Guide

Mean reversion vs momentum

They are opposite bets on the same price. Knowing which one the market is rewarding is most of the skill in trading either.

Two opposite bets

Momentum says a price that has moved will keep moving; mean reversion says a price that has moved too far will come back. They cannot both be right about the same move at the same time, and yet both are durable, well-documented strategies — because they are right at different horizons and in different regimes. Treating them as rivals to be chosen between once, forever, is the mistake. They are tools, and the regime tells you which one to reach for.

The horizon usually decides

As a rough rule that the data keeps supporting, momentum tends to dominate over medium horizons — weeks to months a price that has been rising keeps rising — while reversion tends to dominate at the extremes: the very short term, where overreactions snap back within hours, and the very long term, where stretched valuations eventually correct. A reversion strategy is, in effect, a bet that you are at one of those extremes rather than in the momentum-rewarding middle. That is why the holding clock matters so much: it is also a bet about which force is in charge.

The regime decides the rest

Within any horizon, the market alternates between trending phases, where momentum pays and reversion bleeds, and range-bound phases, where the opposite holds. No one calls these phases perfectly in advance, which is exactly why a disciplined reversion strategy insists on a stop: the stop is the admission that you might be trading reversion in a momentum regime, and it caps the cost of being wrong about the regime. A reversion trader without a stop is implicitly claiming to know the regime with certainty, which no one does.

A useful way to hold both ideas at once: momentum is the trade you put on when the crowd is right and reversion is the trade you put on when the crowd has overshot. The hard part — the part a measured, graded record helps with — is telling those two situations apart before the outcome rather than after.

Why this site is about reversion specifically

This site teaches mean reversion rather than momentum for one reason: it is the harder of the two to verify, and therefore the one where a checkable record is worth the most. A momentum claim is relatively easy to read off a chart after the fact; a reversion claim — “I caught the bottom of a stretch before it turned” — is almost impossible to distinguish from hindsight unless the trade was timestamped before it resolved. That is the whole reason the issued-before-the-outcome test sits at the centre of how this site judges a reversion record.

Keep reading

Is mean reversion profitable?The conditions, including the regime condition.Holding clocksWhy the horizon decides which force is in charge.The statistics of reversionStationarity, the property reversion needs and momentum does not.Issued before the outcomeWhy reversion is the harder claim to verify.