Gap Trading Study Guide & Workbook: A Complete Companion to Peter L. McGrew’s Book

Book: Gap Trading to Profit in 7 Days
Author: Peter L. McGrew
Difficulty: Intermediate

The Big Idea

A price gap is an observation. Turning it into a trading decision requires a definition, a hypothesis, an execution plan, and evidence. This workbook helps you work through the book’s four chapters while building those skills. Keep your copy beside you: reading assignments refer to the supplied 129-page PDF, counting its contents page as page 1. Pagination in other editions may differ.

The explanations, fictional market examples, exercises, answer keys, worksheets, and experimental code below are original CandleChasers teaching material. The seven sessions are a study schedule, not a timetable for becoming profitable. Take longer wherever the calculations or chart work need practice.

If you remember only one thing: define a setup before seeing its outcome, and evaluate the trade you could actually execute.

Seven Sessions, Seven Concrete Outputs

Read first, attempt the exercise with the answers closed, then compare your reasoning with the key. Use historical replay or paper observations for practical work. Do not treat completing an exercise as evidence that a strategy works.

Chapter 1 Companion: Identify and Measure the Gap

Read: Chapter 1, pp. 6–35. Demonstrate: distinguish an opening jump from a gap between completed trading ranges, calculate the opening percentage, and choose a fill reference.

Two comparisons that should not share an undefined label

The chapter introduces gaps through changes between trading periods and visual spaces on charts. Those descriptions can refer to different measurements. An opening gap compares today’s open with the previous close. A gap between completed daily ranges compares today’s low with yesterday’s high, or today’s high with yesterday’s low.

An outside opening can later trade back into the prior range. You cannot know at 9:30 whether the entire completed daily candle will remain separated. A backtest that uses today’s final low to select an opening trade uses information from the future.

Opening gap % = 100 × (current open − previous close) ÷ previous close. The sign gives direction. Use the absolute value only when intentionally studying magnitude without direction.

Worked example: opening gap, overlapping ranges

Fictional Alder Labs closed at $40. Its previous high was $41 and low was $38. The next session opened at $42, traded between $40.50 and $43, and closed at $42.50. Its opening gap was 5%. It opened above the prior high, but the completed daily ranges overlap between $40.50 and $41.

For this workbook, a full opening-gap fill means reaching the previous close. Alder did not reach $40, so that event did not occur. It did reach yesterday’s range. These are separate outcomes.

Clarifications to write in your margin

A wick records an extreme beyond the candle’s body; it is not inherently a false gap. A closing-price line chart usually connects closes and cannot by itself establish the opening gap. Chart session settings also matter: a regular-session opening gap does not prove there were no overnight trades. FINRA’s extended-hours disclosure explains that trading outside regular hours can have lower liquidity, higher volatility, and wider spreads.

The book’s discussion sometimes shifts between gap boundaries and support/resistance. Record the exact price level under consideration. A previous close, previous high, and current open can all be different prices; calling each one “the gap” makes a plan ambiguous.

Exercises 1–4

1. Calculate and classify. Previous close $80, high $82, low $77; current open $84, high $85, low $81. Find the opening gap percentage. Was the opening outside the previous range? Are completed ranges separate? Was the previous close reached?
2. Reverse direction. Previous close $50, low $48; current open $47, high $49, low $45. Answer the same four questions for this gap down.
3. Catch the look-ahead. A researcher buys at the open only when the day’s final low stays above yesterday’s high. What information is unavailable at entry? Rewrite the eligibility rule.
4. Write a definition. Specify your market, session, timezone, gap formula, fill reference, and observation deadline. Explain how extended-hours settings could change what you see.

Reading reflection: Find two places in Chapter 1 where a price boundary needs a more precise name. Rewrite each in your own notes and explain how the clarification changes a calculation.

Chapter 2 Companion: Classify Without Hindsight

Read: Chapter 2, pp. 36–63. Demonstrate: separate what is visible now from what later price action reveals.

The chapter discusses common, breakaway, continuation/runaway, and exhaustion gaps, with additional classification dimensions. These labels mix observations with interpretations. Direction is immediately measurable. Whether a move marks the beginning, middle, or end of a lasting trend often becomes clearer later.

Use two labels, recorded at two times

Before revealing later bars, record candidate classification and the evidence supporting it. After a fixed observation window, record outcome classification. A candidate that fails remains in your sample. Otherwise “breakaway gaps work” can become circular: successful moves receive the name while failures disappear.

There is a wording inconsistency in Chapter 2’s exhaustion discussion: it describes gaps in the prevailing direction but also calls them opposite-direction moves. Later examples describe a trend-direction push followed by failure. For our worksheet, write the sequence explicitly: pre-existing uptrend, upward gap, then observed failure. An upward gap alone does not establish exhaustion.

Make context measurable

These are our possible research definitions, not the author’s universal thresholds: “range breakout” means opening above the highest high of the previous 20 completed sessions; “uptrend” means the previous close exceeds its 20-session moving average and that average exceeds its value five sessions earlier. “High opening volume” could mean first-five-minute volume divided by average first-five-minute volume over the previous 20 sessions exceeds 1.5.

Compare matching time intervals. The first five minutes of volume divided by an average full day’s volume answers a different question. Neither high volume nor a rising candle identifies particular traders’ intentions or guarantees follow-through.

Exercises 5–8

1. Candidate versus conclusion. A stock opens above a 20-session range after an earnings announcement. It has not completed its first candle. What can you label now, and what remains unknown?
2. Spot circular evidence. An analyst collects only gaps followed by ten rising days and calls them continuation gaps. Why can this not establish the chance of continuation at entry?
3. Normalize volume. Today’s first five minutes trade 240,000 shares. The average of the previous 20 opening five-minute bars is 160,000. Calculate the ratio. Does it prove buying by institutions?
4. Blind chart task. Hide future bars on five historical examples. Record your candidate label and two observable reasons. Reveal exactly five sessions and record what changed. Keep disagreements and failed candidates.

Chapter 3 Companion: Turn a Setup Into a Specification

Read: Chapter 3, pp. 64–104. Demonstrate: distinguish continuation from reversal trades and write rules another reader can apply.

The chapter moves among gap-and-go, gap-fill reversals, opening reversals, and inside-range openings. It also discusses catalysts, opening ranges, VWAP, volume, and supply/demand context. Treat each as a separate hypothesis before combining filters.

Three trades can begin with the same gap up

A fill occurring before your entry is not the same strategy as trading toward the fill. For short examples, define the protective stop above the entry or other chosen invalidation level; do not mechanically reuse a long trade’s stop placement.

Exercises 9–12

1. Name the trade. A stock gaps up, falls to yesterday’s close, then rises above a completed recovery candle. You buy that recovery. Is this a fade, continuation breakout, or rebound-after-fill hypothesis?
2. Recalculate after the fill. A long signal closes at $60. Actual fill is $60.40, stop $59.20. What is actual price risk? What is the 2R target from that fill?
3. Audit a short. A short fills at $30 with proposed protective buy-stop at $29. Explain the problem. Suggest a hypothetical invalidation price above entry and calculate a 2R price target.
4. Write a no-trade decision. A candidate meets the chart trigger, but no reliable opening price or previous close is available. What should the research record say? May you estimate convenient prices from the screenshot?

Chapter 4 Companion: Size, Execute, and Review

Read: Chapter 4, pp. 105–129. Demonstrate: distinguish exposure from planned loss, account for costs, stress-test fills, and separate execution quality from outcome.

Size follows the price distance

For ordinary shares in a matching account currency, a simplified calculation is: quantity = floor((cash risk budget − fixed estimated costs) ÷ (entry-to-stop distance + estimated variable cost per share)). Apply an independent cash/exposure limit too. Futures, options, foreign currencies, and leveraged products need their own contract and conversion calculations.

Suppose a fictional account contains $20,000 and the teaching risk budget is $100. Planned entry is $40, stop $39.20, estimated round-trip variable costs are $0.04 per share, and fixed costs are zero. Quantity is floor($100 ÷ $0.84) = 119 shares. Exposure is $4,760; modeled loss including estimated costs is $99.96. The 0.5% budget is an illustration, not a sizing recommendation.

If the entry fills at $40.20 while the stop stays $39.20, estimated risk rises to $1.04 per share: 119 shares now imply $123.76. Recomputing the arithmetic does not retroactively change the filled order. Your execution plan must address price uncertainty and exposure before submission.

A stop is also not a guaranteed maximum loss. The SEC’s stop-order bulletin explains the difference between a stop trigger and its execution price, and why a stop-limit can remain unfilled.

A wider stop does not improve reward/risk by itself

Entry $40, stop $39.20, target $41.60 gives $0.80 price risk and $1.60 reward: 2R. Widen the stop to $38.80 with the target unchanged, and the reward becomes about 1.33R. Reducing quantity can restore a cash-risk budget, but it does not change this price ratio.

Expectancy needs more than the win rate

In a fictional ten-trade sample, four winners earn +2R each and six losers lose −1R each. Gross total is +2R, or +0.20R per trade. If average costs are 0.15R per trade, net expectancy is +0.05R. If costs rise to 0.25R, it becomes −0.05R. These are arithmetic demonstrations, not backtest results.

Record actual net profit/loss divided by initial cash price-risk to obtain realized net R. Freeze the denominator. If your plan includes costs in the risk budget, name that different denominator so your journal does not mix units.

Review behavior separately from money

The chapter emphasizes emotion management and trade review. Translate a vague goal such as “be disciplined” into an observable behavior: did you wait for the completed trigger candle, use the specified quantity, and leave the initial stop unchanged unless the plan allowed an adjustment?

A profitable violation is still a violation. A rule-following loss may be correct execution of an unproven or losing method. Record both dimensions. A useful pause rule for a paper-trading exercise is to stop submitting new simulated orders after a specified number of process errors, then review those errors. The threshold is your experimental operating rule, not proof of improved returns.

Exercises 13–18

1. Size the trade. Cash budget $75, entry $30, stop $29.50, estimated variable round-trip cost $0.05 per share, no fixed cost. Calculate whole shares, exposure, and modeled loss.
2. Apply an exposure cap. For exercise 13, maximum permitted exposure is $3,000. What quantity remains, and what modeled loss follows?
3. Stress the stop. Buy 100 shares at $50 with intended stop $49. A trading halt is followed by an executable exit at $46. Calculate price loss before costs and compare with planned price risk.
4. Expectancy. Win probability 35%, average win 2R, average loss 1R, average costs 0.10R. Calculate net expectancy. Assume no breakeven trades.
5. Drawdown. A $10,000 account loses 20%. What percentage gain on the remaining balance restores $10,000?
6. Process review. You doubled quantity without authorization and made money. Write a review entry that does not reward the violation.

Claims to Investigate While You Read

The supplied book includes examples and confident statements, but it does not supply a complete reproducible dataset and methodology establishing the following ideas for every instrument and period. Keep them as research questions.

Research Lab: Does an Early Breakout Follow an Upward Gap?

The book motivates continuation and opening-range trades. Our experiment asks a narrower question: after an opening gap of at least a chosen percentage, what happens when a later completed five-minute candle closes above the first five-minute high?

This code tests one objective interpretation of the source’s concept. Any mechanical definitions introduced here are ours and are not necessarily the creator’s exact rules.

Choose a standard five-minute candlestick chart for a US stock with a 9:30 New York regular-session open, and switch off extended hours. The script rejects incompatible chart types/timeframes and observed extended-session bars. Start/end inputs select the research period. Leave enough earlier bars loaded for ATR and a prior session close. Use the same split/dividend-adjustment settings for comparisons and inspect corporate-action dates for mechanical gaps.

Market entries are submitted after the signal candle closes and normally become eligible for fill on the next bar’s opening tick. The stop/target instructions are submitted with the entry using relative tick distances, so their reference is the actual simulated fill. The fixed distance is frozen at submission; it is not recomputed from later ATR values. New setups cannot replace an active trade, and the attempt flag remains used even if a stop closes the trade immediately.

Defaults assume a $25,000 account, 10 shares, commission of 0.05% per executed side, and two ticks of simulated slippage. These are teaching assumptions; change commission, slippage, and capital in Strategy Properties to match the instrument and test. Fixed quantity is exposure sizing, not a fixed percentage risk model.

The final-session-bar exit uses an immediate simulated close to avoid intentionally carrying the position overnight. It assumes execution at that bar’s closing price, with the emulator’s applicable costs, rather than reproducing a real market-on-close submission process. Missing final bars, halts, or data gaps can defeat this intended flattening; inspect any overnight positions. Stop/target ordering within a candle remains emulator-dependent. Compare Bar Magnifier where data access permits it, and inspect large winners and losers manually. The broker-emulator mechanics are documented in TradingView’s strategy documentation (paid link).

Validation status: The script has received a static execution review but has not been compiled in TradingView or run against historical data here. No performance results are claimed.

//@version=6
// CandleChasers original research baseline; not the author's complete strategy.
strategy("CC Gap Workbook - Early Continuation Experiment", overlay=true,
     initial_capital=25000, pyramiding=0, margin_long=100,
     commission_type=strategy.commission.percent, commission_value=0.05,
     slippage=2, calc_on_every_tick=false, calc_on_order_fills=false,
     process_orders_on_close=false)

gapMinimum = input.float(4.0, "Minimum upward opening gap (%)", minval=0.0, step=0.5)
shares = input.int(10, "Fixed shares (not fixed cash risk)", minval=1)
atrLength = input.int(14, "Signal ATR length", minval=1)
stopMultiple = input.float(1.0, "Stop distance / signal ATR", minval=0.1, step=0.25)
rewardMultiple = input.float(2.0, "Target distance / stop distance", minval=0.1, step=0.5)
startDate = input.time(timestamp("01 Jan 2020 00:00 +0000"), "First eligible date")
endDate = input.time(timestamp("31 Dec 2030 23:59 +0000"), "Last eligible date")

if barstate.isfirst
    if not chart.is_standard or not timeframe.isminutes or timeframe.multiplier != 5
        runtime.error("Use a standard five-minute candlestick chart.")
    if syminfo.type != "stock"
        runtime.error("This workbook baseline is for US stocks.")
    if endDate < startDate
        runtime.error("End date must be after start date.")
if session.ispremarket or session.ispostmarket
    runtime.error("Switch the chart to regular trading hours only.")

// Previous session close is captured from the prior regular-session chart bar.
// No developing higher-timeframe values or future data are requested.
var float openingRangeHigh = na
var float openingRangeLow = na
var float gapPercent = na
var bool attempted = false
var int frozenStopTicks = na
var int frozenTargetTicks = na

signalATR = ta.atr(atrLength)
nyOpenMinute = hour(time, "America/New_York") * 60 + minute(time, "America/New_York")
nyCloseMinute = hour(time_close, "America/New_York") * 60 + minute(time_close, "America/New_York")

if session.isfirstbar_regular
    if nyOpenMinute != 570
        runtime.error("Use a US stock with a 09:30 New York regular-session open.")
    openingRangeHigh := high
    openingRangeLow := low
    gapPercent := not na(close[1]) and close[1] > 0 ? 100 * (open - close[1]) / close[1] : na
    attempted := false

// First five-minute bar is fully known before a later close can trigger.
// Signals at 10:00 or later are excluded; default first possible signal is 09:40.
eligibleTime = nyCloseMinute > 575 and nyCloseMinute < 600
eligibleDate = time >= startDate and time <= endDate
signal = barstate.isconfirmed and session.ismarket and not session.isfirstbar_regular and
     not session.islastbar_regular and eligibleTime and eligibleDate and
     not attempted and strategy.position_size == 0 and
     not na(gapPercent) and gapPercent >= gapMinimum and
     not na(signalATR) and close > openingRangeHigh

if signal
    // Freeze distances now. Relative brackets anchor them to the actual entry fill.
    frozenStopTicks := math.max(1, int(math.ceil(signalATR * stopMultiple / syminfo.mintick)))
    frozenTargetTicks := math.max(1, int(math.round(frozenStopTicks * rewardMultiple)))
    strategy.entry("GapLong", strategy.long, qty=shares)
    strategy.exit("GapBracket", from_entry="GapLong", loss=frozenStopTicks, profit=frozenTargetTicks)
    attempted := true

// Reissue the same fixed bracket; never move it with later ATR values.
if strategy.position_size > 0
    strategy.exit("GapBracket", from_entry="GapLong", loss=frozenStopTicks, profit=frozenTargetTicks)

// Intentional closing-price simulation; not an actual MOC order model.
// A missing last bar or halt can prevent this intended flattening.
if barstate.isconfirmed and session.islastbar_regular
    strategy.cancel_all()
    strategy.close_all(comment="Modeled session close", immediately=true)

plot(session.ismarket ? openingRangeHigh : na, "Opening range high", color=color.teal, style=plot.style_linebr)
plot(session.ismarket ? openingRangeLow : na, "Opening range low", color=color.orange, style=plot.style_linebr)
plot(gapPercent, "Opening gap percent", display=display.data_window)
plotshape(signal, title="Entry submission", style=shape.triangleup, location=location.belowbar, color=color.teal, size=size.tiny)

What This Code Actually Tests

It tests one long-only early-session breakout model after an upward regular-session opening gap. Eligibility is fixed from the open and previous regular-session close. The opening-range high is fixed after the first bar. A qualifying later close submits one market entry and its relative-distance bracket. Exit occurs at stop, target, or the modeled session close.

What This Code Does Not Test

It does not reproduce the book’s complete discretionary method, premarket-high entries, a one-minute trigger, low-float selection, catalyst verification, volume confirmation, VWAP conditions, a stop at the opening-range low, or a gap-fill strategy. It cannot model queue priority, actual spreads, partial fills, market impact, halts, or a portfolio of simultaneous candidates. A positive result would support further work on this narrow model; a negative result would not refute every gap strategy.

Backtest Lab

Freeze a baseline before running comparisons. Select development dates and a later untouched evaluation period in advance. If you repeatedly inspect and optimize the evaluation results, they become development data. Keep all attempted settings in your research log. A handful of trades, one strong ticker, or one favorable month cannot establish robustness.

Separate observation study: to investigate fill frequency, collect all eligible opening gaps, including days without a breakout entry. Prespecify bands such as 0.5–under 2%, 2–under 4%, and 4–under 8%, separately for up and down gaps. Mark a same-session fill only if price reaches the previous close. This requires a separate event log; the strategy’s winning-trade percentage is not a gap-fill rate.

Exercises 19–22

1. Audit timing. The 9:35–9:40 candle closes above the opening-range high. Under this script, when is the order submitted, and what is the usual earliest simulated fill?
2. Audit the bracket. Frozen stop distance is 50 ticks, tick size $0.01, target 2R, actual entry $20.30. Find the stop and target trigger levels before exit slippage and costs.
3. Audit the sample. You run the script on three stocks selected because they were last year’s biggest winners. What selection problem arises? Propose a better universe rule.
4. Audit the conclusion. A variant has four trades and the highest profit among 40 settings. Write a restrained conclusion and identify the next evidence you need.

Reusable Research Worksheets

Copy these tables into a journal or print this page. Write your answers before revealing future prices. Blank cells are intentional spaces for your work. Paper entries are observations, not submitted brokerage orders.

Worksheet A: Opening observation

Worksheet B: Setup and execution card

Worksheet C: Experiment comparison

Net win rate is profitable trades after the chosen costs divided by all closed trades, with breakevens counted separately in the record. Profit factor is gross gains divided by the absolute gross losses under a stated cost convention; it is undefined if the loss denominator is zero. Maximum drawdown is the largest peak-to-trough decline in the measured equity series. Report whether it is cash, percentage, closed-trade, or intratrade drawdown.

Worksheet D: Weekly process review

1. Which rule did I follow consistently? Cite a specific record.
2. Which behavior failed most often? Count occurrences, including profitable violations.
3. Were valid setups skipped? Record reasons separately from invalid setups.
4. What single operational change will I practise next session?
5. What evidence would show the change helped execution?
6. Am I changing a strategy rule? If yes, create a new version and preserve the old results.

Capstone: Build a Small, Auditable Research Report

Your final assignment is an investigation, not an income target. Choose one hypothesis from Chapter 3. Use the supplied script for its specific hypothesis or write a separate manual protocol for a different setup. Do not describe one as a test of the other.

1. Specify: complete Worksheet B, identify what comes from the reading, and label every added rule.
2. Collect: work through a preselected period without skipping unattractive examples. Aim initially for 30 eligible observations as a workflow exercise; this count is not statistical validation. If fewer exist, report the shortfall rather than widening rules after the fact.
3. Audit: inspect at least five individual cases, including a loss, a skipped setup, and any ambiguous fill. Check decision time against information availability.
4. Compare: change one variable and complete Worksheet C. Preserve the baseline and all attempted variants.
5. Evaluate: report findings on untouched later data when enough observations exist. If the sample remains sparse, label the conclusion inconclusive.
6. Explain: submit a one-page report with the hypothesis, exact rules, sample/exclusions, costs, findings, limitations, and next experiment. State whether the code compiled and which data were actually tested.

Worked micro-review: profitable rule-breaking

Suppose a learner’s five fictional observations include two compliant losses of −1R, one compliant win of +2R, one +1R win entered before the trigger, and one qualifying setup skipped without a rule-based reason. The three compliant executed trades total 0R before costs. The early-entry winner does not establish the compliant method’s result, and the missed setup is a process record, not a zero-return executed trade. Keep the overall account record as well; separating groups must not hide actual losses or costs.

Final assessment: exercises 23–28

23. Can an opening gap up coexist with overlapping completed daily candles? Explain.
24. Why is a filled gap not automatically a profitable gap-fill trade?
25. Does increasing share quantity change the price reward/risk ratio?
26. Why can a high win rate coexist with negative expectancy?
27. How can an “exhaustion” label accidentally leak future information into an entry rule?
28. What is the strongest defensible conclusion after completing this workbook without running any historical tests?

Completion rubric

If a skill is missing, return to its exercise and correct the reasoning. Completing every row supports further learning and investigation; it is not a readiness certificate for risking money.

Final Takeaway

The book introduces gaps, their possible interpretations, setup families, and risk management. This workbook turns those topics into definitions you can check, calculations you can reproduce, and hypotheses you can investigate. Your next useful step is to freeze one specification and collect honest evidence, including the cases that do not support it.

Risk Disclosure

This material is for education and research, not personalized financial advice or a recommendation to trade. Trading can cause substantial losses. Stops do not guarantee a maximum loss; liquidity, halts, gaps, spreads, fees, slippage, and leverage can materially change outcomes. Backtests and paper trades are hypothetical and cannot reproduce every aspect of live execution. Examples and research settings do not establish future profitability.

This is an independent educational companion to the book and is not affiliated with, sponsored by, endorsed by, or authorized by its author or publisher.