If we look at the Dow over a period of years, we can see that it is up a little over 20% on a three-year basis, a little more than 10% on a two-year basis, and a little more than 5% on a one-year basis. I decided to go back all the way to 1942 and investigate all historical periods that had a similar trajectory: three year returns between 20-30%; two-year returns between 10-20%; and one-year returns between 5-10%. The resulting sample, drawn from 15,740 trading days, gave me 191 occasions, but many of these were clustered within the same cyclical periods.
The most recent occasion when we saw the market trading similarly was February, 2001. If you recall, the NASDAQ had taken quite a tumble over the previous year, but the large cap Dow stocks had held up well. Over the next two years, however, the Dow fell a grievous 26%.
The largest number of occasions cluster in the 1992-1995 time period--particularly during March-October, 1994. Here, if you recall, the market faced rising interest rates, but the economy did not go into deep recession. The market drop was quite mild and, two years later, we were up by about 50%.
We next see a cluster in January and February of 1981. Sadly, I'm showing my age by remembering that market quite well. It was also a period of high interest rates. We had bounced off a 1980 low, but were headed lower later that year and steadily until August, 1982, when we launched the great bull market. One year after the 1981 period that resembled our current market, we were down about 10%, but two years later we were up by the same amount, thanks to that 1982 lift off.
February, 1973 provides the next cluster of days similar to the current one. It was not a good time in the market, as we bounced off a decline from the 1972 highs, but then fell precipitously into the December, 1974 lows. This was a time of high energy prices, high interest rates, and deep recession. Two years later, the market was down by more than 25%.
February, 1962 is our next correlate. We had the missile scare and market plunge that year, but quickly righted ourself. The market was down about 3% one year later, but up a little over 10% two years afterward.
Next we have our second largest cluster of similar market periods spanning 1953-1954, with the majority of occasions in Feb.-May, 1953. That market had risen sharply off 1949 lows and was showing modestly higher highs and lower lows for over a year. One year later, the market was up about 10% on average, but two years later it was up by about 50%, reflecting the strong growth of the post-war period.
If we lump all the occasions, we can see that the market was up on average by 10.9% one year later and 36.1% two years later. That two-year performance is stronger than the market's average two-year gain of 17.09% covering the entire sample. Clearly, the two periods of real weakness (2001 and 1973) occurred in the context of major economic slowdown, augmented by event shocks (oil, WTC bombing).
Here's the interesting part. If we look *three* years out, all periods similar to the current one were profitable, except for those 2001 and 1973 occasions, which only lost less than 3% at worst. Indeed, the average three-year gain was 72.4%, much stronger than the average three-year gain of 26.3% for the sample overall.
This time may be different, I suppose, but the historical record suggests that buying for the longer run on near-term weakness would not be a bad strategy. Unless you're anticipating a rerun of 1929, 1973, or 2001, buying a three-year up market in which gains have been moderate over the past year is an attractive proposition. I'll post more on this in the Trading Psychology Weblog tonight.