Testing TradingMarket's Rule #2: Buy the Market after It's Dropped, Not after It's Risen

This is part two of this week’s series on TradingMarket.com’s (TM) 10 Trading Rules. In this post, I’ll look at TM’s Rule #2: buy the market after it’s dropped not after it’s risen.

As evidence, TM looks at the market after it has either fallen or risen for three consecutive days and shows that the former is much more bullish than the latter. Longtime readers know that I’m a firm believer in this type of short-term mean reversion, which I’ll demonstrate below in the simplest way I know how.

click to enlarge

[logarithmically-scaled]

The graph above shows two strategies – one going long the S&P 500 at today’s close if today closes down (red) or if today closes up (green), from 2002. This is a proof of concept, so this test is frictionless (ignores transaction costs and slippage).

Clearly, in recent history, the market has been a mean-reverting animal on a day-to-day basis; up days have tended to follow down days (and vice-versa).

For the number-lovers:

click to enlarge

However, all of this ignores the fact that this short-term contrarian tendency is a very recent phenomenon. The next graph shows the same test from 1950 buying after a close down (red) or close up (green).

click to enlarge

Prior to this decade, the market tended to follow-through after short-term moves, not mean-revert. I’ve talked about this inflection point in the market before in the "Evolution of Daily Follow-Through" and offered one possible solution in "The Simple Made Powerful with Adaptation."

So, do I agree with TM’s rule to buy the market in the short-term after it’s dropped, not after it’s risen? I do, wholeheartedly, but I would be wary of trading that approach without a healthy respect for the fact that at some point (maybe 100 years from now or maybe tomorrow), the market will shift again.

The other articles in the series can be found here: Testing TradingMarket's 10 Trading Rules, Part 1, Part 3 and Part 4.