ES Mean Reversion: An Intraday Trading Plan Synthesized from 100+ Trading Books - Extended Summary

Author: Greeny | Categories: Day Trading, Mean Reversion, Futures, Trading Systems, ES Futures

About This Summary

This is the complete, unabridged trading plan for intraday mean reversion on ES (E-mini S&P 500) futures. It was created by synthesizing the collective wisdom of over 100 trading books into a single, actionable framework with minimum inputs and maximum clarity. Every principle in this plan has been cross-validated across multiple independent sources. This is not a theoretical exercise -- it is a ready-to-execute playbook. Full credit to the 63+ authors whose original work made this synthesis possible (see Sources section at the end).

PHILOSOPHY

Markets spend 70-80% of time in balance (trading ranges). Price that deviates from value attracts responsive participants who push it back. This plan exploits that tendency using only price, volume, and VWAP -- no exotic indicators.

Cardinal Rule: Mean reversion is ONLY traded in the direction of the larger trend or within confirmed balance. Never fade a one-timeframe trending market.

The plan is built on three foundational insights that appeared consistently across all 100+ books:

1. Mean reversion is the dominant market behavior -- but trend days, though less frequent (20-30% of sessions), produce outsized moves that can destroy a mean-reversion trader who doesn't recognize them
2. The hierarchy most traders invert -- position sizing and exits matter more than entries, yet 90% of traders spend 90% of their time on entries
3. Discipline is the edge -- the difference between profitable and unprofitable traders is not what they know, but whether they execute what they know consistently

INPUTS (Only 4)

No other indicators needed. Each additional indicator adds complexity, lag, and curve-fitting risk without proportional edge improvement (Tomasini/Jaekle, Kaufman). Montier's research shows that beyond 5-7 variables, additional data increases confidence without increasing accuracy.

Total free parameters: 4 (RSI entry threshold, RSI exit threshold, max hold time, daily loss limit). This is well within the robustness guidelines from Pardo and Tomasini/Jaekle, who recommend 2-4 parameters maximum for a system to resist overfitting.

PRE-MARKET ROUTINE (30 min before RTH)

1. Mark prior day's VAH, VAL, POC (Steidlmayer, Dalton)
2. Mark overnight high/low and overnight developing VA
3. Mark any naked POCs from prior sessions -- untested fair value levels that act as magnets (Dalton)
4. Check economic calendar -- do NOT trade mean reversion into FOMC, CPI, or NFP. Mean reversion assumes normal distribution of returns; news events produce fat-tailed moves that violate this assumption (Taleb)
5. Calculate daily max loss and per-trade position size (Elder's 2% rule, Platt's 3% rule via Schwager)
6. Note prior day's ATR -- if overnight range already exceeds 80% of it, expect a trend day
7. Check whether prior day's value area overlaps the day before's. If yes, bias toward mean reversion (Dalton)

SESSION FILTER: IS TODAY A MEAN-REVERSION DAY?

This is the most important decision of the day. Getting this wrong -- fading a trend day -- is the single fastest path to ruin for a mean-reversion trader. The transition from balance to trend is the most dangerous moment (Lowenstein/LTCM, Dalton).

Decide this in the first 30-60 minutes. Do not guess before the open.

Step 1: Classify the Opening (first 15 min)

Step 2: Measure the Initial Balance (IB = first hour's range, 9:30-10:30 ET)

Step 3: Quick Regime Checks

• Is price within or overlapping the prior day's value area? If yes, balance is likely. (Dalton)
• Is the profile developing symmetrically (bell-curve shape)? If yes, mean reversion is the right approach. (Steidlmayer)
• Is the market making single prints (one-timeframe directional move with gaps in the profile)? If yes, stop fading immediately. (Dalton)
• ADX below 20 and declining = mean reversion environment. ADX above 30 and rising = trend-following only. (Wilder, Connors/Raschke)
• Kaufman's Efficiency Ratio below 0.1 = mean-reverting. Above 0.3 = trending. (Kaufman)

If in doubt, stand aside. The cost of missing a trade is zero. The cost of fighting a trend is real.

Market Type Classification (Tharp)

ENTRY RULES

Where to Enter (Structural Levels -- Priority Order)

1. Prior day's VAH / VAL (Steidlmayer, Dalton)
2. VWAP, especially first pullback to VWAP after morning move (Aziz, Johnson)
3. IB High / IB Low (Dalton)
4. Prior day's POC or naked POC (Dalton)
5. Prior day's High / Low
6. 2nd standard deviation VWAP band -- extreme extension (Aziz)
7. Fibonacci cluster zones from prior session OHLC (Boroden)

The Value Area Rule (Steidlmayer)

One of the few mechanical mean-reversion applications endorsed by the Market Profile creator:

When to Enter (Entry Conditions)

Minimum 4 out of 6 conditions must be met. 5 or 6 = high conviction (A+ setup).

The Two-Reason Rule (Brooks): Never enter with fewer than two independent reasons. Examples:

• Price at VAL + RSI(2) below 10
• Price at outer Bollinger Band + volume climax at support
• Spring at prior day's low + delta divergence positive

The Three Highest-Probability Mean Reversion Setups

1. The Spring / False Breakdown (Wyckoff)

• The highest cross-reference frequency of any single setup across all 802 books in the library
• Price breaks below support (VAL, IB low, prior day's low), triggers stops, then immediately snaps back above
• Volume should be LOW on the false break (exhaustion, not initiative). Spring #3 (shallow penetration, low volume) is the highest reliability
• Enter on the bar that reclaims the broken level
• Stop: below the false break low + 1 tick buffer
• The trapped breakout traders' stops fuel the reversal (Brooks, Bulkowski)
• Bulkowski's data: busted patterns produce moves often 2x larger than the original expected breakout

2. VWAP Reclaim

• The 802-book compilation calls this "fundamentally a mean-reversion entry within a trending context"
• Price dips below VWAP on declining volume in a session that is otherwise balanced/bullish
• Price reclaims VWAP on expanding volume
• Enter on the close of the reclaim bar
• Stop: below the dip low
• Win rate approximately 60-70% per Aziz
• Works because institutional algorithms benchmark to VWAP, creating gravitational pull (Johnson)

3. Failed IB Extension

• Price extends beyond IB high or low but fails to follow through
• The extension bar closes back inside the IB range
• Enter in the direction back toward the IB midpoint
• Stop: beyond the extension extreme + 1 tick
• AMT logic: the market tested for OTF participation beyond the IB and did not find it (Dalton)

Additional Named Setups (from Street Smarts -- Connors/Raschke)

Turtle Soup / Turtle Soup +1: Fade breakouts beyond the 20-period high/low when ADX is declining. Enter on a stop back through the prior extreme. The prior extreme must have occurred at least 3 bars earlier (ensures "stale" enough to have attracted breakout traders). Win rate approximately 55-60%, R:R 1:1.5 to 1:2.

80-20 Setup: When ES opens in the top/bottom 20% of its daily range and closes in the opposite 20%, fade the move the next session if price exceeds the prior session's extreme and reverses. Textbook exhaustion/excess pattern. Win rate approximately 55-65%.

Momentum Pinball: 3-period RSI rate-of-change dropping below 30 as an oversold filter, then enter long above the IB high. Combines short-term oscillator exhaustion with IB breakout confirmation.

Wyckoff's Effort vs. Result -- The Master Volume Principle

For mean reversion: look for high volume / small price movement at the extreme. This signals that the counter-force is absorbing without conceding ground -- the setup for the snap-back.

Entry Execution

• Use limit orders at structural levels, not market orders. Every tick of slippage directly erodes the mean-reversion edge (Narang).
• If using order flow tools: confirm absorption (large passive orders consuming aggressive orders without price moving through) or delta divergence (price makes new extreme, delta does not confirm) before entry.
• When you see a front-loaded aggressive pattern (implementation shortfall algorithm), the price displacement is more likely permanent. Do NOT fade it. When you see metronomic or VWAP-style flow, the displacement is more likely temporary. Fade it. (Johnson)

VSA Signals That Precede Mean Reversion Entries (Coulling, Wyckoff)

• Stopping Volume: High volume on a down bar that closes in upper portion of range = institutional buying absorbing selling
• No Supply: Narrow spread, low volume, down close in downtrend = selling exhausted
• Shakeout: High volume sharp decline that reverses quickly = stop hunting, institutional accumulation

EXIT RULES

Scaling Protocol

Target Hierarchy

Bulkowski's distribution data across tens of thousands of patterns shows outcomes are front-loaded. Most of the gain from reversion plays comes quickly. Scaling out captures the high-probability portion.

Connors' RSI-Based Exit (Alternative)

Exit longs when RSI(2) crosses above 65-70. Exit shorts when RSI(2) crosses below 30-35. This lets the mean reversion play out fully rather than capping upside with arbitrary targets. Can be used instead of or in combination with structural targets.

Time Stop

If the trade has not moved 4+ ticks in your favor within 15 minutes, exit. A mean-reversion trade that stalls is sending information that the level may not hold. Capital sitting flat in a non-performing trade has an opportunity cost (Damir).

Hard Invalidation

If the level you faded breaks with conviction -- aggressive market orders overwhelming absorption, depth evaporating, single prints developing -- exit immediately. Do not wait for the stop. The market is transitioning from balance to trend.

End of Day

Close all positions before the cash close. Overnight gap risk is uncompensated for intraday mean reversion.

RISK MANAGEMENT

Position Sizing

Position Size = (Account Equity x 0.01) / (Stop Distance in Points x $50)

Example: $50,000 account, 1% risk = $500. Stop is 4 points ($200/contract). Size = 2 contracts.

Never risk more than 1% of account equity on a single trade. (Elder's 2% rule adapted to the more conservative 1% for futures)

Source hierarchy: Tharp found that position sizing -- not entries, not indicators, not market selection -- is the primary determinant of whether a system meets its objectives. Most traders spend 90% of effort on entries, which account for perhaps 10% of performance variance.

ATR-Based Volatility Adjustment (Wilder, Turtles)

ATR automatically widens stops in volatile sessions and tightens in quiet ones.

The Asymmetry of Loss Recovery

This non-linear relationship is the mathematical argument for small position sizes and strict daily loss limits (Damir, Kaufman, Elder).

Daily Loss Limits

Portfolio Heat (Elder)

• 2% Rule: Never risk more than 2% of account equity on a single trade
• 6% Rule: Never exceed 6% total account risk across all open positions
• All positions in correlated instruments (ES, NQ, YM, RTY) are effectively one position with different betas. Size accordingly.

Non-Negotiable Rules

1. Hard stops only. Live orders in the market, never mental stops. Loss aversion (Kahneman: losses felt 2x as intensely as equivalent gains) will talk you out of exiting.
2. Never widen a stop. "Every catastrophic loss in the compilation began as a small loss where the stop was moved." (802-book meta-analysis)
3. Never average down. It increases exposure AND psychological commitment to a losing thesis. LTCM was destroyed by this exact behavior at the macro scale (Lowenstein).
4. Never revenge trade. After a loss, the next entry must meet all the same criteria as any other.
5. Maximum 4-5 trades per session. More means quality standards have been relaxed.

Risk of Ruin (Eckhardt, via Faith)

The relationship is non-linear. Doubling position size from 5% to 10% only halves the time to double your money, but quadruples the risk of ruin.

Modified Kelly Criterion (Sinclair, Thorp via Schwager)

f* = edge / variance_of_payoff

Use half-Kelly or less in practice. Reasons:

• Edge estimation is uncertain; overestimating the edge leads to catastrophic overbetting
• Full Kelly can produce drawdowns exceeding 50%
• The Kelly criterion assumes infinite time horizon and log-utility, which do not match real trading constraints

Kaufman recommends 1/4 to 1/2 Kelly as upper bound.

Expectancy Formula (Penfold)

Expectancy = (Win Rate x Avg Win) - (Loss Rate x Avg Loss)

A mean reversion ES system should target:

• Win rate: 55-65%
• Avg win: 1.5-2.0 R
• Avg loss: 1.0 R
• Expectancy: +0.475R to +0.70R per trade

If you cannot demonstrate positive expectancy over 200+ trades (backtested or forward-tested), you do not have a system yet.

TIME-OF-DAY FRAMEWORK

The midday dead zone is "the single most commonly cited cause of profit erosion" across all 802 books in the library.

PSYCHOLOGY CHECKLIST

Before Every Trade (under 30 seconds)

1. State the thesis in one sentence: "Price is at [level], extended on RSI(2), with [confirmation]. Target is [level]."
2. Confirm the stop price.
3. Confirm position size is within risk limits.
4. Ask: "What would make me wrong?" -- that is where the stop goes.
5. Articulate the strongest argument AGAINST the trade in 15 seconds (Montier: confirmation bias defense).

Five Truths to Internalize (Mark Douglas, Trading in the Zone)

1. Anything can happen on any single trade. Your VA level can be steamrolled by a surprise headline.
2. You don't need to know what happens next to make money. You need to know the probability and manage accordingly.
3. There is a random distribution of wins and losses for any edge. You will have 5 losers in a row even with a 60% win rate.
4. An edge is just a higher probability, not a certainty. Mean reversion at VA extremes works more often than not, but "more often than not" means it fails regularly.
5. Every moment in the market is unique. The VA level that held yesterday may not hold today.

The Three Killers

The Drawdown Spiral (Tendler, Schwartz)

The specific cascade that kills mean-reversion traders:

1. Loss at a VA level triggers doubt about the level
2. Doubt causes hesitation at the next valid setup
3. Hesitation causes you to enter late or not at all
4. Missing a winner after taking a loser triggers frustration
5. Frustration triggers an impulse entry at a marginal level (revenge trade)
6. The revenge trade loses because the level was marginal
7. Now down 3 trades and emotional -- exactly when you should stop but exactly when most traders size up

Recovery protocol (Schwartz):

1. Stop trading for 1-3 days
2. Exercise intensely (reset cortisol levels)
3. Review journal without judgment -- identify process violations, not P&L
4. Reduce position size by 50-75%
5. Trade only A+ setups for 2-4 weeks
6. Gradually restore size as discipline is consistently maintained

Post-Win Danger (Schwartz)

Post-win overconfidence is MORE dangerous than post-loss fear. Schwartz's repeated observation: his worst losses followed his biggest wins. The mechanism: big win triggers dopamine, generates overconfidence, leads to increased risk-taking (sizing up, fading marginal levels, ignoring trend-day signals), and exposes you to outsized losses. Treat big wins as triggers for increased caution, not increased aggression.

Cognitive Biases Most Dangerous for Mean-Reversion Traders

Process Over Outcome

Judge every trade by whether the process was followed, not by P&L.

Physical Fitness as Edge (Schwartz)

Schwartz found a direct, observable correlation between his cardiovascular fitness and his P&L. Modern neuroscience confirms: cardiovascular exercise improves prefrontal cortex function (executive decision-making), reduces baseline cortisol, and accelerates recovery from acute stress. Sleep deprivation measurably degrades decision quality -- prioritize 7-8 hours. Front-load trading activity to the first 2-3 hours when both market volume and cognitive resources peak.

POST-SESSION JOURNAL

For every trade, record:

1. Date and time
2. Setup type (Spring, VWAP Reclaim, Failed IB Extension, Turtle Soup, 80-20, other)
3. Entry price, stop price, target price(s)
4. Actual exit price(s) and P&L
5. R multiple achieved (actual profit or loss / initial risk)
6. Screenshot of the trade
7. Was the process followed? (Yes/No)
8. If no, what was the violation?
9. Emotional state at entry (1-10 scale)
10. Day type that developed (balanced, trend, double-distribution)
11. One lesson

Weekly Review Questions

• What was my win rate this week?
• What was my average R per trade?
• Which setups performed best?
• Which time-of-day produced the best results?
• Did I violate any rules? How many times?
• Am I trading within my daily loss limits?
• Am I doing more of what works and less of what doesn't? (Steve Clark via Schwager)

SYSTEM VALIDATION FRAMEWORK

Before trading this plan live (Pardo, Kaufman, Tomasini/Jaekle):

1. Walk-Forward Analysis: Optimize on in-sample window, test on out-of-sample, slide forward, repeat. Concatenate all OOS results for true equity curve.
2. Parameter sensitivity: Plot performance across a range of parameter values. Robust systems show broad plateaus, not sharp peaks. If changing RSI threshold from 5 to 10 causes performance to collapse, the system is fragile.
3. Minimum 500+ trades in backtesting sample.
4. Degrees of freedom: Trades-to-parameters ratio should be 10:1 or better.
5. Monte Carlo simulation: 1,000-10,000 iterations. Examine 95th percentile worst-case drawdown.
6. Minimum thresholds: Profit factor > 1.5, Sharpe > 1.0, Recovery factor (Net Profit / Max Drawdown) > 3.0.
7. Paper trade 50+ trades before going live. Start with minimum position size.

"If your system has 10 parameters and 100 trades, you have not found a trading system. You have found a description of the past." (Kaufman)

WHAT TO IGNORE

• Financial news, opinions, social media during trading hours (Darvas: isolation is edge)
• More than 4-5 data inputs (Montier: additional data beyond 5-7 increases confidence without accuracy)
• Trades during news events (Taleb: mean reversion assumes normal distribution; news produces fat tails)
• Exotic indicators, oscillators, or proprietary systems
• Any trade that does not meet the minimum 4/6 entry criteria
• The midday dead zone (11:30-14:00 ET)
• The urge to "make back" a loss
• Trading educators who do not trade (Taleb: skin-in-the-game filter)

THE 10 LAWS OF CAPITAL DESTRUCTION (What NOT to Do)

From the 802-book meta-analysis, these are the most frequently cited causes of account destruction:

1. Never widen stops -- the #1 account killer across all books
2. Never average down -- increases exposure AND psychological commitment
3. Never trade the dead zone (11:30-14:00) -- your edge evaporates
4. Never revenge trade -- the self-reinforcing feedback loop that terminates accounts
5. Never counter-trend scalp without confluence -- requires >60% win rate just to break even
6. Never overtrade -- more than 4-5 trades/session means quality standards have been relaxed
7. Never ignore volume -- price without volume is an incomplete signal (Wyckoff, Coulling)
8. Never trade without a journal -- no data means no improvement
9. Never confuse ensemble probability with time probability -- "works 65% of the time" is irrelevant if one loss wipes you out (Taleb)
10. Never ignore regime changes -- a strategy that worked yesterday in balance will destroy you today in trend

THE HIERARCHY MOST TRADERS INVERT

This plan is designed to reflect the correct hierarchy. Entries are the simplest part. Risk management, exits, and discipline are where the edge lives.

QUICK REFERENCE CARD

PRE-MARKET:  Mark VAH/VAL/POC, ON hi/lo, naked POCs. Check calendar. Set daily max loss.
OPEN:        Observe 30 min. Classify opening type. Measure IB at 10:30.
FILTER:      IB normal + balanced profile = mean reversion ON.
             IB narrow + directional open = stand aside.
ENTRY:       Price at structural level + RSI(2) extreme + declining volume/absorption + reversal candle.
             Minimum 4/6 conditions. Limit orders only.
STOP:        Beyond the level that defines the trade. Hard stop. Never widen.
TARGET 1:    POC/IB midpoint. Take 50% off. Move stop to breakeven.
TARGET 2:    VWAP or opposite VA boundary. Exit remainder.
TIME STOP:   15 minutes of no progress = exit.
MAX TRADES:  4-5 per day.
MAX LOSS:    3 consecutive stops = done. 3% of equity = done.
POST-CLOSE:  Journal every trade. Rate discipline 1-10. Review weekly.

SOURCE BOOKS & AUTHORS

This plan synthesizes the work of the following authors and titles, who deserve full credit for the original ideas:

Plus ~40 additional titles from the full 802-book library covering market history, behavioral finance, portfolio theory, and trading biography that informed the cross-validation of principles.

"The cost of missing a trade is zero. The cost of fighting a trend is real."