Module 6 · Lesson 4 of 17 · Risk, process, and mindset · About 5 minutes

Building a Trading Plan Document

On this page
  1. What you will learn
  2. What goes into a plan?
  3. Plan diagram
  4. Why write limits in advance?
  5. How does a plan help review?
  6. What this does not tell you about trading plans
  7. Method in Practice
  8. Check yourself
  9. Reflect on your own process
  10. Sources

Before you read: Trading Plans, Journaling and Honest Backtesting

What you will learn

  • What sections a written trading plan commonly includes.
  • Why limits on risk and time are written down before trading starts.
  • How a plan supports review without telling you what will happen.

What goes into a plan?

If someone else read your plan, could they tell what you intend to do?

Completed fictional worked path within SIM-P v1.0 — written before Study Day D1, 08:55 CST. D1–D5 below are invented successive study sessions, not real calendar dates or historical sessions. Purpose: practice recognizing whether a previously written rule was met and preserve evidence. This deliberately arbitrary rule is not a validated trading approach or a recommendation to copy it. Scope: one fully specified worked execution/sizing path, CONDITIONAL on the stipulated 100 fill. The bounded 99<C≤100 range classifies eligibility only; it is NOT an exhaustive executable protocol for every close in that range. Other eligible closes alone do not specify entry fill or size and do not justify reusing $22. No missing execution branch is invented here.

Field Filled v1.0 decision
Scope Manual simulator only; invented linear instrument SIM-X; no orders, borrowing, real money or member data. One price point=$1 per unit; quote increment 0.25; whole units only.
Study window Each D1–D5, 09:00–09:30 CST; at most one entered simulation per session. No need to create a real appointment.
Data/context Authored closed-bar input sheet, not a licensed feed. Before each session, three prior bars have (H,L,C) = (99,97,98), (99,97.5,98.5), (98.75,97.75,98.25). Freeze observed range 97–99 before seeing the next bar.
Observable eligibility First completed study bar closing strictly above 99 and at or below 100, with a timestamped before-entry note and all needed inputs present. This is only a dummy classification condition, not an execution/fill/size instruction or evidence of an edge.
Invalidation A close at 99 or below fails eligibility; above 100 also fails this bounded exercise. Once a simulated long is entered, the preset adverse reference is 98; do not widen it.
Entry/exit assumptions For this worked path ONLY, close 100 is assumed filled at 100 with zero additional slippage; no fill/size rule is supplied for other eligible closes; exit at the first simulated observation reaching 98, otherwise at exactly 09:30 CST. These are stipulated simulator fills, not what a live market guarantees.
Funds/size units Dummy available cash $2,000; maximum 10 units. At entry 100, 10 units represent $1,000 notional. No futures contract multiplier or margin rule is inferred.
Planned amount Chosen teaching cap $25 per simulation, including $2 fixed total position round-trip costs, not $2 per unit or per side. At entry 100/exit reference 98: per-unit price risk $2. floor[(25−2)/2]=11 units mathematically; apply separate 10-unit cap → 10 units. Gross planned amount 10×2=$20; plus $2 costs=$22, leaving $3 of the $25 comparison amount.
Session stop After one entered simulation, no second entry that day. If net loss reaches the separate $25 review threshold, stop new simulation entries and review records; a gap/slippage could exceed a threshold, so it is not a maximum-loss guarantee. Missing inputs or inability to record the before-note also mean no new entry.
Records Plan version; snapshot ID/time; frozen band; completed close; rule result; chosen quantity; entry/98 reference; assumed fixed costs; actual simulated exit and reason; after-note; record uncertainty without filling missing facts from later outcomes.
Review version boundary Keep v1.0 fixed through D5; review at D5, 16:00 CST. Record any later workflow change as a new version applying only to subsequent study opportunities. No retrospective change to eligibility.

Apply the eligibility classification; only the first situation includes the completed 100-fill execution/sizing assumptions:

  • Before the next bar, frozen band 97–99 and all required fields are present. Completed close 100 passes 99<100≤100; hypothetical 10 units at 100 with 98 adverse reference use $22 including fixed costs. “Eligible” means the stated classification predicate is met; the separate assumed 100 fill is required for this $22 worked amount, not inferred for all eligible observations.
  • Classification-only counterexample: with complete inputs, close 99.50 satisfies 99<99.50≤100. It is NOT an automatic eligibility failure. Its entry fill and quantity are UNSPECIFIED here; classification alone cannot supply either or justify copying $22.
  • Same context, completed close 98.75: it fails the strict above 99 condition. Record no entry even if the next invented bar later rises to 103. The future result cannot change the earlier failed gate.
  • Completed close 100 but the input/snapshot record is unavailable: record no entry due to missing information, not a fabricated pass. A close 100.25 also fails the upper boundary even with a complete record.

These situations distinguish an objective rule result, a missing-data state and a later price outcome. For foundational arithmetic/checklist context see Risk Management Arithmetic: Position Size, Risk-Reward, Drawdown, Trading Plans, Journaling and Honest Backtesting and A Pre-Trade Checklist: Slowing Down Before Acting.

Plan diagram

Fictional simulator plan. A strip runs from SIM-X / manual simulation only to freezing the input range first. Left, the completed-close classifier 99 < C ≤ 100: a number line from 98.75 (fail) to 100.25 (fail) with the pass zone above 99 (open circle) up to 100 (closed circle) and 99.50 marked; below it five cases: 100 and 99.50 are classification passes, 98.75 and 100.25 are classification fails, and 100 with missing evidence fails. Right, stipulated fill at 100: entry 100 against adverse reference 98, a separate 10-unit cap giving 10 units, and 10 × (100 − 98) × $1 = $20, then a bar of $20 plus $2 leaving $3: $22 planned versus $25 cap. Footer: stop and preserve evidence.
Read the diagram step by step
  1. Completed fictional SIM-P v1.0. Simulator only; no live strategy or size recommendation.
  2. D1–D5, 09:00–09:30 CST. Invented study sessions, not real calendar dates. One entered case per day.
  3. Frozen range 97–99. Completed-close classifier: 99 < C ≤ 100. Needed inputs and timestamped before-entry note must be present.
  4. CLASSIFICATION ONLY
  5. Eligibility does not specify a fill or size.
  6. The completed worked path below additionally stipulates a fill at 100.
  7. 99.50: classification pass ONLY
  8. Fill/size unspecified. Do not reuse $22 from the conditional 100-fill path.
  9. Below the strict above-99 boundary → no entry.
  10. Above the at-or-below-100 boundary → no entry.
  11. Missing inputs/snapshot or before-note → no entry, not a fabricated pass.
  12. CONDITIONAL ON 100 FILL
  13. Entry 100 → adverse reference 98. 10 units; $1/point/unit. Stipulated fill at 100, zero additional slippage.
  14. Per-unit price risk = $2. floor[(25 − 2)/2] = 11 whole units. Separate 10-unit cap → 10 units.
  15. 10 × (100 − 98) × $1 = $20. + $2 FIXED total POSITION round-trip cost = $22 planned versus $25 cap. Not $2 per unit or per side.
  16. $20 price-distance component + $2 fixed position costs; $3 of the $25 comparison amount remains. Planned amount is not a maximum-loss guarantee.
  17. One entered case/day. Missing evidence → no entry. Review D5, 16:00 CST; future changes versioned.
  18. Fictional simulator exercise. No live setup, size or schedule is recommended.

Illustrative diagram constructed for teaching. Not real market data.

One fictional 100-fill path has reproducible sizing/cost arithmetic. The broader 99<C≤100 predicate classifies observations but does not complete an execution protocol for every eligible close. No live setup, size or schedule is recommended.

Why write limits in advance?

Why decide a daily loss limit before the day starts?

Decisions made during a fast market are often affected by emotion. A daily loss limit and a per-trade risk written in advance set boundaries before pressure arrives. The arithmetic from the risk lesson helps turn those limits into position sizes. Limits only work if they are followed, which is why review checks whether they were.

How does a plan help review?

How can you tell if a bad day came from the plan or from not following it?

Compare the before-event version with the information available then. A compliant loss and a rule-breaking gain are different process outcomes even if the gain looks attractive afterward. A missing before-record leaves eligibility unknown; a later winning result is not substitute evidence.

For the completed 100-fill worked path, v1.0 asks two separate questions: “Was 99<completed close≤100 known?” and “Were the assumed 100 fill, 98 reference, 10 units and fixed $2 cost actually the stipulated case supporting $22?” A lower eligible close such as 99.50 answers the first classification question but does not supply the second question’s missing fill/size assumptions. It does not ask “Did the next bar happen to rise?” Keep the original note, then append dated corrections of clerical facts or a new version for future sessions. See Journaling Your Reasoning, Not Just Your Results and The Weekly Review: Process Before Outcome for the supplied completed records and review; no profitable method is established by following this example.

What this does not tell you about trading plans

  • What any plan should contain for you.
  • Which conditions, methods or limits anyone should use.
  • Whether following a plan leads to gains.
  • Nothing here suggests any trade.

Method in Practice: John's complete plan

Fictional example. John, Harbor Index futures, the prices, times, costs and results are all invented for teaching. This shows how a written plan can be built and tested. It is not a recommendation, and nothing here says how often this idea works.

Try it on a real chart. Test this idea yourself on past sessions with bar replay. Keep your own count, including the days with no trade.

The idea. Trading Plans, Journaling and Honest Backtesting lists what plans commonly cover. Here is John's plan for one method, filled in under each of those headings, so you can see what "precise enough that someone else could check it" looks like.

John's plan headings: markets and hours · preparation · conditions · risk · orders · limits · review (full document below)

Three outcomes (how the plan handled the Volume Profile Explained: POC, Value Area, HVN, LVN lesson's replays)

  • Worked: every step was covered by a written line, including the time limit that closed the trade

  • Failed: the "wrong if" line and the risk section decided the exit and kept the loss inside the cap

  • No trade: the conditions section said two closes in a row, so one close meant no trade

Show John's full plan
HeadingJohn's plan (version 2)
Markets and hoursHarbor Index futures, front month. Regular session only, 08:30–15:00 CT. No positions overnight
PreparationBefore 08:15: mark the prior value area (70% volume, same platform settings every day), the prior high and low, and the overnight range. Check the calendar for releases in the session
ConditionsOpen below the prior value area low, by no more than 1 point. Two 30-minute closes in a row back above the value area low. Not within an hour of a scheduled major release. No setup by 12:00 means no trade
Risk$150 maximum per idea, costs included. Size from the Sizing Futures Positions in Contracts: Ticks, Tick Value and Whole Numbers lesson's arithmetic, at most 2 contracts. Stop after $300 lost in a day
OrdersEntry: market after the second close. Wrong if: market after a 30-minute close below the value area low. Profits: limit orders at the POC (half) and the value area high (rest). After the first profit, a stop at the entry price. Time limit 14:30
Limits on activityOne idea per day. No new trade to "make back" a loss
ReviewJournal entry every session, including no-trade days. Weekly review on Fridays: process first, then results. One change at most, re-tested on held-back sessions

Test it yourself

  • Give your plan to someone else and ask them to replay one session from it. Where do they have to ask you?
  • Check each heading has at least one line you could prove or disprove from a chart.
  • Date your plan and give it a version number.

Where it breaks

  • A plan can be followed perfectly and still lose.
  • A plan with vague words ("strong", "clearly") can't be checked.
  • Plans go stale. The review section is what keeps them current.

Reflection question: Which heading in your plan is the thinnest, and what would make it checkable?

Check yourself

  1. Under the stipulated 100-fill path, why do ten units use $22 rather than $40? Does the same amount follow from eligibility alone?
  2. Does close 100.25 qualify? What about close 99.50 with complete inputs, or close 100 with missing inputs?
  3. Can a later rise to 103 retroactively make 98.75 eligible?
  4. Name four kinds of information a trading plan can record without supplying recommended numerical limits. Hint: instruments, hours, limits, exits.
  5. Why record a session loss-limit rule before the session rather than invent or relax it under pressure? Hint: emotion under pressure.
  6. How can records separate a plan-design question from an execution-adherence question? Hint: compare each trade with it.

Reflect on your own process

For your own notes; there is no score and no right answer.

  1. Could someone else read your written plan and tell, without asking you, whether a situation met its conditions?
  2. Where does your plan record its version number, and how do you mark changes made after a review?
  3. What does your written plan say to do when a required input or record is missing?
  4. Which heading in your plan is the thinnest, and what would make it checkable?

Sources

Last reviewed . Spotted a mistake? Report an error.