Detailed Outline
Title: Technical Analysis for Futures Traders: A Quant-Desk Perspective on Liquidity, Order Flow, and Systematic Edge Working title, refine before publication. Alternates considered: The Futures Trader's Stack, Reading the Futures Tape, Liquidity, Flow, and Structure.
Length target: 90,000–120,000 words across 21 chapters and 4 appendices. Audience: aspiring prop / desk traders; systematic researchers; experienced discretionary futures traders upgrading to institutional frameworks. Difficulty arc: beginner-accessible Part I, intermediate-rigorous Parts II–III, advanced/quant Parts IV–V.
Front Matter
- Preface: Why this book exists. What the gap is between retail TA and what futures desks actually use. Honest disclaimers about uncertainty, regime dependence, and the limits of TA. Reading-paths for three audience types.
- How to use this book: Recommended order, "skim if experienced" markers, exercise pacing, links between chapters.
- A note on data and platforms: CME data sources, RTH vs. ETH conventions, why we use unadjusted contracts for level work and continuous contracts for trend work, supported platforms (Pine, NinjaTrader, ATAS, Sierra Chart, Python with Databento/CME tickers).
PART I: FOUNDATIONS (Beginner)
Chapter 1: What You Are Actually Trading
Abstract. Before any chart appears, you must understand that a futures contract is a leveraged, dated, exchange-cleared agreement, not a stock. Tick value, contract size, expiry/roll, basis, and the RTH/ETH session split fundamentally change how everything in this book behaves. We walk through the canonical contracts (ES, NQ, GC, CL), the math of margin and leverage, and why "back-adjusted" charts lie about absolute price.
Core concepts. Contract specs · tick / point / handle · margin (initial, maintenance, intraday) · roll mechanics and the front-month convention · basis and term structure · RTH vs. ETH · settlement vs. close.
Key takeaways. (1) Tick value drives stop placement and sizing far more than chart aesthetics. (2) The RTH session is the institutional session; treat ETH as a different market. (3) Continuous-contract charts distort absolute-level analysis, know when to switch to unadjusted contracts.
Exercises. Compute the dollar value of a 1-σ daily move on ES, NQ, GC, CL using the prior 20 sessions. Compute the basis between the front and second-month ES contract over the past 5 days. Identify the most recent roll date for each of the four contracts.
Chapter 2: Market Structure: Trend, Range, and Regime
Abstract. Market structure is the chassis of every setup that follows. Higher-highs / higher-lows is not mysticism, it is a tractable, falsifiable language for describing price. We define swing structure formally, classify regimes (trend, range, volatility), and build a composite regime score the rest of the book leans on.
Core concepts. Swing pivots and the ZigZag formalization · trend / range / squeeze classification · ATR percentile, Bollinger Band Width percentile, Kaufman Efficiency Ratio, ADX · the regime composite · volatility regimes (calm, normal, expanded, crisis).
Key takeaways. (1) The regime determines which tool is applicable, full stop. (2) No single indicator classifies regime reliably, composites are mandatory. (3) Roughly 15% of sessions are unclassifiable; those are skip days, not "trade smaller" days.
Exercises. For ten recent ES sessions, manually classify each as Trend-Up / Trend-Down / Range / Squeeze using the composite criteria. Compare to the indicator output. Note where your eye and the indicator disagree.
Chapter 3: Liquidity and Order Flow Primer
Abstract. This chapter is the bridge from "what every retail trader thinks TA is" to "what an institutional trader actually watches." We introduce the liquidity primitives, pools, equal highs/lows, fair-value gaps, sweeps, and the order-flow primitives, delta, CVD, footprint, absorption, and we explain why setups built on these primitives dominate setups built on classical indicators alone.
Core concepts. Liquidity pools · equal H/L · round-number magnets · fair-value gap (FVG) · stop run / liquidity sweep · delta and Cumulative Volume Delta · footprint chart · stacked imbalance · absorption · BVC (Easley/López de Prado/O'Hara) as the inference engine.
Key takeaways. (1) Liquidity is not where price is; it is where stops are parked. (2) Sweeps without order-flow confirmation are noise. (3) Delta/CVD are powerful as confirmation, useless as triggers.
Exercises. On a recent NQ session, mark every equal-low and equal-high. Annotate which were swept. Of the swept, which reversed cleanly within 10 bars? Note the pattern.
PART II: CLASSICAL TECHNICAL ANALYSIS (Intermediate, with critical lens)
Chapter 4: Price Action: Swings, Pivots, Highs and Lows
Abstract. Pure price action, without indicators, is the most data-efficient way to read a chart. We treat swings formally, define the ATR-conditioned ZigZag, and walk through how pivot points compose into trend channels and structure breaks.
Core concepts. Swing high / swing low · pivot strength · break of structure (BoS) and change of character (CHoCH) · trend channel construction · the failed-swing pattern.
Key takeaways. (1) Most retail "swing" labels are subjective; the ATR-conditioned definition makes them mechanical. (2) BoS without a CHoCH is a continuation, not a reversal. (3) Failed swings at structure are some of the highest-quality reversal signals in TA.
Exercises. Implement (or pseudocode) an ATR-conditioned ZigZag on closing prices for ES 5-min, March 2026. Count swings. Compare to a naive percent-based ZigZag.
Chapter 5: Support, Resistance, and the Quality of a Level
Abstract. Not all levels are equal. We build a quality score for any horizontal level, baseWeight × confluence × age-decay × touch-quality, and discuss why this maps to how desks actually rank levels on a chart.
Core concepts. Pivot levels · prior-day H/L/C, prior-week H/L · prior-month extremes · POC / VAH / VAL · naked POCs · round-number levels (boost only) · clustering / confluence · age decay · sweep history.
Key takeaways. (1) A "level" is a cluster of types, not a single line. (2) Sweep-then-defend is structurally bullish-for-the-defender. (3) Round numbers don't matter alone; they matter when they coincide with two other source types.
Exercises. For one current ES day session, list every level type within ±1.5 × daily ATR of close. Cluster them by ATR proximity. Score each cluster. Compare to the levels actually defended in the next session.
Chapter 6: Trend Tools: Moving Averages, MACD, and ADX
Abstract. Trend indicators are useful as regime filters and dynamic deviation references, not as crossover signal generators. We review the math, dismiss the crossover-system mythology with the relevant published evidence, and rebuild MAs as a context layer.
Core concepts. SMA vs. EMA · Wilder smoothing · MACD as a derivative of EMA difference · ADX as direction-blind trend strength · the 200-day MA as a regime filter.
Key takeaways. (1) MA crossovers in liquid futures, traded as standalone signals, do not beat slippage on intraday horizons (Faber 2007 ensemble case is portfolio-level, not single-instrument). (2) Distance from a long MA is a useful conditioning variable. (3) ADX > 25 is a necessary but not sufficient condition for trend setups.
Exercises. Backtest a 9/21 EMA crossover on ES 15-min from 2020–2025 with realistic slippage. Compare the equity curve to a "trade only in the direction of the 200-period EMA" filter applied to the same setup.
Chapter 7: Oscillators: RSI, Stochastics, and the Mean-Reversion Trap
Abstract. Oscillators are bounded transformations of price momentum. They tell you what regime would be appropriate to assume, they do not tell you which regime is in effect. The chapter dismantles the standard "RSI > 70 is overbought" usage and rebuilds oscillators as conditional signals.
Core concepts. RSI · Stochastics · CCI · Williams %R · classical and hidden divergence · oscillator failure swings · the regime-conditioning rule.
Key takeaways. (1) "Overbought" in a trend day extends; "overbought" in a range fades. The oscillator does not tell you which. (2) Divergences are useful only at structure, never mid-range. (3) The hidden divergence (continuation) is more reliable than classical divergence (reversal).
Exercises. Mark every RSI 14 > 70 reading on NQ 5-min during the most recent five trend days. How many faded? Repeat on five range days. Note the asymmetry.
Chapter 8: Volatility: ATR, Bollinger Bands, and the Squeeze
Abstract. Volatility is the measurement underlying every other indicator's interpretation. ATR is your sizing unit; Bollinger Bands are your normalised deviation envelope; the squeeze is the most robust precondition for an expansion move that retail TA recognises.
Core concepts. ATR (true range derivation, Wilder smoothing) · Bollinger Bands (μ ± 2σ on close) · BBW (Band Width) percentile · the squeeze (BBW < 20th-percentile rolling) · Yang-Zhang as the right volatility estimator for futures.
Key takeaways. (1) ATR is not a signal, it is a unit. (2) Squeeze + directional break in a trend regime is one of the cleanest setups in this book. (3) Yang-Zhang preserves opening-jump information; close-to-close volatility throws it away.
Exercises. On GC daily for the last 12 months, plot BBW. Mark every reading below the 20th percentile. Which resolved into expansions ≥ 1.5 × ATR within 10 sessions?
PART III: INSTITUTIONAL FRAMEWORKS (Advanced)
Chapter 9: Volume Profile and Market Profile (Dalton)
Abstract. Market Profile is the single most important framework Dalton bequeathed to non-floor traders. We walk through TPO construction, the value-area derivation, day-type classification, and the multi-day composite profile.
Core concepts. TPO (Time Price Opportunity) · POC (Point of Control) · Value Area (70%) · IB (Initial Balance) · day types (Trend, Normal, NV, Neutral, Non-Trend, Double Distribution) · single prints · naked POC · composite (multi-day) profile.
Key takeaways. (1) The value area is where institutional volume agreed; extremes are where one side rejected. (2) Day-type classification by 11:00 ET is the highest-value daily decision a profile trader makes. (3) Naked POCs retest with high empirical frequency (Dalton: ~75% within 30 sessions on ES), they are durable magnets.
Exercises. Build a TPO profile manually for one ES RTH session. Identify the POC and VA edges. Compare to the platform-rendered profile. Note any divergences and their cause.
Chapter 10: VWAP and Anchored VWAP
Abstract. VWAP is not a trading "indicator", it is the institutional execution benchmark. We define VWAP rigorously, derive volume-weighted standard-deviation bands, and develop the anchored-VWAP family (session, week, month, prior-settlement, event-anchored).
Core concepts. Session VWAP · weekly / monthly VWAP · anchored VWAP from arbitrary pivot · volume-weighted σ bands · multi-AVWAP confluence · the prior-settlement-VWAP retest pattern.
Key takeaways. (1) AVWAP from a clear event (FOMC, CPI, prior settlement) is a magnet because real participants reference it. (2) Reversion-to-VWAP is statistically real intraday in balanced ES regimes; it breaks down on trend days. (3) Multi-AVWAP confluence (session AVWAP intersecting with prior-settlement AVWAP) marks high-probability reaction zones.
Exercises. Anchor a VWAP from the most recent FOMC release on ES. Track its retest history over the next 10 sessions. Note the conditions under which retests held versus broke.
Chapter 11: Order Flow: Footprint, Delta, CVD
Abstract. Order flow is the deepest screen-trader-accessible layer. We treat the four primitives, delta, CVD, footprint, absorption, and the inference engine (Bulk Volume Classification) that powers most platforms' delta in the absence of true L2.
Core concepts. Per-bar delta · CVD construction · CVD divergence at swing high/low · footprint cells · stacked imbalance (3+ levels at 2:1+ skew) · absorption signature · BVC: Easley, López de Prado, O'Hara (2012).
Key takeaways. (1) Order flow is the cleanest measurable proxy for institutional aggression, but only at structure. (2) Footprint signals during news are mechanical, not informational. (3) Stacked imbalances at PDH/PDL are textbook continuation/reversal triggers.
Exercises. On a recent NQ session, find a CVD divergence at a swing high during RTH. Was it confirmed by order-flow absorption? Did the price reverse within 10 bars? Repeat the analysis at a swing high not coinciding with structure. Note the difference in outcome distribution.
Chapter 12: Liquidity: Pools, Sweeps, Imbalances
Abstract. Liquidity is the substrate beneath price. Stops cluster predictably; institutions know this; the sweep-and-reverse pattern is, at its core, an institutional liquidity-harvest. We codify it.
Core concepts. Liquidity pool taxonomy (PDH/PDL stops, equal H/L stops, round-number stops, IB extension stops) · the sweep mechanic (pierce ≥ 0.10–0.30 × ATR + close-back-inside) · velocity gating · the FVG (fair-value gap) · imbalance-fill probability.
Key takeaways. (1) Sweep + close-back-inside + order-flow confirmation = high-quality reversal trigger. (2) FVGs in low-liquidity contexts are not edges. (3) The same sweep mechanic in a strong trend regime continues, regime always wins.
Exercises. Identify the three most recent sweeps of equal-highs on ES intraday. For each, classify regime, note order-flow agreement, and tabulate outcome (reverse within 10 bars: yes/no).
Chapter 13: The Open: Open Type, IB, and Day Structure
Abstract. The open is the highest-information event of the trading day. Dalton's open-type classification, Open Drive, Test Drive, Auction, Auction in Range, Rejection-Reverse, is the fastest playbook selector available.
Core concepts. Open Drive (immediate one-directional move from open price) · Test Drive (probe other side then reverse) · Open Auction (around prior close, slow) · Rejection-Reverse · Initial Balance and IB extension probabilities · the 10:00 / 10:30 ET reassessment.
Key takeaways. (1) Open Drive in the direction of overnight bias resolves into a trend day with high probability. (2) Open Auction with no IB extension by 11:00 ET is a strong range-day signal. (3) Rejection-Reverse is the highest-volatility open type and the easiest to mis-trade.
Exercises. Classify the open on every ES RTH session for the past two weeks. Tabulate which open type produced which day type. Note your hit rate.
Chapter 14: Cross-Market Context: VIX, DXY, Yields, Correlated Futures
Abstract. No futures contract trades in isolation. The chapter develops the cross-market context map, VIX for risk regime, DXY for dollar regime, US10Y yields for risk-free rate, correlated futures (NQ/ES correlation, GC/DXY anti-correlation, CL/risk-on coupling).
Core concepts. VIX percentile · DXY direction · US10Y trend · 30-day rolling correlation between contracts · risk-on / risk-off classification · the divergence flag (when expected correlations break).
Key takeaways. (1) VIX > 80th percentile is an automatic size-down regime. (2) DXY direction calibrates the GC trade. (3) When ES and NQ decouple, treat the divergent move with suspicion.
Exercises. Build a one-screen cross-market dashboard (VIX %ile, DXY 20-day return, US10Y level, ES–NQ correlation) and tag every day for the past month with a regime label.
PART IV: SYSTEMATIC EDGE & RISK (Advanced)
Chapter 15: From Discretionary to Systematic
Abstract. Discretionary traders make more decisions than they realise. Systematizing forces those decisions into the open, where they can be tested. We walk through the discretionary-to-systematic translation pipeline: rule extraction, parameterisation, walk-forward validation.
Core concepts. Decision-rule extraction · parameter sensitivity analysis · walk-forward (anchored, rolling) · in-sample / out-of-sample protocol · selection bias · the Deflated Sharpe Ratio (Bailey & López de Prado, 2014).
Key takeaways. (1) Most "discretionary edges" do not survive parameterisation; that is the diagnostic, not the failure. (2) Walk-forward beats single-split backtest. (3) DSR corrects for the hundreds of strategies a researcher implicitly tested before writing the paper.
Exercises. Take one discretionary setup from your own journal. Parameterise it. Run a walk-forward backtest with 5 train/test splits. Report the DSR.
Chapter 16: Statistical Validation: Avoiding the Curve-Fit Trap
Abstract. The most common failure mode in systematic TA is overfit. We treat the four pillars of validation: (1) walk-forward, (2) parameter robustness, (3) regime stratification, (4) multiple-testing correction.
Core concepts. Bias-variance tradeoff in indicator parameter selection · parameter heatmap diagnostics · regime-stratified Sharpe · Bonferroni / Holm correction · the Probability of Backtest Overfitting (PBO) · combinatorially symmetric cross-validation (CSCV).
Key takeaways. (1) A robust strategy degrades gracefully under parameter perturbation. (2) Regime-stratified Sharpe reveals strategies that are entirely powered by one regime. (3) PBO > 0.5 means the in-sample Sharpe is uninformative.
Exercises. For an existing strategy, build a 2D parameter heatmap of Sharpe ratios. Identify whether the chosen parameters sit on a plateau or a peak. Re-run the strategy on the next-best parameter cell. How much does the Sharpe move?
Chapter 17: Risk Management for Futures
Abstract. Position sizing, stop placement, and the daily-loss cap are the only edges that compound across regime changes. We treat each rigorously, with the futures-specific math (tick value, leverage, fat tails) made explicit.
Core concepts. Per-trade risk fraction · per-session loss cap · Kelly and fractional Kelly · ATR-based stop sizing · structural stop placement · position-size formula given (account, risk %, stop distance, tick value) · max-correlated-position-cluster cap.
Key takeaways. (1) Per-session loss cap is a hard rule, not a guideline; below it you fight to stay in the game, above it you can recover. (2) Stops belong behind structure, never at round numbers. (3) Sizing must be a function of realized recent volatility, not nominal risk.
Exercises. Given a $50K account, 1% per-trade risk, ES at 6,200, recent 14-day ATR = 60 points, structural stop = 8 points: compute contract size. Repeat with stop = 25 points. Note how the contract count drops, not the dollar risk.
Chapter 18: Execution: Slippage, Spread, Queue, Fill Quality
Abstract. A backtest at the mid is a fantasy. We treat execution realism: slippage models, market vs. limit vs. stop-limit, queue position, the cost of being wrong on a fast-mover, and how execution quality interacts with strategy frequency.
Core concepts. Bid-ask spread by product · slippage models (constant tick, ATR-conditioned, volume-conditioned) · partial fills · queue priority on a price level · the round-trip cost as a fraction of edge.
Key takeaways. (1) Slippage destroys high-frequency mean-reversion strategies first. (2) On news, your "stop" is the next available bid/ask, not the price you set. (3) Queue position on a passive limit at structure is a real edge, and a perishable one.
Exercises. Estimate round-trip cost (commission + slippage + spread) per ES contract on a 30-trade-per-day strategy versus a 3-trade-per-day strategy. Annualise. Note the magnitude.
PART V: CASE STUDIES & FAILURE MODES
Chapter 19: When TA Fails: Regime Shifts, News, Liquidity Crises
Abstract. Cataloguing the failure modes is half of the edge. We walk through the historical failures (Feb 2018 vol-mageddon, March 2020 COVID, Aug 2024 yen carry), the structural reasons TA broke, and what could have saved a disciplined trader (regime detection, hard caps, news blackouts).
Core concepts. Regime-shift detection · news-blackout protocol · liquidity-crisis signatures · the difference between "the trade is wrong" and "the regime is wrong."
Key takeaways. (1) The trades that ruin you are the ones you took because the chart "looked the same as it did last week." (2) A hard time-window blackout around tier-one news (FOMC, NFP, CPI) saves more accounts than any indicator improvement. (3) When liquidity inverts, every backtested edge inverts with it.
Exercises. Pull the ES tape for August 5, 2024. Annotate the 90 minutes around the gap-down open. Which conventional setups would you have taken? Which would have blown up? What rule would you add to prevent that?
Chapter 20: Worked Examples: NQ Open Drive, ES Trend Day, GC Range, CL News
Abstract. Four full-session walkthroughs, each with screenshots, footprint cells, AVWAP overlays, and minute-by-minute commentary. Each example is chosen to illustrate a different setup family in its native habitat.
- §20.1 NQ Open Drive trend day (ideal trend continuation playbook).
- §20.2 ES range day with VAH/VAL fade (mean reversion at structure).
- §20.3 GC slow grind with multi-AVWAP confluence (institutional accumulation).
- §20.4 CL news-driven sweep and reverse (event-driven liquidity harvest).
Key takeaways. (1) The same indicator stack reads differently in each environment, that is the whole point. (2) The decision points (entry, scaling, exit) are mechanical when the regime is correctly classified. (3) Most retail traders' losing trades are not "wrong analysis", they are "right analysis, wrong regime."
Exercises. For each of the four sessions, write a one-page post-mortem in your own words. Compare your reasoning to the chapter's annotations. Note divergences.
Chapter 21: Putting It All Together: The Daily Plan
Abstract. A condensed daily workflow that operationalizes everything: pre-market checklist, open-type classification, regime tagging, watch-level construction, trade-plan template, post-session journal protocol.
Core concepts. Pre-market checklist (overnight range, key economic events, cross-asset context) · Open-type classification by 09:45 ET · regime tag · watch-level priority · trade-plan template fields · journal tag schema · weekly statistical review.
Key takeaways. (1) The plan is the edge. The execution of the plan is the discipline. (2) A statistical review is what converts intuition into calibrated belief. (3) If you cannot articulate the trade in one paragraph before entry, you do not have a trade.
Exercises. Run the daily-plan template for one full trading week. Submit each completed form to your journal. At week's end, score yourself against the plan, not against P&L.
Appendices
Appendix A, Mathematical Reference
- Wilder smoothing derivation
- Yang-Zhang volatility estimator
- BVC (Bulk Volume Classification) formula
- Volume-weighted standard deviation
- Kelly formula and fractional Kelly
- Deflated Sharpe Ratio
Appendix B, Empirical Studies
The studies opened in §11 of the Research Summary, executed and reported.
Appendix C, Platform Notes
Pine Script v6 quirks, NinjaTrader, ATAS, Sierra Chart, Python (Databento), what each does well, what each gets wrong, and what to avoid.
Appendix D, Annotated Bibliography
Primary sources organized by topic (microstructure, market profile, technical analysis, risk management, statistical validation), each with a one-paragraph annotation on what to read it for and what to skip.
Cross-cutting design choices
- Every chapter ends with Key Takeaways (3 bullets max) and Exercises (1–3 short prompts). The exercises are designed for paper or live screens; nothing requires proprietary data.
- Every chapter that proposes a tool or rule lists its failure conditions explicitly. This is non-negotiable.
- Visual diagrams are described in
Visuals/Diagram_Concepts.mdso they can be designed independently. The text always reads coherently without them. - Code is sparing. Where it appears, it is in pseudocode or in a single supported language (Pine v6 for indicators, Python for statistical work). Implementation language is not the bottleneck; understanding is.
Drafting status
| Section | Status | Words |
|---|---|---|
| Front matter (Preface, How to use, Data note) | Drafted | ~1,500 |
| Ch 1: Foundations | Drafted | ~4,500 |
| Ch 2: Market Structure | Drafted | ~3,700 |
| Ch 3: Liquidity & Order Flow Primer | Drafted | ~4,300 |
| Ch 4: Price Action | Drafted | ~3,100 |
| Ch 5: Support, Resistance, Level Quality | Drafted | ~3,300 |
| Ch 6: Trend Tools | Drafted | ~2,900 |
| Ch 7: Oscillators | Drafted | ~2,600 |
| Ch 8: Volatility | Drafted | ~3,000 |
| Ch 9: Volume Profile / Market Profile | Drafted | ~3,400 |
| Ch 10: VWAP and Anchored VWAP | Drafted | ~3,100 |
| Ch 11: Order Flow Detailed | Drafted | ~3,100 |
| Ch 12: Liquidity Detailed | Drafted | ~3,600 |
| Ch 13: Open Type and Day Structure | Drafted | ~2,800 |
| Ch 14: Cross-Market Context | Drafted | ~2,600 |
| Ch 15: Discretionary to Systematic | Drafted | ~2,900 |
| Ch 16: Statistical Validation | Drafted | ~2,700 |
| Ch 17: Risk Management | Drafted | ~3,300 |
| Ch 18: Execution | Drafted | ~2,900 |
| Ch 19: When TA Fails | Drafted | ~3,100 |
| Ch 20: Worked Examples | Drafted | ~3,800 |
| Ch 21: Daily Plan | Drafted | ~3,200 |
| Appendix A: Mathematical Reference | Drafted | ~2,200 |
| Appendix B: Empirical Studies | Drafted | ~2,100 |
| Appendix C: Platform Notes | Drafted | ~1,900 |
| Appendix D: Annotated Bibliography | Drafted | ~2,400 |
| Research Summary (companion doc) | Drafted | ~3,400 |
| Trading Frameworks (companion doc) | Drafted | ~4,200 |
| Visual / Diagram Concepts (companion doc) | Drafted | ~3,600 |
| TOTAL | First draft complete | ~90,000+ |
First-draft pass complete across all 21 chapters, 4 appendices, front matter, and three companion documents. Subsequent passes (when the user directs) will revise for tone, evidence completeness, exercise quality, internal consistency, and any reader-validation feedback.