As with all other markets, the S&P is momentum based. The onset of a new up or down direction generally won't fizzle right away. Initial buy/sell imbalance will slowly and surely bring in other players who want to either initiate or terminate a position before it becomes too late. Market orders feed on themselves.

All markets also mean revert. That is, markets back off of overextended levels. If you didn't have both components, you wouldn't have the market movement we all recognize. The S&P, though, follows mean-reversion in unexpected ways. The usual pattern of smaller moving averages being the direction to follow versus larger ones is sometimes flipped on its head.

Nearly two decades ago, I published the original full-book version of "Beating the Financial Futures Market." That was the first of my books to include a substantial number of biases and actual stand-alone systems of my own design. I offered so-called "either-or" building blocks, which were simple indicators that would signal a long or short on the 8:30 a.m. day session open and an exit on the 3:15 p.m. session close of nearly every day.

One of the most durable indicators of all was the 5-day versus the 2-day average closes. Again, contrary to expectations, you would go long when the 5-day average closed above the 2-day average, and short when the 5-day average was under the 2-day average. In the book, I combined the indicator with other components to achieve better results, but remember markets and systems do drift over time. Nineteen years have not been kind to most of the compound methodologies. The one basic element that survived the eons was the 5-day average versus 2-day average.

The 2007 through September 2025 time span represents virtual virgin territory. This system was invented prior to 2007 and has rolled forward unaltered since then. The S&P generated $70,790 during the span. Granted, all we're receiving here is a most persistent bias rather than a tradable system. The worst drawdown was $75,197, and the return on account was a puny 94.14 percent.

Think of the result mainly as an affirmation of the mean-reverting aspect. The other four markets confirmed it.

The Dow made $13,895, with a $68,160 worst drawdown, and 20.37 percent ROA.
The Nasdaq made $103,010, with a $54,275 worst drawdown, and 189.79 percent ROA.
The Russell made $94,590, with a $26,305 worst drawdown, and 359.59 percent ROA.
The Midcap made $195,039, with a $60,921, worst drawdown, and 320.15 percent ROA.

Consider how arbitrary the exit-on-close rule is. If we never exit a position until the averages flip around, at which time exiting and reversing, we get the following net profit, worst drawdown, and ROAs.

S&P: $184,725, $51,063, 361.76 percent.
Dow: $43,850, $64,520, 67.96 percent.
Nasdaq: $285,310, $79,105, 360.67 percent.
Russell: $109,310, $34,025, 321.26 percent.
Midcap: $255,510, $57,420, 444.98 percent.

From here, it shouldn't be hard to imagine leveraging the unique 5-day versus 2-day average cross bias into something viable. We certainly have history on our side.

Art Collins has been a member of the Chicago Board of Trade for over 36 years.