Thursday, November 13, 2008

Should You Make Investment Decisions Based on Performance History?
"Past performance is not a predictor of future results." We have all seen this phrase in virtually every document promoting an investment fund offered by an investment advisor or mutual fund company. It could be that it appears so often that investors may be de-sensitized to it and simply ignore the warning. After all, why are mutual funds any different from our favorite sports team? If our team has an unbeaten streak and the best record in the league, wouldn't we be reasonably safe in thinking that our team will win its next game?
But successful investment performance is not the same as successful athletic performance. While some renowned investors such as Warren Buffet, Peter Lynch and George Soros have indeed produced excellent and consistent investment performance, investors must exercise caution when selecting a fund or even an asset class such as stocks based solely on the fund's or asset's recent performance. Why? Because one can rarely, if ever, statistically prove that the performance history of a fund is a predictor of future performance. So, when making a decision on how to invest your savings, you should not focus on the investment's recent performance record. Understanding the concept of serial correlation will help you understand why.


Serial Correlation: There's no proof history will repeat itself
The term serial correlation is simply a statistical exercise that determines the degree of confidence that one might have when predicting the future based on past observations. One looks at a series of numbers or events to see if there is a pattern and tries to determine how likely it is that the pattern will continue. If a series has perfect serial correlation, then one should be very confident that he or she could predict the future based on past observations. If a series has no serial correlation, it means that future observations will be random and are completely unrelated to the past.

Are you trying to predict the market? You might as well flip a coin
The best example of a series that is random is the flip of a coin. If you pulled a coin from your pocket or purse and tossed it in the air five times and all five times it came up heads, what do you think the chance is that the next flip will be heads? Do you think that it would be heads again just because the first five flips were heads? Of course not!
As in flipping coins, it has been mathematically proven that the future performance of the stock market has no serial correlation with past performance. It is random. The stock market going up on a day after five straight days or during a year after five straight years of positive performance is like a coin flip. It is purely a random event.
The same can be said for mutual fund performance. Just because a fund produced five years of excellent performance does not mean that one should expect a sixth great year. Whether the fund performs well or not for the short term is just like a coin flip. As for any investment, one needs to think long-term. One simply cannot predict long-term mutual fund performance based on a short-term record.


It Doesn't Always Pay to Follow the Leader
It was not unusual for staff at the General Board to receive phone calls from participants during 2002 who said something like: "I have just switched all of my money from the Domestic Stock Fund to the Domestic Bond Fund. Bonds are performing well and stocks are performing poorly." In 2003 the story is different. During the summer, participants have called saying: "I am switching my money out of the Domestic Bond Fund and into the other funds that are performing better."
When one says, "performing better", one is implying that they believe that the past is a predictor of the future. For investors, this is almost always a loser's game. Investors who follow a philosophy of buying after a class or a fund has performed well are doomed to a lifetime of "buying high and selling low". These investors will experience significant degradation of wealth and a lot of sleepless nights.


So, what is an investor to do?
First, you must resolve never to make an investment decision based solely on historical investment performance. Think of it like a coin flip that cannot be predicted and resolve to use another method. Second, you should consider seeking the advice of a fee-only financial planner. Such an individual will help tailor a financial plan for you based on your own unique circumstances, financial goals and risk tolerance. Third, with the help of your financial planner, develop a plan and stick to it.
Too often investors abandon a carefully thought out plan just at the wrong time. In the late 1990s, many people with diversified investment portfolios jettisoned their bonds and loaded up on technology and telecommunications stocks right before the bubble burst. As a result, these people saw significant losses in their investment portfolios, and many have had to defer their retirement plans. In 2002, many people jettisoned stocks in their investment portfolios in favor of better performing bonds. As a result, they missed the recent rally in stock prices. In both cases investors abandoned their long-term investment strategy at the wrong time and their wealth accumulation was significantly affected.
There is a good reason why you see the phrase "past performance is not a predictor of future results". The reason is because it is the truth! While at first it may appear that there is a pattern to stock or fund performance, in fact, it is merely random and therefore unpredictable. You can provide yourself with greater assurance of financial independence by following a well thought-out strategy. That way you can avoid disappointing results from selecting investments based solely on past performance.


*article by http://www.gbophb.org/sri_funds/articles/perfhistory.asp

No comments: