Last week I presented a data analysis of the historical behavior of the S&P 500 Index over the last forty years (approximately). I showed how buying all-time (meaning at least 5 years) high prices when days closed at such highs could be used as a very profitable trading strategy, and that relatively volatility could have been used effectively as a filter to make the strategy more profitable still.
The first thing I want to add to this in this second part of the piece, is introducing a control value. For example, I showed that buying all-time highs at the daily close and holding for one month produced an average return of 0.43%, but what about doing the same on any day? We need this information as a comparison control, so let’s make a table of how all the results compare, using all-time highs, 50-day highs, and all days which feel into neither category, over the same time period.
Time Period of High Price | 1-Day Holding Period | 1-Week Holding Period | 1-Month Holding Period | |||
Win Rate | Average Result | Win Rate | Average Result | Win Rate | Average Result | |
5-Year | 51.54% | 0.03% | 56.23% | 0.04% | 61.28% | 0.41% |
50-Day | 50.55% | 0.00% | 54.60% | 0.07% | 57.41% | 0.61% |
Neither | 53.40% | 0.04% | 56.63% | 0.20% | 61.95% | 0.79% |
This is a surprise – the lows produce better results than the highs! It seems that when it comes to the broad stock market, the saying “buy the dips” is valued for a good reason – buying when the price is off the highs has produced better results, on average, than buying the highs.
There is one further thing to consider before drawing conclusions from this analysis. The last 40 years have seen some very strong and long-lasting bull markets. It might be that if I made this same analysis going back over 100 years instead of only 40, the all-time highs would come out ahead. This is because buying at all-time highs would tend to keep you out of trouble in bear markets. So, there is something else to consider here, which I will return to in the future.