There is another inconvenient truth that has nothing to do with Al Gore or carbon footprints, and it’s not that Al Gore seems to be leaving deeper footprints these days. The truth is that, despite every subtle and not-so-subtle attempt to imply otherwise, past performance rarely has anything to do with whether an investment manager is anything more than lucky. Yes, we all know that “past performance is not a guarantee of future returns”, but in reality not only does past performance not guarantee future returns, more often than not it has absolutely nothing to do with future returns.

Even a monkey throwing darts can pick investments correctly on occasion, but this luck won’t be repeated consistently given enough time. So how long does a monkey (or investment advisor) need to consistently beat a benchmark before he can be considered good instead of just lucky?

The Quarter-Flipping Example

Although this example is not the most accurate, it’s probably the easiest to visualize. If you give 2000 advisors each a quarter with the goal to flip heads, the first round of flips would result in roughly 1000 of them succeeding. This is similar to assuming that each has a 50/50 chance of beating the S&P 500 in a given year. After 9 flips (or 9 years), three advisors are likely to have flipped heads 9 times in a row, at which time all the other advisors will be scrambling to find their “secret” and duplicate it. The three lucky advisors will have their own radio shows where they sell a 5-DVD set showing how you can follow their system. In April of this year, there were 56,709 people with the CFP® designation. Chartered Financial Analyst (CFA) numbers are harder to come by, but in 2006 alone, 116,000 people took at least one of the CFA exams. So with 50/50 odds, how long will it take to separate good money managers from lucky ones? As I mentioned, this is not the most accurate example, so let’s leave it at “a long time”.

Statistical Example

This example is far more accurate because it has the word “statistics” in it. It also uses an equation that factors in standard deviation, correlation, and other accurate-sounding terms. It even begins with a more precise question:

If an advisor beats his benchmark by 2% each year (most advisors would sell their grandmother for 2%/year), how many years does it take to determine this is due to skill rather than luck with a statistical 95% confidence level?

The answer depends on the volatility, or standard deviation, of the portfolio (higher volatility takes more years), so I’m assuming the standard deviation is 20%, which is close to the S&P 500. And the answer is…56 years. At 4% per year the number drops all the way to 16 years, and at 6%, an advisor is statistically good instead of lucky after only 8 years. Warren Buffet has achieved “good” as opposed to “lucky” status, but most advisors need to squeeze as much as they can out of their positive years before their luck runs out.

Does It Take 50 Years To Choose An Advisor?

This doesn’t mean that performance is not important, it just means that performance alone is a poor indicator of whether an advisor will outperform his benchmark in the future. So if you believe in luck, simply try to pick a lucky money manager. Previous lottery winners would be a good start. Otherwise, simply make sure the advisor’s philosophy is aligned with yours, that his (or her) strategy makes logical sense, and that sales fees, expenses and commissions don’t put you in a hole before you even start, and don’t get too hung up on each year’s performance.