Mean reversion

Mean reversion on Hyperliquid

Mean-reversion traders make money from the market's tendency to overshoot and snap back. The equity curve is smooth and the win rate is flattering — right up until a move does not come back.

In short

A mean-reversion trader fades extended moves, expecting price to return toward a recent average. Win rates are high and individual gains are small, but losses are comparatively large because the losing scenario is a move that keeps going. The archetype is the natural counterweight to trend following in a basket.

The mechanics

Positions are opened against the immediate direction of price, usually after a rapid extension, and closed on a partial retracement rather than a full reversal. Holding periods are short, turnover is high and the profit per trade is modest.

The risk profile is the mirror image of trend following: frequent small wins, infrequent large losses. Read in isolation, a 75% win rate looks like skill; read alongside average win/loss it may only describe a payoff structure that has not yet met its bad day.

High win rate, small average win, larger average loss.
Sensitive to how quickly a losing position is cut.
Adding to losers is common in this archetype and is a discipline red flag.
Performance degrades sharply when a market shifts into sustained trend.

Where the tail risk sits

The failure mode is not gradual. A mean-reversion book can look immaculate for months and then give back a large share of accumulated profit in a single regime break, because every position is on the wrong side of the same move at once.

Position discipline is therefore weighted heavily for this archetype during scoring: sizing stability and the absence of martingale-style adds separate a durable mean-reversion trader from one that is quietly selling insurance.

Why it belongs in a diversified basket

Mean reversion earns during exactly the chop that starves trend followers. Held together and isolated in separate sub-accounts, the two archetypes take turns contributing rather than cancelling each other inside one netted position.

Past performance is not indicative of future results. Perpetual futures are leveraged instruments and carry a substantial risk of loss, including the loss of your entire position.

Questions

Frequently asked

Is a high win rate a good sign?

Only with context. High win rate plus a large average loss can describe a strategy whose worst outcome has not happened yet. Profit factor and drawdown history matter more.

How is adding to losers detected?

From the public fill record: increases in position size while the position is underwater, especially repeated ones, are visible and counted against position discipline.

Does mean reversion copy well?

Reasonably. Holds are short but not sub-minute, so replication lag introduces tracking error without destroying the edge. Very short-horizon variants copy poorly.

What breaks the strategy?

A sustained directional regime. The archetype assumes overshoots revert; when they do not, losses accumulate on every open position simultaneously.

Can a basket be all mean reversion?

It can, but it should not be. That concentrates the same tail risk across every sub-account at once.

Keep reading

Trend followingThe counterweight archetype that profits from sustained moves.Strategy correlationMeasuring whether two leaders actually diversify each other.Holding period and turnoverHow trade frequency changes replication fidelity.Risk disclosureWhat can go wrong, stated plainly.

Start Autopilot View live performance

Diversified copy trading. On autopilot.

Score-weighted allocation across up to 10 elite Hyperliquid traders, each isolated in its own sub-account. Your funds never leave your account.