stay-at-home investor, buying and holding and rebalancing is something that you can do. To take a crack at the 14.5% annual return you might want to retain the services of a professional advisor.

But that’s not the point of this blog. I’m fascinated with investor fixation on average returns. The average return for a time period is one of the most useless pieces of information I can imagine. Malkiel tells us that the buy and hold return from 1995 to 2009 is 8% on average. But what if you get to the average by earning 0% on your money for the first 7.5 years and then earning 16% on the next 7.5 years? If you are withdrawing money from your portfolio to fund your retirement, the results could be catastrophic. Your spending would decrease your capital to the point that you would possibly not have enough left to meet your objectives, even though your money was earning 16% per year for the second half of the time period. Financial planning research tells us that it isn’t the average of returns that matters. Instead, it is the order of returns that matters. Active and tactical portfolio management allow investors to defend against disorderly markets when they can least afford them.

Average portfolio returns are like the average depth of a lake. If I tell you the average depth is one foot deep and you can’t swim, are you going to try and walk across to a far shore that is a mile in the distance? Obviously the average depth is a useful piece of information for someone, but it isn’t relevant to the decision at hand. What if the path you take across the lake is actually 100 feet deep, even though the average is one foot? It would take a giant leap of faith to start walking across if you can’t swim. To focus on average returns for asset classes creates a similar problem. Without knowing why assets returned whatever they returned, the information is useful, but somewhat irrelevant. Investors hoping to earn the average returns of stocks should beware. The data strongly suggests that when buying at high normalized (or smoothed) P/E ratios, the odds of earning historical average returns are very low. Unfortunately, the 10-year normalized P/E ratio is about 22 times earnings currently. Buying and holding from here is like walking across that lake.