Wednesday, October 18, 2006

When Lots of Stocks Are High, Should You Buy?

Each day in the Trading Psychology Weblog, I track the number of stocks making new 20-day highs and 20-day lows. These give us an intermediate-term sense of whether the market is strengthening or weakening. How about the bigger picture, however? If many stocks are making 52-week highs, does it pay to buy?

I went back to 1990 (4196 trading days) and computed the number of 52-week NYSE highs as a percentage of the total issues traded each day. What we find is that there were 136 occasions in which 10% or more of traded stocks made fresh 52-week highs. One such occasion occurred this past Monday.

What we find is that, 40 days after we get a large number of new highs, the S&P 500 Index ($SPX) is up on average by only .32% (68 up, 68 down). That is a subnormal return when compared to the average 40-day return of 1.47% (2695 up, 1501 down) over that period.

In other words, buying stocks when lots of stocks are high has not been a good buy.

How about the reverse?

When only .5% or fewer of stocks are making new highs (N = 202), the next 40 days in $SPX average a sizable gain of 4.34% (164 up, 38 down), much better than average.

When very few stocks are making new highs, it's been worth a buy.

It is only human nature to become more bullish as more stocks are making fresh highs. But if human nature were rewarded in the marketplace, we'd see more trading millionaires.