As bad as the news might have been in the September account statement from your broker, the best thing to do now is forget it. No one should shift to a new adviser or strategy based on one sad month of performance.
I reached that conclusion after studying the performance of investment newsletters during and after other sharply negative months of past decades. Almost always investors chasing good performance then would have regretted their decisions.
A spectacular illustration of the pattern comes from the market crash of Oct. 1987, which was far worse for US stocks than were the events of Sept. 11 this year. In fact, the combined value of the country's equities fell 22.8 percent in October 1987, versus 9 percent last month.
Nevertheless, as a group, the newsletters that performed the best during October 1987 -- which, on balance, were aggressively bearish -- have been among the poorest performers over the years since.
Consider the top 10 letters from that month, ranked by performance, among those monitored by The Hulbert Financial Digest. Over the next 14 years, those 10 letters have produced an average annualized gain of just 4.5 percent, or just over one-third the 12.9 percent registered by the Wilshire 5000. Even 90-day Treasury bills have done better, at 5.3 percent a year, on average. Now consider the 10 newsletters with the worst performances in Oct. 1987. Six of them have since outperformed the Wilshire 5000. And as a group, those 10 have produced an annualized gain of 10.2 percent, less than that of the Wilshire 5000 but more than double the average return of the 10 best-performing letters that month.
You may object that the results reflect little more than the bull market over most of the last 14 years. Yet the market's performance could well have turned out differently in the wake of 1987's Black Monday. There was no guarantee that stocks would not enter a long-term bear market, and if they had, the pattern might have been different: the newsletters with the best subsequent performances could have been among those with the best gains in Oct. 1987.
But my conclusion still holds. My argument is based on performance over the last 14 years, and bear markets hardly ever last that long. Since 1802, there have been 186 different, though overlapping, 14-year periods, and in fewer than 4 percent of them have stocks failed to produce a positive real return, according Jeremy Siegel, a professor at the Wharton School of the University of Pennsylvania.
If the future is like the past, the probability is more than 96 percent that the advisers with the best returns last month, generally market bears, will be poor long-term bets. Even in the unlikely event that stocks decline over the next 14 years, the top performers last month would still be poor bets.
That is because the best advisers over a period as short as one month tend to be those that have incurred the greatest risk. Regardless of the market's overall direction, it is most unlikely that such advisers can make enough money to overcome the huge losses that their high risk inevitably brings when they misstep -- something that even the best advisers do occasionally.
A more intelligent bet is on advisers with the best long-term records. Only when track records are measured over many years does past performance become even a part guide to the future.
RSS