20 Trading Rule Examples You Can Actually Score
A useful trading rule defines an observable trigger, an action, an exception policy, and the evidence needed to score it. It cannot eliminate losing trades or guarantee discipline. It can prevent a named action, expose a violation, and make the next review less dependent on memory.
When [observable condition], I will [specific action], unless [predefined exception]. I will verify it with [field or artifact].
Replace every bracket with a value derived from your instrument, strategy, account constraints and loss tolerance. Do not copy a percentage or trade count merely because it appears in somebody else's rulebook.
The 20 examples below cover risk, entry, execution, session control and review. They are templates, not personalized investment advice. Put the selected rules inside a complete trading plan, test them in a simulator or other controlled environment where appropriate, and define what evidence would cause you to revise them.
What Makes a Trading Rule Enforceable?
Binary language helps, but not every valid rule needs a number. “Trade only when the named setup is present” can be enforceable if the setup has observable criteria and screenshots. “Be disciplined” cannot be scored because it does not identify an action or boundary.
| Vague wish | Auditable rule | Evidence |
|---|---|---|
| Manage risk | Initial planned loss may not exceed the account's approved per-trade risk budget. | Balance/equity basis, stop, quantity, tick or pip value, costs |
| Do not overtrade | After the session trade cap or loss cap triggers, no new order may be submitted. | Order timestamps, trade sequence, cap state |
| Follow the plan | Every required setup field must be recorded before entry; missing means no trade. | Pre-entry checklist and timestamp |
| Cut losses | The exit method and invalidation are defined before entry; any override requires its own recorded rule. | Pre-trade plan, order audit trail, fill |
| Stay calm | If the named behavior trigger occurs, pause new entries for the predefined reset procedure. | Behavior tag, trigger time, next eligible time |
A rule is incomplete when it cannot answer four questions: What state activates it? What action follows? What happens if the platform or market prevents that action? Which source proves compliance? The last two questions matter because a planned stop is not the same as a guaranteed fill.
Risk Management Rules
Rule 1: Cap Initial Planned Loss Per Trade
Template: “Before entry, initial planned loss including estimated costs must be no greater than [risk budget] of [declared balance or equity basis].” Calculate quantity from the stop distance and instrument value. A valid risk budget comes from your capacity to absorb losses and the strategy's tested distribution; this article does not supply a universal percentage.
Worked placeholder: a rule may use 0.5%–1% per trade when that range fits the trader's tested distribution, total exposure, loss capacity, and account constraints. The numbers are common template inputs, not a recommendation; replace the range with the actual approved limit and its balance/equity basis.
Score: compare the pre-entry calculation with the position actually opened. Keep planned risk separate from realized loss because gaps, slippage and fees can make them differ.
Rule 2: Cap Total Open Risk
Template: “No new position may raise aggregate planned loss above [portfolio cap]. Positions sharing [defined risk factor] are also subject to [cluster cap].” Five individually small positions can express one concentrated exposure. The rule therefore needs both position-level and portfolio-level logic.
Score: snapshot every open position, stop, quantity, currency conversion and cluster label at the decision time. “Different symbols” is not evidence of diversification.
Rule 3: Stop New Entries at the Session Loss Boundary
Template: “When realized loss plus the declared treatment of open risk reaches [session boundary], cancel eligible entry orders and submit no new entries until [reset time].” Define whether commissions, funding and open P&L count. If the account is governed by broker or program rules, the official rule calculation wins.
Score: save the boundary, timezone, reset time and equity inputs. A local journal cannot claim a live lockout unless it receives current account and order state from a verified route.
Rule 4: Define the Protective-Order Failure Path
Template: “Before entry, choose [stop, stop-limit, option hedge or manual contingency], record its trigger basis, and define what happens if it rejects, gaps or does not fill.” A stop order can constrain the plan without guaranteeing the execution price. A stop-limit can control price but may remain unfilled.
The SEC's updated stop-order bulletin explains those trade-offs for securities, while the CFTC's required commodity-risk language warns that market conditions can prevent contingent orders from limiting losses as intended.
Rule 5: Add Only Under a Prewritten Scale Plan
Template: “I may add only at [defined condition], with [new stop/invalidation], while total planned loss remains inside Rules 1 and 2.” This is more precise than “never add to a loser” or “always add to winners.” Price relative to entry does not by itself prove whether an add belongs to a tested strategy.
Score: record the original scale schedule and compare every child order with it. An unplanned add is a violation even if the combined trade later wins.
Entry Rules
Rule 6: Every Entry Has a Defined Setup ID
Template: “No order is eligible until setup [ID/version] is selected and all required conditions are recorded.” A setup must define instrument universe, timeframe or holding horizon, trigger, invalidation and any market-state filter. “It looked right” is not a setup.
Score: require the setup version and pre-entry evidence. Do not let a post-trade label retroactively turn an impulse into a planned trade.
Rule 7: The Entry Zone and Expiry Are Predefined
Template: “The entry is valid only inside [price or condition zone] until [time or state expiry]. If price leaves the zone first, the idea expires and must be reassessed.” This prevents a planned trade from quietly becoming a chase with different risk and payoff.
Score: compare intended zone, order timestamp and fill. If the strategy explicitly allows re-entry, give that action a separate trigger.
Rule 8: Missing Decision Data Means No Trade
Template: “If [required quote, spread, volatility input, market status or account state] is missing or stale, submit no new entry.” This is a data-integrity rule. Guessing a stop value, exchange rate or remaining account limit produces a precise-looking decision from unknown inputs.
Score: store source timestamps and freshness thresholds. Mark unavailable values Unknown or Not verified; zero is a number, not a substitute for missing data.
Rule 9: Scheduled-Event Exposure Follows the Tested Strategy
Template: “From [lead time] before [named event class] until [normalization condition], new entries are [blocked or allowed only by strategy version].” A universal number of minutes does not fit every instrument or strategy. A news strategy and a strategy that avoids event volatility need different rules.
Score: save the calendar source, event, timestamp, strategy version and decision. Unexpected events remain residual risk.
Trade Management and Execution Rules
Rule 10: Map Invalidation and Exit Logic Before Entry
Template: “Before entry, record the thesis-invalidating condition, protective-order plan, target or trailing logic, time exit and end-of-session treatment.” A target is not automatically required, and a fixed reward-to-risk threshold is not universally profitable. The exit model must belong to the tested strategy.
Score: compare every exit with the pre-entry rule and allowed adjustments. Outcome alone does not establish decision quality.
Rule 11: Reject Entries Outside the Execution Tolerance
Template: “Do not enter when displayed spread, estimated market impact, fee, borrow/funding condition or expected slippage exceeds [documented limit].” The limit can vary by instrument and order size, but it must be fixed before seeing whether the trade wins.
Score: store arrival quote, order size, fill sequence, fee asset and benchmark. The broader risk-management guide shows why planned position loss and realized execution loss must be reconciled separately.
Rule 12: Every Rejection, Partial Fill and Cancel Has a Response
Template: “If an order is [rejected, partially filled, pending beyond timeout or disconnected], take [cancel/reconcile/reduce/no-resubmit action] before any replacement order.” Blind resubmission can duplicate exposure. Assuming a rejected protective order exists can leave exposure unmanaged.
Score: preserve client order IDs, venue order IDs, status events, fills and replacement linkage. After a connection gap, fetch authoritative order and position state before acting.
Rule 13: Do Not Expand Risk Without a Predefined Rule
Template: “After entry, no stop, quantity, hedge or expiry change may increase worst-case planned loss beyond [approved remaining budget] unless [named contingency rule] applies.” “Never widen a stop” is simpler but can misdescribe strategies with planned volatility- or structure-based management. The invariant is the approved risk boundary.
Score: replay every modification against the versioned plan. A profitable override is still an override.
Session and Portfolio Rules
Rule 14: Trade Only in the Defined Operating Window
Template: “New entries are permitted only during [session and timezone], except for [explicit strategy exception].” The rule reduces untested regime drift and defines when monitoring is available. It is not a claim that one market session is universally better.
Score: normalize source timestamps to UTC, then derive the named local session. Daylight-saving changes need an explicit policy.
Rule 15: Cap Activity Using Your Own Degradation Evidence
Template: “After [trade/order/decision count] in one session, no further entry is allowed unless [predefined exception].” A fixed cap of three trades is not universal; a scalper and a swing trader have different units of activity. Choose the cap from strategy design and observed error rates, not folklore.
Score: compare trade sequence with plan compliance, execution quality and setup mix. The overtrading workflow separates high legitimate frequency from low-quality escalation.
Rule 16: Stop After a Risk-Control or Platform Failure
Template: “After [duplicate order, missing quote, rejected protection, unknown position, stale account state or platform disconnect], submit no new entries until reconciliation passes.” This circuit breaker can matter more than a consecutive-loss rule because it responds to loss of control, not a random outcome sequence.
Score: record failure type, detected time, state-recovery evidence and authorized restart time.
Rule 17: Use a Defined Cooldown After a Violation
Template: “When rule [ID] is violated, pause new entries for [procedure or interval], reconcile open risk, and complete [reset checklist] before resuming.” Consecutive losses do not necessarily prove poor execution; a rule violation is direct evidence that the process left its boundary.
Score: the violation must be logged before the next entry. If the reset is repeatedly bypassed, redesign the environment—alerts, order permissions, reduced size or session termination—rather than relying on another promise.
Review and Behavior Rules
Rule 18: Score Compliance Before Explaining the Outcome
Template: “Record pass, fail, not applicable or unknown for each active rule before writing the trade narrative.” A winning trade can violate the plan; a losing trade can comply. Scoring the process first reduces the temptation to excuse a winner or invent a flaw in a loser.
Score: require the rule version, status and evidence link. Unknown is different from pass.
Rule 19: Close the Record Within a Defined Window
Template: “By [time or session-close event], reconcile fills and costs, attach evidence, score compliance and record any exception.” Ten minutes may fit one workflow and interrupt another. The useful boundary is one you can follow without compromising live risk management.
Score: compare close time, review-completed time and missing fields. Raw exports should remain immutable even if annotations change later.
Rule 20: Change One Rule Through a Versioned Test
Template: “Change rule [ID] from version [A] to [B] only after [defined evidence threshold]; test it for [window]; revert when [rollback condition].” Avoid editing several rules after one emotional trade. If the population, strategy or market regime changed, say so rather than attributing the result to the rule alone.
Score: keep both versions, effective timestamps, reason, population, exclusions and rollback decision. Use the trade-review process to test one question against comparable records.
How to Stop Breaking Trading Rules
Repeated violations can come from different causes: the rule is ambiguous, the evidence arrives too late, the platform does not support the control, the rule conflicts with the strategy, or the trader chooses to override it. The repair should match the cause.
| Failure | Evidence | Structural response |
|---|---|---|
| Could not tell whether it applied | Several unknown or disputed scores | Rewrite trigger, action and exception |
| Forgot the rule | No pre-entry acknowledgement | Put the checklist in the order workflow |
| Data was unavailable | Missing or stale input | Block entry or define an explicit fallback |
| Platform could not enforce it | Rejected order or unsupported control | Choose another control or stop the route |
| Override after loss or urgency | Rule known, then bypassed | Cooldown, permissions, reduced access or session stop |
| Rule harms the tested process | Comparable compliant records reveal the conflict | Run a versioned change with rollback |
Do not punish a loss merely because it feels bad. Diagnose the process evidence. If a loss triggers an urge to recover immediately, the revenge-trading guide provides a separate interruption and review sequence.
Test a rule before relying on it
First, run the sentence test: two independent reviewers reading the same record should reach the same compliance status. If one calls it a pass and the other calls it a fail, the trigger, action or exception is underspecified. Add the missing definition before collecting performance data.
Second, run the availability test. Confirm that every required input exists at decision time—not only after the trade closes. A rule that depends on final slippage, end-of-day margin or a reconstructed emotion cannot control the entry. Separate a real-time gate from a post-trade diagnostic.
Third, run the failure test. Simulate a rejected order, partial fill, stale quote, disconnected platform and missing journal record. State whether the safe response is cancel, flatten, stop new entries, reduce exposure or escalate for manual review. The response must fit the instrument and platform; there is no universal emergency order.
Fourth, run the conflict test. Two individually sensible rules can disagree—for example, “exit at session close” and “do not cross a wide spread.” The rulebook needs priority or exception logic before both triggers occur. Record which rule controls and why.
Finally, run the audit test on a small historical or simulated sample. Count pass, fail, not applicable and unknown separately. Reconcile the denominator: excluding unknown records can make compliance look artificially high. Only after the rule produces stable classifications should you ask whether compliant and noncompliant populations differ. Even then, association does not prove the rule caused the outcome.
Track the Rule, Version and Evidence in TSB
Ownership disclosure: Trader's Second Brain is our product. It does not enforce broker orders, prevent losses, guarantee discipline or make a strategy profitable. It can support the evidence loop after a recognized import or manual record: preserve trades, attach notes, group reviewed behavior, compare segments and revisit violations.
TSB recognizes 328 structured source profiles through canonical runtime truth. That is a parser and mapping count, not guaranteed compatibility with every account or custom file. Preview a representative source and reconcile fills, quantities, fees, currencies and account scope before relying on any rule score.
Minimal compliance record: rule ID, rule version, applicable/not applicable, pass/fail/unknown, trigger timestamp, action timestamp, exception ID, evidence link and reviewer note. Keep the imported economic record immutable; corrections to interpretation should be traceable.
All 20 Trading Rules
| # | Rule | Primary evidence |
|---|---|---|
| 1 | Cap initial planned loss per trade | Sizing calculation |
| 2 | Cap total and clustered open risk | Portfolio snapshot |
| 3 | Stop new entries at the session boundary | Timestamped equity inputs |
| 4 | Define the protective-order failure path | Order audit trail |
| 5 | Add only under a prewritten scale plan | Parent/child order plan |
| 6 | Assign every entry a setup ID | Pre-entry setup version |
| 7 | Predefine entry zone and expiry | Plan versus fill |
| 8 | Do not trade with missing decision data | Source freshness |
| 9 | Use a tested scheduled-event policy | Calendar and decision time |
| 10 | Map invalidation and exits before entry | Versioned exit plan |
| 11 | Reject entries outside execution tolerance | Quote, fills and fees |
| 12 | Handle every reject, partial fill and cancel | Linked order states |
| 13 | Do not expand unapproved risk | Modification replay |
| 14 | Use a defined operating window | Normalized timestamps |
| 15 | Cap activity from degradation evidence | Trade sequence analysis |
| 16 | Stop after a control or platform failure | Recovery checklist |
| 17 | Use a defined violation cooldown | Violation and restart time |
| 18 | Score compliance before outcome narrative | Rule status and link |
| 19 | Close the record in a defined window | Review completion time |
| 20 | Change one rule through a versioned test | Population and rollback |
Start With the Rules You Can Prove
Choose the smallest set that covers catastrophic risk, entry eligibility, execution failure and review. For each rule, fill in the trigger, action, exception, evidence and owner. Run the set without changing definitions mid-sample. Add a rule only when it solves a documented failure that the existing system does not cover.
FINRA's day-trading risk disclosure notes that frequent trading can produce substantial costs, system failures can cause losses, and margin or short selling can create losses beyond the initial investment. Those realities are why the rulebook needs costs, failure paths and a capital boundary—not motivational slogans. A rule cannot make trading safe; it can make a specific decision auditable.
