Position Trading vs Mean Reversion
Position trading is a holding period of months to years and supplies no reason to enter. Mean reversion is a reason to enter — that a stretched move returns to a typical value — and it is most reliable at short horizons, which makes the combination an awkward one.
Most pairings of a holding period and an edge combine cleanly. This one does not, and the reason is that the edge’s evidence and the holding period’s horizon are measured on different scales.
What each one is
Position trading is a holding period of months to years, and it says nothing about entries. Position trading covers it.
Mean reversion is an edge: price stretched away from a typical value tends to return. Mean reversion covers it, and trend following covers the exit rule that suits this holding period far better.
The horizons do not match. Whereas the reversion effect is most visible over a handful of bars — on this site’s shared series direction runs average 2.01 — a position trade is held for a span in which the market has ample time to establish a new typical value.
Where they differ
Which horizon each is about. The holding period is months. The reversion effect measured here operates over bars — 2.01 on average, 11 at the longest. Extending the horizon does not strengthen the edge, it gives the market time to make the stretched price the ordinary one.
What the drift does. On this series 54% of 566 ten-bar windows finished higher than they started, and that tax compounds with time — the longer you hold a position that bets against upward movement, the more of it you pay.
How the loss develops. Mean reversion’s loss argues for itself — further from the mean is a stronger version of the original case — and a long holding period gives that argument months to operate rather than minutes.
What would rescue it. Value investing is mean reversion on a long horizon and it works because the reference value is an earnings figure rather than a price average. Remove the fundamental anchor and the long horizon has nothing holding the mean in place.
Where they agree
Both need a market that is not trending permanently, which is the shared requirement and the one that is hardest to establish in advance.
Both require external position sizing. Neither the schedule nor the edge says how much to risk.
Both are indifferent to the round trip at 0.0098 against targets measured in many bar ranges — the execution cost is not what decides this one.
And both live through drawdown. On this series 95% of bars sat below a prior peak with the longest wait for a new high at 73 bars.
Which one to use
Use a trend-following exit with this holding period instead. It is the rule that actually fits months — a trailing stop survived a median of 32 bars at four average true ranges here across 562 trials — and it does not require betting against direction.
Use mean reversion over months only when the mean is anchored to something real. An earnings multiple, a yield, a spread between two related instruments — anything that exists outside the price chart and does not move just because price did.
Use mean reversion on a short horizon if you want the version the measurements support. The 2.01-bar run length is the evidence, and it is evidence about bars rather than months.
And when the anchor is just a long moving average, take neither. A mean defined by past price adjusts to the price, so it will meet you on the way down rather than pulling you back.
Why a long horizon weakens this particular edge
Because a price-based mean is not fixed. Any average computed from price follows price, so over months the reference value migrates toward whatever happened — the gap you were trading closes from the wrong end, and the position never benefits.
And because time turns overshoot into repricing. A market that moves a long way and stays there has not failed to revert; it has revalued, and nothing in the reversion framework can distinguish the two in advance.
The original data
Of the 24,971 unique videos in the search corpus, no title compares these two directly. Mean reversion appears in 151 titles at a median of 3,835 views across 102 channels. Position trading appears in 48, at a median of 3,115 across 40.
Two of the smallest audiences in the style group. Both are unglamorous and both are searched far less than the short-horizon styles, which fits the pattern throughout this corpus — attention tracks speed rather than evidence.
On the chart above the length of time it has stayed there is the argument against the trade. A price that has been extreme for months is a price the market has had months to reconsider and has not.
When it fails
The characteristic failure is shorting something expensive and holding it for months. The edge says the price is stretched, the holding period supplies the patience, and the position gets worse in a way that the method interprets as improvement — for months rather than minutes. Meanwhile the drift measured here finished higher in 54% of 566 ten-bar windows, so the passage of time is itself working against the position. There is no point at which the framework produces a stop, because the entire logic is that a larger gap is a better opportunity, and the horizon removes the only thing that used to limit the damage.
A second failure is using a long moving average as the mean. It follows price, so it meets you rather than pulling you back.
A third is treating a holding period as a strategy, which leaves a schedule and no reason to trade.
A fourth is ignoring the annual drag on a months-long position, which removes 20.2% of a thirty-year pot at 75 basis points regardless of the edge.
And a fifth is averaging down over a long horizon, where each addition has months to compound the error before any verdict arrives.
Related
Position trading covers the long holding period. Mean reversion covers the contrarian edge and the horizon it suits. And trend following covers the exit rule that fits this duration properly.
Value investing is mean reversion over years and it works for a reason that has nothing to do with the technical version — it rests on a company’s earnings, not on a price being far from its average. Stripping out the fundamental half and keeping the long horizon leaves you with the weakest form of both.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.