A stock closes red seven days running and your brain fills in the rest. The bears have had their turn. Somebody has to cover. It has to bounce.
That’s the same instinct that makes people bet red after roulette lands black six times. In roulette it’s provably useless because each spin is independent. In markets it feels different, because there are humans on the other side with positions to close. So I tested it.
424 losing streaks of seven days or more since 2000, measured against a proper baseline: buying randomly in stocks already below their 200-day moving average. Next day, worse than a coin flip. Next week, worse. Over a month, worse than the baseline. At three months, identical. Winning streaks don’t collapse either. Currencies, nothing. Commodities, nothing.
The streak count on its own is not an edge. People remember the handful of times it worked and forget the hundreds it didn’t.
One exception, and it’s the indices. When the SPY is down seven straight, that’s not a company having a bad fortnight, that’s national news and two weeks of relentless selling. The win rate there is high, on a sample of about 24 occurrences, which is small enough that I’ll say so out loud rather than dress it up.
That’s what most research actually looks like. A reasonable question, a null result, and one nugget inside the noise worth pulling on.
If you want the systems that survived this kind of testing, that’s Stats Edge Pro. $149.99 a month, 30-day money-back guarantee.
Michael Nauss, CMT, CAIA, CDMS