Mean Reversion Strategies

In trading, there are many theories and strategies, but one of the most proven and understandable for beginners is Mean Reversion. Its essence is simple: the price of an asset over time returns to its long-term average value after significant deviations. When the asset has grown too much – you sell, when it has fallen too much – you buy, expecting a correction. In this article we will take a detailed look at how the theory works, which indicators to use, which strategies to apply in practice, how to manage risks and when it is better to refuse to enter a trade. We will also talk about Pairs Trading – one of the most famous implementations of this concept.

Introduction to Mean Reversion

Short answer: what is Mean Reversion

Mean Reversion is a trading concept based on the assumption that the price of an asset over time returns to its conditional average level after significant deviations.

In simple terms: if the price has grown too much or fallen too much relative to the “norm”, it is highly likely to correct in the opposite direction. Mean reversion traders earn precisely on these corrections:

The average value is usually calculated as a moving average over a certain period – for example, 20, 50 or 200 days.

Important warning for beginners: Mean reversion is an effective concept, but it does not guarantee profitable trading:

• The price may continue to move away from the average for a long time and unexpectedly
• The average value also moves – the price may “stall”, and the average “catches up” with it (this is also considered reversion)
• The trend may change direction, and the strategy will not work

Therefore, professional traders always additionally use strict risk management protocols: stop-loss and determining the exit point if the price does not go in the expected direction.

Historical and statistical context

The concept of mean reversion is not new – it is rooted in statistics and econometrics:

• In statistics, the term regression toward the mean describes the phenomenon when extreme values in a sequence of data over time approach the average.
• In finance, mean reversion has become a separate theory, suggesting that asset prices and historical returns ultimately return to the long-term average level.
• One of the classic mathematical models is the Ornstein-Uhlenbeck process – it describes the return to the mean, which is used to model price behavior with a tendency toward the mean.
• Stock reporting services have long offered traders moving averages for 50 and 100 days as a basic tool for determining buy/sell points.

Mean reversion is observed in many asset classes: stocks, currency pairs, commodities, even exchange rates. However, it is important to understand: the return process can take years, which makes the strategy less suitable for short-term investors.

Why it works

Mean reversion works due to several fundamental mechanisms:

• Feedback in market processes. The process of returning to the mean in any time series is due to the influence of positive or negative feedback. When the price approaches the upper or lower limit, the probability of reversal increases.
• Thermostatic analogy. Time series with reversion to the mean behave like the temperature in a room with a thermostat: when it is above the set level, the thermostat turns on cooling to return it to normal; when below – heating.
• Limited sustainability of extremes. Extreme price movements in any direction are temporary and cannot last long. There are two reasons:
• In overbought conditions (when the price is too high), buyers become more cautious, sellers more active, which puts pressure on the decline.
• In oversold conditions (when the price is too low), buyers find the asset attractive, putting pressure on growth.
• Use of technical indicators. Traders use tools based on measuring deviation from the average, for example: Bollinger Bands, RSI (Relative Strength Index), MACD or PPO.

Research confirming the theory

Additionally, the theory is confirmed by a number of major studies and articles in specialized literature. For example:

• Mean Reversion in International Stock Markets: An Empirical Analysis of the 20th Century – an article analyzing Mean reversion in international stock markets in the period 1900-2008.
• Optimal Mean Reversion Trading with Transaction Costs and Stop-Loss Exit – analytics illustrating the dependence of timing strategies on model parameters such as transaction costs and stop-loss level. Originally published in International Journal of Theoretical and Applied Finance, Vol. 18, No. 3, 2015.
• Efficacy of a Mean Reversion Trading Strategy Using True Strength Index – a comprehensive analysis of a trading strategy based on return to the mean value using the True Strength Index (TSI), applied to ETF SPY (S&P 500) and QQQ (Nasdaq index). The study covers historical data from 1996 to 2022, including various market conditions, to assess the strategy's sustainability.

Key Takeaways

• Mean Reversion is a strategy based on the fact that after significant deviations, the price returns to its average value.
• It works best in flat (sideways movement) and on liquid assets.
• Main tools: Bollinger Bands, RSI, Z-score, Stochastic, MACD.
• Always use confirmation from 2–3 indicators and strict risk management (risk no more than 1–2% per trade).
• Do not trade against a strong trend — in this case, it is better to use Trend Following.
• Pairs Trading is one of the most reliable implementations of Mean Reversion for stocks.
• Combine Mean Reversion and Trend Following in your portfolio for stability in any market conditions.
• Before every trade, be sure to go through the checklist.

Now let’s take a closer look at how exactly the Mean Reversion theory works in practice.

How Mean Reversion Works in Trading

Now let's take a detailed look at the mechanics of mean reversion strategies: how to calculate the average value, how to measure the price deviation from it, what signals indicate buying or selling, and what mathematical model describes this process.

Identification of the average value

The central element of mean reversion is determining the average value to which the price will return. In trading, the average is calculated as a moving average over a certain period. Two main ones are used:

• SMA (Simple Moving Average) – calculated as a simple arithmetic average of prices (usually closing prices) over the selected period. Each new period “shifts” the calculation window, discarding the oldest value and adding a new one. SMA smooths out price fluctuations, helps determine the trend direction and serves as dynamic support/resistance. The longer the period, the smoother and less sensitive to market noise the SMA line becomes.
• EMA (Exponential Moving Average) – an exponential moving average, a more sensitive to recent prices variant of the moving average. Unlike SMA, EMA gives more weight to recent prices (usually closing prices) and gradually reduces the weight of old data according to an exponential law. The calculation uses a special smoothing coefficient that determines how quickly the indicator reacts to price changes. Thanks to this, EMA reacts faster to trend reversals and new market impulses, but it can more often give false signals in a sideways market. The most popular periods are 9, 12, 26 and 50. EMA is widely used in combination with SMA and in MACD indicators.

Formula for SMA over n periods:

SMA = (P₁ + P₂ + … + Pₙ) / n

where P_i – closing price on the i-th period.

Formula for EMA:

EMAₜ = Pₜ × k + EMAₜ₋₁ × (1 − k)

where k = 2/(n+1) – multiple smoothing coefficient.

What period to choose?

• 20 days – short-term average, sensitive to rapid changes
• 50 days – medium-term, often used to determine the trend
• 200 days – long-term, “long-term norm” of value

The choice of period depends on your timeframe: for intraday trading (H1–H4) use 20-day EMA, for daily (D1) – 50 or 200-day SMA.

Measuring deviations

After calculating the average, you need to understand how far the price has deviated from it. Only significant deviations give a signal to enter. The following calculation methods are usually used for this.

Standard Deviation

Standard deviation measures the degree of dispersion of prices around the average. The greater the deviation, the stronger the price has “gone” from the norm.

Formula for standard deviation over n periods:

σ = √[Σᵢ₌₁ⁿ(Pᵢ − μ)² / n]

where μ is the average value (SMA or EMA).

How to use in trading:

• Price at ±1 standard deviation from the average – normal fluctuation, signal is weak
• Price at ±2 standard deviations – significant deviation, potential signal
• Price at ±3 standard deviations – extreme deviation, high probability of correction

Z-score

Z-score is a standardized measure of deviation, showing how many standard deviations the price is from the average.

Formula:

Z = (P − μ) / σ

where:

• P – current price
• μ – average value (SMA/EMA)
• σ – standard deviation

Z-score interpretation:

Z-score is especially useful for comparing different assets: it standardizes deviations, making them comparable regardless of absolute price.

Signals: buy (oversold), sell (overbought)

The next main point of Mean Reversion is a pair of basic signals.

Buy signal (oversold)

When it occurs:

• The price is significantly below the average (usually below the lower Bollinger Band or Z < -2)
• RSI below 30 (oversold according to RSI – we will talk about this indicator in more detail later)

What we do:

• Open long expecting the price to return up to the average
• Stop-loss: set slightly below the minimum to protect against continued decline
• Take-profit: target price – the middle line (SMA/EMA) or slightly higher

Example:

• SMA(20) = $100
• Current price = $90
• Z-score = -2.5
• Signal: buy, we expect growth to $100

Sell signal (overbought)

When it occurs:

• The price is significantly above the average (above the upper Bollinger Band or Z > +2)
• RSI above 70 (overbought according to RSI)

What we do:

• Open short expecting the price to return down to the average
• Stop-loss: slightly above the maximum
• Take-profit: middle line (SMA/EMA) or slightly lower

Example:

• SMA(20) = $100
• Current price = $110
• Z-score = +2.3
• Signal: sell, we expect decline to $100

Additional indicators for signal confirmation

Important: Never enter based on only one indicator. Use confirmation from 2+ tools (for example, Bollinger Bands + RSI + Z-score) to reduce false signals.

Mathematical model: Ornstein-Uhlenbeck process

For the mathematical description of mean reversion, the Ornstein–Uhlenbeck process is used – a stochastic model that describes the behavior of a time series with a tendency to return to the mean.

Differential equation Ornstein-Uhlenbeck:

dXₜ = θ(μ − Xₜ)dt + σdWₜ

where:

• X_t – asset price at time t
• μ – long-term average value (mean)
• θ – speed of return to the mean (the larger, the faster the return)
• σ – standard deviation (volatility)
• W_t – Wiener process (stochastic component, “noise”)

How to interpret the equation:

• θ(μ−X_t)dt – deterministic part describing the return to the mean
• σdW_t – stochastic part describing random fluctuations (market volatility)

The Ornstein-Uhlenbeck process is used in quantitative trading, parameter evaluation and strategy testing.

Practical advice: If you are building a quantitative mean reversion model, evaluate the θ parameter for each asset. Stocks with high θ (fast return) are better suited for short-term mean reversion.

Key Indicators and Strategies Based on Them

In modern trading terminals, you do not need to calculate everything manually. Instead, you can focus on the behavior of basic indicators. Let's look at the most popular types.

Bollinger Bands

The first of the two mentioned above and, perhaps, the most popular indicator in mean reversion is Bollinger Bands. It visualizes the average value and deviations from it in the form of “bands” and is built on the basis of standard deviation:

• Upper band: SMA + 2σ
• Lower band: SMA − 2σ
• Center line: SMA (usually 20 periods)

where σ is the standard deviation of prices over 20 periods

Typical signals:

• Touching or breaking the upper band signals overbought, time to sell.
• Lower band – about oversold – time to buy.
• The price returns to the center line – We close the position in profit.

Important nuances:

• Bands expand with high volatility, meaning deviations can be stronger.
• Bands contract with low volatility, often before a strong move.
• Do not enter immediately on a breakout: sometimes the price “walks the band” in a strong trend.

It works much more reliably in combination with others (for example, with RSI and Z-score).

RSI (Relative Strength Index)

The second of the mentioned and popular ones is RSI – a momentum indicator that measures the speed and change in price movements. It perfectly shows overbought and oversold conditions.

How it is built: RSI = 100 – (100 / (1+RS))

where RS = (average growth over n periods) / (average decline over n periods)

Standard period: 14 (you can use 9 for short-term trading).

How to use for mean reversion:

• RSI > 70 – overbought – sell, expect decline.
• RSI < 30 – oversold – buy, expect growth.
• 30–70 – normal zone, no signal, wait.

Examples of signals:

Buy-signal (oversold):

• RSI dropped below 30
• Price breaks the lower Bollinger Band
• Entry: long, stop-loss below minimum, take – center SMA line

Sell-signal (overbought):

• RSI rose above 70
• Price breaks the upper Bollinger Band
• Entry: short, stop-loss above maximum, take – center SMA line

Divergence on RSI

An additional strong signal is divergence between price and RSI:

• Positive divergence: price makes a new low, but RSI – a higher low → buy
• Negative divergence: price makes a new high, but RSI – a lower high → sell

Stochastic Oscillator

Stochastic is another overbought/oversold indicator, similar to RSI, but more sensitive to rapid changes.

Consists of two lines:

• %K (fast): (C − L_n) / (H_n − L_n) × 100
• %D (slow): SMA(%K) over 3 periods

Standard parameters: (14, 3, 3).

How to use:

• %K > 80 – Overbought – sell if %K crosses %D downward
• %K < 20 – Oversold – buy if %K crosses %D upward

MACD (Moving Average Convergence Divergence)

MACD is a trend and momentum indicator.

Consists of:

• MACD line: EMA12 − EMA26
• Signal line: SMA(MACD) over 9 periods
• Histogram: difference between MACD and signal line

Signals for reversal:

• MACD upward crossover after deep decline → buy
• MACD downward crossover after strong growth → sell
• Divergence → strong reversal signal

MACD works best in combination with Bollinger Bands and RSI.

Moving Averages

Moving average is the basis for all mean reversion strategies.

How to use:

• Use SMA(20) or EMA(20) as the center line.
• Golden Cross / Death Cross help determine the overall trend (mean reversion works better in flat).
• Price deviation > 2 standard deviations from SMA(20) → potential entry.

Summary table: main indicators

Recommended combinations of indicators

Experienced traders recommend not relying on one indicator, but using 2–3 tools at once for additional confirmation.

Combination 1: “Basic” (for beginners)

• Bollinger Bands + RSI
• Buy: price below the lower band + RSI < 30
• Sell: price above the upper band + RSI > 70

Combination 2: “Quantitative” (for advanced users)

• Bollinger Bands + RSI + Z-score
• Buy: price below the lower band + RSI < 30 + Z < -2
• Sell: price above the upper band + RSI > 70 + Z > +2

Combination 3: “Momentum” (for short timeframes)

• Stochastic + MACD + EMA(20)
• Buy: %K < 20 + upward crossover + MACD upward crossover + price < EMA
• Sell: %K > 80 + downward crossover + MACD downward crossover + price > EMA

Practical advice: how not to overload the chart

Many beginners add all indicators to the chart at once – this is a mistake that only complicates trading. Instead, use the following recommendations:

• Choose maximum 3 indicators: 1 for the average (Bollinger/MA), 1 for overbought (RSI/Stochastic), 1 for momentum (MACD)
• Use the rest of the indicators only for additional confirmation, if the first signal is weak
• If the signals contradict each other – do not enter the trade, wait for a clear match

Intraday vs Swing vs Long-term

Mean reversion can be applied on different timeframes, but the approaches to trading will differ substantially. Let's break down what suits each style.

Intraday trading

Intraday traders work on short timeframes: from 5 minutes to 1 hour. Their goal is to catch several small price corrections during the day and close before the end of the trading session.

What to use:

• Timeframes: M5, M15, H1
• Indicators: Bollinger Bands (period 20), Stochastic (14,3,3), EMA(20)
• Trade frequency: 3–10 positions per day
• Time per one trade: from 15 minutes to 2 hours

Features:

• Intraday mean reversion requires quick decisions. The price can quickly bounce off the Bollinger band and return back, and if you do not have time to enter, the signal will be lost.
• Use Stochastic instead of RSI – it is more sensitive and works better on short timeframes.
• Stop-loss should be narrow (usually 10–30 points on forex), because you do not want to hold a losing position.
• Take-profit is also small: usually this is a return to the center line of Bollinger Bands (15–40 points).

Who it suits: Traders who can sit at the chart for several hours a day and quickly react to signals. Not suitable for those who work a full day and want to trade “on autopilot”.

Swing-trading (medium-term)

Swing traders hold positions from several days to several weeks. They catch larger corrections and are not so dependent on intraday fluctuations.

What to use:

• Timeframes: H4, D1
• Indicators: Bollinger Bands (period 20), RSI (14), SMA(20) and SMA(50)
• Trade frequency: 1–3 positions per week
• Time per one trade: from 2 days to 3 weeks

Features:

• Swing trading is calmer, because you do not need to constantly monitor the chart. You can check signals in the morning and evening, and spend the rest of the time on your own affairs.
• On the daily timeframe RSI works better than Stochastic – it gives fewer false signals.
• Stop-loss is wider: usually 50–150 points on forex, because the price can fluctuate more strongly during the day.
• Take-profit is also larger: 100–300 points, as you catch a deeper correction.

Who it suits: Traders who work a full day and want to trade without constant chart monitoring. Ideal for those who prefer a calmer style with fewer trades.

Long-term trading

Long-term traders and investors hold positions from several months to several years. They use mean reversion to determine entry points into assets that are temporarily cheapening or becoming more expensive relative to their “normal” value.

What to use:

• Timeframes: D1, W1 (weekly), M1 (monthly)
• Indicators: SMA(50), SMA(200), RSI (14)
• Trade frequency: 2–6 positions per year
• Time per one trade: from 3 months to several years

Features:

• On long timeframes mean reversion works slower. The price can remain oversold (RSI < 30) for several weeks or even months before correction.
• Use long-term moving averages: SMA(50) and SMA(200). They show the “long-term norm” of the asset's value.
• Stop-loss is very wide: 200–500 points on forex, or even 10–20% drawdown on stocks.
• Take-profit is large: 300–1000 points on forex, 20–50% on stocks.

Who it suits: Investors and long-term traders who believe in the fundamental value of assets and are ready to wait months for price correction. Not suitable for those who want quick profits.

Which style to choose?

The answer depends on your schedule and psychology:

• If you work a full day and cannot sit at the chart for several hours, swing trading is better for you. You check signals in the morning and evening, set stop-loss and take-profit, and can calmly do your work. The position is held for days or weeks, and you do not depend on intraday fluctuations.
• If you are a student or work a flexible schedule, and you have 3–4 hours a day for monitoring, you can try intraday trading. It is more dynamic, gives more trades, but also requires greater concentration and can bring more stress.
• If you are an investor and believe in the long-term value of assets, long-term mean reversion will give you the opportunity to buy good companies or assets “cheap” during temporary declines and sell “expensive” during temporary growths.

Why risk management is critical

Mean reversion is a strategy with naturally high risk. You enter a position when the price has already gone far from the average, and expect it to return, but:

• The correction may not start immediately. The price may remain oversold or overbought for several days, weeks or even months.
• It may “spread” along the band. In a strong trend, the price may break the upper or lower Bollinger Band and continue moving in the direction of the trend, rather than returning to the average.
• The average also moves. Sometimes the price does not return to the average, but the average “catches up” with the price. This is also considered a return to the average, but for you it is a loss if you are already in a position.

Therefore, without strict risk management, mean reversion can quickly deplete your deposit. Even professional traders use strict capital protection protocols.

Rule 1: Position Sizing

One of the most important parameters is how much capital you allocate to one trade.

Recommendation for beginners:

• Do not risk more than 1–2% of the deposit on one trade.
• If you have a $10,000 deposit, the maximum loss on one position should be no more than $100–$200.

How to calculate: Position size = (Deposit × Risk per trade) / (Stop-loss in points × Point value)

Example:

• Deposit: $10,000
• Risk per trade: 1% = $100
• Stop-loss: 50 points
• Point value (on EURUSD): $0.10 per point on 1 micro-lot

Position size = 100 / (50 × 0.10) = 100/5 = 20 micro-lots = 0.20 lots

That is, you open 0.20 lots, and in the worst case you will lose only $100 (1% of the deposit).

Important: If the stop-loss is wider (for example, 100 points), the position size needs to be reduced to 0.10 lots so that the risk remains $100.

Rule 2: Stop-Loss

Stop-loss is your automatic protection against large losses. In mean reversion, it needs to be set correctly, otherwise the strategy will not work.

How to set stop-loss in mean reversion:

• For Long (buy) – slightly below the last minimum (recent low) that was before the entry.
• For Short (sell) – Slightly above the last maximum (recent high) that was before the entry.

Example for long:

• Price breaks the lower Bollinger Band
• RSI < 30 (oversold)
• Last minimum before entry: $95
• You enter long at $96
• Stop-loss set at $94.50 (slightly below the minimum)

If you set the stop-loss too close to the current price (for example, at $95.50 when entry at $96), the price may simply “jerk” and knock out your stop-loss before the correction. Leave a small margin (usually 5–10 points on Forex).

If the stop-loss is too far (for example, at $90), you risk losing too much on one trade. This violates the 1–2% risk per trade rule.

Rule 3: Take-Profit

Take-profit is your target exit point in profit. In mean reversion, the goal is usually clear: return of the price to the average value.

How to set take-profit in mean reversion:

• For Long (buy) – on the center line of Bollinger Bands (SMA 20) or slightly higher.
• For Short (sell) – on the center line of Bollinger Bands (SMA 20) or slightly below the entry.

Example for long:

• SMA(20) = $100
• Price breaks the lower band at $90
• You enter long at $91
• Stop-loss at $89.50 (below the minimum)
• Take-profit at $100 (center line) or $101 (slightly higher)

Ideally, the risk/reward ratio should be at least 1:1.5 or 1:2.

Example:

• Stop-loss: 50 points (risk $100)
• Take-profit: 100 points (profit $200)
• Ratio: 1:2

If the ratio is less than 1:1 (for example, stop 50 points, take 40 points), the trade is statistically unprofitable even with a 50% success probability.

Rule 4: Maximum number of open positions

Do not hold too many positions at the same time. If you open 5–10 trades at the same time, even with proper risk management on one trade, the total risk can become too high.

Experienced traders recommend the following:

• Intraday: 1–3 positions at the same time
• Swing: 2–5 positions at the same time
• Long-term: 3–10 positions at the same time (diversification is more important here)

The total risk of all open positions should not exceed 5–10% of the deposit.

Example:

• Deposit: $10,000
• Maximum total risk: 5% = $500
• Risk per trade: 1% = $100
• Maximum positions: 500 / 100 = 5 positions

If you already hold 5 positions with $100 risk each, do not open a sixth until one of the first ones closes, regardless of profit or loss.

Rule 5: Diversification

Diversification is the distribution of risks across different assets. If you trade only one currency pair (for example, EURUSD), and something unpleasant happens on it (for example, sudden news), you will lose everything. But a competent distribution of the deposit will help avoid this.

How to diversify:

• Work with at least 3–5 different assets (currency pairs, stocks, precious metals).
• Do not trade highly correlated assets at the same time (for example, EURUSD and GBPUSD often move in the same direction).
• Combine different types: forex + stocks.

Example of diversification:

• EURUSD (forex)
• AAPL (Apple stock)
• XAUUSD (gold)
• CLUSD (oil)

If one position is unprofitable, others may be in profit, and the overall result will be positive.

Rule 6: Trailing Stop

Trailing stop is a dynamic stop-loss that moves with the profit. It allows you to “fix” part of the profit if the price suddenly reverses.

In this case, it can be used as follows: when the price has reached 50% of the way to take-profit, move the stop-loss to breakeven (to the entry point). When the price has reached 75% of the way, move the stop-loss to 50% of the profit.

Example:

• Long entry: $91
• Stop-loss: $89.50 (risk $1.50)
• Take-profit: $100 (profit $9)
• Price reached $95.50 (50% of the way): move stop-loss to $91 (breakeven)
• Price reached $98.25 (75% of the way): move stop-loss to $95.50 (fix $4.50 profit)

The advantage of this approach is that if the price suddenly reverses after 75% of the way, you will close with $4.50 profit instead of zero.

At the same time, it is important to remember that in mean reversion the price can “fluctuate” near the center. Use trailing stop carefully.

Pairs Trading – one of the most famous implementations of Mean Reversion

Pairs Trading is a statistical arbitrage strategy based on the idea of mean reversion. Its essence is simple: you trade the price difference between two correlated assets, and not the direction of the assets themselves. The Ornstein-Uhlenbeck model mentioned earlier is used as the mathematical basis.

How Pairs Trading works

• You choose two correlated stocks (for example, AAPL and MSFT, Coca-Cola and Pepsi)
• You look at the price difference (or difference in returns) between them
• When the difference deviates from the average (one stock is significantly more expensive than the other relative to the norm):
• Buy the “cheap” stock (long)
• Sell the “expensive” stock (short)
• Expect the difference to return to the average – then close both positions in profit

Practical example

Situation:

• Historically, AAPL and MSFT move approximately the same (correlation ~0.8)
• Usually their price ratio: AAPL = 1.2 × MSFT
• Today: AAPL = $180, MSFT = $150, ratio = 1.2 – norm
• In a week: AAPL = $190, MSFT = $150, ratio increases to 1.27 (AAPL overheated)

Action:

• Buy MSFT (cheap relative to norm)
• Sell AAPL (expensive relative to norm)
• Expect the ratio to return to 1.2
• If AAPL falls to $180 and MSFT rises to $155, the ratio drops to 1.16 and the order can be closed in profit

Important: You are protected from market movement. If the entire market falls, but AAPL falls more than MSFT – you are still in profit.

How to choose good pairs

Examples of successful pairs: AAPL vs MSFT, Coca-Cola vs Pepsi, JPM vs BofA, Exxon vs Chevron.

Methods for identifying pairs

• Principal Component Analysis (PCA)
• OPTICS (clustering algorithm)
• Cointegration testing

Hedge Ratio

For proper Pairs Trading, you need to calculate the hedge ratio – how many units of one stock to sell for each unit of the other.

Calculation methods:

Practical advice for beginners

If you are a beginner in Pairs Trading:

• Start with 2–3 simple pairs from the same sector (AAPL/MSFT, Coca-Cola/Pepsi).
• Use OLS for hedge ratio – simple but workable.
• Trade on a demo account until you understand the dynamics.
• Do not trade more than 5 pairs at the same time, as a large number is too difficult to control.
• Check the correlation daily – if it is < 0.5, close the pair.

Mean Reversion vs Trend Following

In trading, there are two fundamental approaches: mean reversion and trend following. They are opposite in logic, but both can be profitable. The question is not which strategy is “better”, but in what conditions each of them works more effectively, and how to use them together correctly.

What is the main difference

Let's immediately designate the difference:

In simple terms:

• Mean reversion – you buy when the price has fallen too much, and sell when it has jumped too high. In both cases, you initially assume that the deviation is temporary, and the price will return to a conditionally initial position.
• Trend following – you buy when the price has started to rise, and hold the position while the trend continues, hoping that the deviation will be maximum.

When Mean Reversion works better

Mean reversion is effective in flat conditions – when the market is in balance, there is no clear direction, and the price fluctuates around the average value.

Specific situations:

• Horizontal moving averages: if SMA(50) and SMA(200) are almost horizontal and not far from each other – this is flat. Mean reversion works well.
• Low or moderate volatility: when Bollinger Bands are not too expanded, the price more often bounces off the bands rather than “spreads” along them.
• Absence of important news: in periods between news (NFP, FOMC, central banks) the market is more stable, and corrections occur predictably.
• Overbought/oversold according to RSI: if RSI regularly goes into zones >70 and <30, and then quickly returns to the 30–70 range, the market is prone to mean reversion.
• Assets with high liquidity: currency pairs (EURUSD, GBPUSD), large stocks (AAPL, MSFT), gold. They have less “noise” in movement, and the price more often returns to the average.

When Trend Following works better

Trend following is effective in strong trend conditions – when the price moves in one direction with high energy, and corrections are weak or short-term.

Specific situations:

• SMA(50) significantly above/below SMA(200): if SMA(50) > SMA(200) and both are growing – uptrend. If SMA(50) < SMA(200) and both are falling – downtrend.
• High volatility in one direction: when Bollinger Bands are expanded, and the price has “spread” along the upper or lower band for several days – this is a trend, not flat.
• Consecutive highs/lows: if the price makes a new high after a new high (upward) or a new low after a new low (downward) – the trend is strong.
• RSI remains in the extreme zone for a long time: if RSI >70 holds for several days without returning to 50, the trend is upward and strong. If RSI <30 holds for a long time – downward.
• Macroeconomic drivers: when there is a fundamental reason for the trend (for example, Fed rate growth, war, crisis, technological boom), the trend can continue for months and years.

How to determine whether it is flat or trend now?

Before entering any trade, it is important to understand what is happening in the market now. There are several ways to do this:

• Check by moving averages:
• Flat: SMA(50) and SMA(200) are horizontal, close to each other, often intersect.
• Trend: SMA(50) and SMA(200) diverge, one is significantly above the other, both are directed in the direction of the trend.
• Check by Bollinger Bands:
• Flat: Bands have contracted, narrow, price bounces off them.
• Trend: Bands are expanded, price has “spread” along the upper (upward) or lower (downward) band.
• Check by RSI:
• Flat: RSI quickly goes into >70 or <30, but quickly returns to 50.
• Trend: RSI stays in >70 (upward) or <30 (downward) for a long time, does not return to 50.
• Check by highs/lows:
• Flat: Price makes highs and lows in approximately the same range, no clear direction.
• Trend: new highs above previous (upward) or new lows below previous (downward).

Practical advice: Before each trade, open the chart on D1 and check all 4 criteria. If at least 3 out of 4 indicate flat, use mean reversion. If 3 out of 4 indicate trend – trend following.

Combining strategies in a portfolio

Instead of choosing one strategy for constant work, professional traders combine mean reversion and trend following in a portfolio. This allows:

• To be profitable in different market conditions (flat and trend)
• To diversify risks (if one strategy is unprofitable, the other may be in profit)
• To increase the overall portfolio profitability

Moreover, you can combine in different ways.

1. By assets

Assign different strategies to different assets:

2. By timeframes

On short timeframes (H1–H4) mean reversion more often works, on long ones (D1–W1) – trend following.

3. Division by market situation

Monitor the overall market context and change strategy priorities:

• If the market is in flat (for example, after a period of instability, before important events), increase the share of mean reversion trades to 70–80%.
• If the market is in trend (for example, during a macroeconomic crisis, technological boom), increase the share of trend following trades to 70–80%.

4. Portfolio balance

Recommended distribution:

• 50–60% Mean Reversion. In this case, losses will only be in a strong trend.
• 40–50% Trend Following. Losses in flat due to a large number of false signals.

But it is worth understanding that this is not a strict rule, but a guideline. The ratio is better adjusted to your style, experience and current market conditions.

3 common mistakes when combining

• Using mean reversion in a trend and trend following in flat due to incorrect assessment of the market state.
• Too much variety of strategies: if you have 10 positions with different strategies, different timeframes and different assets, you will not be able to control the risk. Limit yourself to 3–7 positions.
• Lack of clear rules for each strategy: if you do not know when to use mean reversion and when trend following, you will enter chaotically. Use our checklist (later in the text) to control the situation.

When mean reversion does not work?

Despite the overall effectiveness, there are typical cases when the strategy does not work as it should. We partially mentioned them above in the text, but for greater convenience we will combine them into a general table.

Checklist before entering a trade

Before each use of mean reversion, go through this checklist:

• Trend checked? SMA(50) and SMA(200) horizontal or weakly directed? If strong trend, do not enter.
• Signal confirmed? At least 2 indicators match (for example, Bollinger + RSI + Z-score).
• Stop-loss set? Slightly below/above minimum/maximum, margin 5–10 points.
• Take-profit calculated? On the center line SMA 20, risk/reward ratio at least 1:1.5.
• Position size correct? Risk per trade no more than 1–2% of deposit.
• Total risk normal? All open positions do not exceed 5–10% of deposit.
• Diversification observed? You do not trade more than 1–2 positions on one asset with high correlation.

If the answer to any question is “no” – do not enter the trade. Wait for better conditions. It is better to wait a little than to lose the deposit.

FAQ

• What is Mean Reversion (return to the mean)?
  Mean Reversion is a financial theory stating that asset prices and historical returns ultimately return to their long-term average level after significant deviations.
• How does the Mean Reversion strategy work in trading?
  The strategy is based on the idea that when the price deviates too far from the average (overbought or oversold), it is highly likely to correct back. Traders buy on oversold and sell on overbought, expecting a return to the average.
• Which indicators are best to use for Mean Reversion?
  Main indicators:Bollinger Bands – deviation from the average (SMA),RSI – overbought/oversold,Z-score – standardized deviation,Stochastic – fast overbought/oversold,MACD – momentum and crossovers.
• What popular Mean Reversion strategies exist?
  The most popular:Bollinger Bands strategy: band breakout,RSI-based strategy: RSI ;30 for buy, ;70 for sell,MA crossover: moving average crossover,Z-score strategy: Z; -2 for buy, Z; +2 for sell.
• Is Mean Reversion suitable for stocks, Forex or ETFs?
  Yes, it works in many asset classes:stocks (AAPL, MSFT),Forex (EURUSD, GBPUSD),>ETFs and commodities (gold, oil).
• What is the difference between Mean Reversion and Trend Following?
  Mean reversion enters against the current price movement (expects correction), works in flat. Trend following enters in the direction of the trend (expects continuation), works in strong trend. At the same time, the approaches do not replace, but complement each other.
• How is Z-score calculated in the Mean Reversion strategy?
  Formula: Z = (P − μ) / σwhere P – current price, μ – average (SMA/EMA), σ – standard deviation. Z &lt; -2 = oversold (buy), Z +2 = overbought (sell).
• Is Python / code needed to implement Mean Reversion?
  For simple trading on manual indicators (Bollinger, RSI in TradingView/MetaTrader) code is not needed. For quantitative trading, testing on historical data and automation – Python is useful, but not mandatory for beginners.
• What risks and disadvantages do Mean Reversion strategies have?
  The price may continue to move away from the average for a long time and unexpectedly The average also moves – the price may “stall”, and the average “catches up” with it In a strong trend the strategy does not work, gives losses False signals with contradictory indicators
• Which timeframe is best suited for Mean Reversion?
  It works best on H1–D1 (intraday and swing trading). On M5–M15 there is too much noise, on W1–M1 corrections can take months.
• Can Mean Reversion be combined with other strategies?
  Yes, professionals combine mean reversion with trend following in a 50/50 or 60/40 ratio in the portfolio for risk diversification and stable returns in different market states.
• Does Mean Reversion work in real trading?
  Yes, it works. Many quantitative funds use mean reversion as part of their strategies. Testing on historical data shows positive returns with proper risk management, especially on assets with high liquidity (EURUSD, AAPL, gold).
• How to manage risks in the Mean Reversion strategy?
  Here are a few simple rules: Risk 1–2% of deposit per trade Stop-loss below minimum (long) / above maximum (short)Take-profit on center line SMA 20, risk/reward ratio at least 1:1.5Diversification: 3–5 different assets, total risk 10%
• When does the strategy not work?
  In a strong trend (SMA 50 significantly above/below SMA 200, both grow/fall). During important news due to high volatility.When the price “spreads” along the Bollinger band for several days.
• Is Mean Reversion suitable for beginner traders?
  Yes, it is suitable. This is one of the most understandable strategies for beginners: the logic is simple (bought cheap – sold expensive), indicators are visual (Bollinger, RSI), risks are controllable (with proper risk management).
• What is Pairs Trading and how is it related to Mean Reversion?
  \Pairs Trading is a strategy where you trade the price difference between two correlated assets (for example, AAPL vs MSFT). When the difference deviates from the average, you buy the “cheap” and sell the “expensive”, expecting the difference to return to the average – this is classic mean reversion.
• How to avoid false signals in Mean Reversion?
  Use confirmation from at least 2–3 indicators (for example, Bollinger + RSI + Z-score). Do not enter if signals contradict each other. Always preliminarily check the market state: flat or trend?Do not enter during important news.

Conclusion

Mean Reversion is one of the most proven and understandable strategies for beginner traders. Its essence is simple: the price tends to return to its average value after significant deviations. You buy when the asset is oversold, and sell when it is overbought, expecting a correction to the average. With proper risk management and selection of indicators, this strategy can become a stable source of profit in the long term.

However, it is important to remember: it does not work in a strong trend. Always check the market context before entering, use confirmation from several indicators and never risk more than 1–2% of the deposit on one trade. Combine mean reversion with trend following in the portfolio to make a profit in different conditions and reduce overall risks.

Want to try Mean Reversion in practice? Open a demo account on j2t.com and practice without risk to real capital. If you already feel confident, register a full trading account and start working on real markets. Good luck!