Mean Reversion Signals That Actually Work

Markets oscillate. They overshoot on fear, overcorrect on greed, and spend most of their time reverting to equilibrium after brief extremes. This is not a theory -- it is a statistical property that has persisted across asset classes and time periods for as long as markets have existed. The mechanism is straightforward. When price moves sharply in one direction, several forces push it back: profit-taking by traders who rode the move, value buyers stepping in at perceived discounts, options dealers hedging their gamma exposure (which creates counter-directional flow), and algorithmic mean reversion strategies that are explicitly designed to capture this pattern. The persistence of mean reversion is not a market inefficiency waiting to be arbitraged away. It is a structural feature of how markets process information. Overreactions happen because participants are human (or because algorithms are responding to other algorithms). The reversion happens because prices that deviate significantly from fair value attract capital that pushes them back. The challenge is not whether mean reversion exists -- it clearly does. The challenge is identifying WHEN it is likely to work (low volatility, no catalyst) versus when it is likely to fail (high volatility, trend regime). This is where volatility context becomes essential.

Internal Bar Strength (IBS) is the most straightforward mean reversion indicator available. The formula is trivial: (Close - Low) / (High - Low). It produces a value between 0 and 1. An IBS below 0.2 means the market closed near the low of its range -- sellers dominated at the close. An IBS above 0.8 means the market closed near its high -- buyers were in control at the close. The mean reversion thesis: extreme IBS readings tend to reverse the following session. The documented performance is notable for its simplicity. Over 10 years of ES futures data, buying after an IBS below 0.2 and selling after an IBS above 0.8 produces a win rate in the range of 65-70% on the next session's direction. This is not a complete strategy -- there is no stop loss, no position sizing, no risk management in the raw signal. But as a directional filter, it is remarkably consistent. IBS works best in quiet to normal regimes (CRC 0-40). In elevated and crisis regimes, extreme IBS readings are more likely to represent the beginning of a trend than an overextension. A close near the low during a market selloff is not a buying opportunity -- it is confirmation that selling pressure is genuine. This is why combining IBS with volatility context matters: the signal is the same, but the interpretation depends on regime.

Larry Connors popularized the 2-period RSI as a short-term mean reversion signal. The logic: a 2-period RSI captures very short-term momentum extremes. When RSI(2) drops below 10, the market has been falling aggressively for 2 consecutive periods. When it rises above 90, it has been rallying hard. The documented win rate for buying RSI(2) below 10 on SPY (and by extension ES) is approximately 68% for a positive return over the next 1-5 trading days. The effect is strongest at 1-2 day horizons and decays beyond 5 days. The average winning trade is modest in size, which means this is a high-frequency, high-win-rate, small-edge strategy -- not a home run approach. RSI(2) and IBS are correlated but not identical. IBS is a single-bar measure (where did price close within today's range?). RSI(2) is a two-bar momentum measure (how aggressive was the recent move?). When both are at extremes simultaneously -- IBS below 0.2 AND RSI(2) below 10 -- the combined signal is stronger than either alone. The overlap indicates both single-bar and multi-bar overextension. The same regime caveat applies. RSI(2) below 10 during a trending selloff is not a reversal signal -- it is a confirmation of trend strength. Filtering RSI(2) signals to only fire in low-volatility regimes (rating 1-4) significantly improves the win rate at the cost of fewer signals.

Certain calendar effects create reliable mean reversion setups. The most well-documented is Turnaround Tuesday: after a Monday decline, Tuesday has a statistically significant tendency to reverse. The effect has been documented in academic literature dating back decades and persists in modern data. The mechanism is partially structural. Monday declines often reflect weekend risk repricing -- news that accumulates over the weekend hits the market at Monday's open. By Tuesday, the overreaction has been absorbed and value buyers step in. Institutional rebalancing flows also concentrate on Mondays and Tuesdays, creating predictable counter-directional pressure. Consecutive down days create another calendar-based setup. After 3 or more consecutive declining sessions in ES, the probability of a positive session increases to approximately 62-65%. After 4 consecutive down days, the probability rises further. This is not a large enough edge to trade blindly, but combined with IBS or RSI(2) confirmation, it provides additional conviction. End-of-month and beginning-of-month effects are also documented. Institutional rebalancing -- pension funds, mutual funds, and index funds adjusting allocations -- creates predictable flow patterns. The last 3 trading days of the month and the first 3 of the new month tend to have a slight positive bias in equities, driven by systematic buying from these flows.

No mean reversion signal works in all conditions. The common thread across IBS, RSI(2), calendar patterns, and Counter VIX Reversal (CVR) signals is that they perform best in low-to-normal volatility environments and degrade or invert in high-volatility trending regimes. CVR deserves specific mention. The Counter VIX Reversal signal fires when VIX spikes above its 10-day moving average by a threshold amount while the underlying market declines. The thesis: VIX spikes represent fear overshoots that tend to normalize, pulling the market back up. Documented win rates are in the 65-70% range for ES over the subsequent 1-3 days. The practical framework for combining these signals: 1. Check the regime first. If CRC is above 60, mean reversion signals are unreliable. Do not fade the trend. 2. In calm regimes (CRC 0-40), look for signal convergence. IBS below 0.2 plus RSI(2) below 10 plus a calendar pattern (Tuesday after Monday decline, or 3+ consecutive down days) creates a high-conviction mean reversion setup. 3. Size for the signal quality. A single signal (IBS alone) gets minimum size. Two converging signals get standard size. Three or more get maximum size (within your risk rules). 4. Use the volatility rating for stop placement. The expected range from the forecast defines how much room the trade needs. A mean reversion entry with a stop at 40% of the expected daily range gives the trade room to work while capping downside. 5. Take profits early. Mean reversion is a high-win-rate, small-average-win strategy. Holding for home runs turns a winning approach into a losing one because the reversion move is typically modest -- a return to the mean, not a reversal to the opposite extreme. Every signal referenced here is documented with its data source and sample period in our Sources and Citations documentation. These are not backtested curve-fits -- they are persistent statistical effects measured across decades of data.