The Long-Term Relative Strength Report with Random Timing

As a follow-up to our Long-Term Relative Strength with Market Timing report, we did a control study of sorts. The purpose of the study was to discover what contribution the Market Timing added to the use of the Long-Term Relative Strength - Strong report.

What we did

The goal was to create a series of random trades to compare to the market-timed trades. We took the length of each in-the-market and out-of-the-market period from 11/09/88 to 01/06/97 and randomized them. This gave us the same number periods in the market (44) as our Relative Strength with Market Timing test, and started and ended on the same dates. Of course, the other trade dates were not aligned.

What we found

This graph shows our findings most clearly. Note that not only is the horizontal axis not a linear time scale, the trades from the different systems are not aligned (except that the first trades all started on 11/10/88 and the last trades all ended on 01/07/97)..

Specifically we found::

• Buying the top 5 stocks from the Relative Strength-Strong report on the S&P 500 on the random dates did surprisingly well, averaging 31.7% per year compounded for the 8+ years studied. In addition, the equity curve was surprisingly smooth.
• As reported in the Relative Strength-Strong with Market Timing report, buying the top 5 stocks from the Relative Strength-Strong report on the S&P 500 averaged 42.9% per year compounded for the 8+ years studied.
• The return for the randomly timed S&P 500 index dropped to 9.4% per year, substantially below the Buy and Hold strategy at 13.1% per year. Interestingly, the randomly-timed index return was almost exactly 70% of the Buy and Hold Return, which corresponds closely to the 70% of the time that was spent in the market.
• We did not try principal protection stops or minimum volumes.

But wait a minute, what's this?

While we were running the Relative Strength with Random Timing study, we noticed that buying the top 5 stocks from the report was no longer the optimum number! Buying 4 was better than 5, buying 3 was better yet, and buying just 2 gave the best results of all. We don't know what this means, but it will probably give rise to a future study. In the meantime, here is what the equity chart looks like:

Buying the top 2 stocks from the Relative Strength-Strong report on the S&P 500 on the random dates averaged an amazing 51.7% per year compounded for the 8+ years studied. In addition, the equity curve was also surprisingly smooth.

Now what?

Send an email and let us know what you think. We'll have to look through our back issues of AIQ Opening Bell Monthly and talk to David Vomund. In the meantime, take a look at the Short-Term Relative Strength with Market Timing study and the Relative Strength Summary, too.

Home AIQ Software Contacts Studies