Mean Reversion vs Momentum Trading
Mean reversion says a move away from a typical value tends to return. Momentum trading says a move that has started tends to continue. They are direct opposites about the same price action, and the evidence for each sits at a different horizon.
These two make directly opposing claims about the same bars. Most comparisons of them end in a preference; this one has a measurable resolution, and it is that they are describing different lengths of time.
What each one is
Mean reversion says a stretched move returns to a typical value. Buy weakness, sell strength. Mean reversion covers it.
Momentum trading says a move that has begun tends to continue. Buy strength. Momentum trading covers it, and trend following covers the exit rule momentum is usually paired with.
They cannot both be true of the same move at the same horizon. Whereas most pairs on this site divide labor, these two give opposite instructions about an identical chart, which is why the horizon question is not a technicality.
Where they differ
Which measurement supports each. Over a single day the S&P 500 has been close to a coin flip since 1950, closing up on 53.1% of sessions — price changes direction quickly and often, which is the reversion case. Over longer windows the same index leans one way, with 59% of ten-day windows finishing higher, which says direction over a longer span is less of a coin flip, the ground momentum stands on. Both figures come from one index, so the two methods are reading different horizons of the same market rather than contradicting each other.
How the loss behaves. This is the asymmetry that matters. A momentum trade that fails reverses, and the invalidation sits behind the level that broke — a defined place. A mean-reversion trade that fails gets more attractive as it worsens, because further from the mean is a stronger version of the original argument.
What each feels like. Mean reversion feels intelligent — you are buying cheap and selling dear. Momentum feels foolish — you are buying what has already gone up. Neither feeling has any relationship to how either method actually behaves.
Which is easier to run consistently. Momentum, because the exit is structural. Reversion requires a stop that the method’s own logic argues against, so the discipline has to be imported from outside it.
Where they agree
Both are entry rules and neither supplies an exit. That decision comes from somewhere else in each case.
Both are read off price alone, with no volume and no second input to check either against.
Both cost a round trip per trade — say 1 cent on a $0.50 average bar, or 2% of a bar, which the reversion trader pays far more often because the holds are short.
And both need external position sizing, which is the thing that actually determines whether either survives a bad run.
Which one to use: Mean Reversion or Momentum Trading?
Use mean reversion when your horizon is a handful of bars. Its case rests on price turning back quickly after short moves, and that is a claim about short intervals — stretch the hold and the argument weakens rather than grows.
Use momentum when your horizon is weeks or months. Longer definitions of strength filter out more of the short-lived moves, and a longer horizon gives a trend the time it needs to show itself.
Use momentum when you cannot supervise the position closely. Its failure is bounded by a stop the structure provides; reversion’s is not, and an unsupervised reversion trade is the configuration that ends accounts.
And when you cannot state your horizon, do not trade either. The contradiction between them is only resolvable by naming the timeframe, so a method without one will alternate between them at random.
Why horizon resolves the contradiction
Because both effects are present in the same data at different scales. Nothing about short-run oscillation prevents long-run persistence — a series can wobble bar to bar and still trend over fifty, and the index figures above show exactly that in one market.
And because the samples are small enough to hold loosely. Even the index figures describe one market over one stretch of history, and any single instrument’s own record is shorter still — these are pointers, not a law, and the reconciliation is a reasonable reading of them rather than a proof.
The original data
Mean reversion appears in 157 titles at a median of 3,857 views across 106 channels. Momentum trading appears in 82, at a median of 6,167 across 70.
Twice the videos on the contrarian side and two-thirds the audience. Buying weakness is the more appealing story and the more frequently told one, which fits the general shape of this corpus: the idea that feels clever outdraws the idea that feels uncomfortable.
On the chart above both frameworks give a confident and opposite answer. Only naming your horizon tells you which one is speaking to your situation.
When it fails
The characteristic failure is switching between them based on the last trade. A momentum entry fails, so the next stretched move gets sold instead; that reversion trade fails, so the next breakout gets bought. Because both edges are real at their own horizons, each switch is supported by a genuine argument, and the trader ends up systematically taking whichever one has just stopped working — buying strength after the trending period ended and selling it after the ranging period did. The record shows a sequence of defensible decisions and a consistently negative result, which is the hardest kind of problem to diagnose from a trade log.
A second failure is averaging into a reversion loss, where the method’s logic recommends more of exactly what is going wrong.
A third is running momentum in a range, where every breakout fails and each pays a round trip.
A fourth is treating a few remembered breakouts as large-sample evidence. A few dozen cases are not.
And a fifth is combining them into one system to hedge the disagreement, which produces a portfolio whose two halves cancel while both pay costs.
Related
Mean reversion covers the contrarian edge and its horizon. Momentum trading covers the continuation edge. And trend following covers the exit rule momentum needs.
This is the only pair on the site that genuinely contradicts, and the resolution is not that one side is wrong. Short-horizon price oscillates and longer moves persist, both measurably, so the question is never which is true — it is which horizon you are trading.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.