Skip to article content

Executable invalidation · explicit branches · evidence trail

6Z Trade Management: Preflight, Events, Stops and Review

A stop can be analytically valid and operationally unusable. If the order path, current spread, expected quantity or event state cannot support the exit, the trade fails before entry.

No order until the record is complete

Freeze the Decision Before Market Exposure

Pre-entry authorization card

Instrument
Exact 6Z expiry, side, account, venue and quantity.
Thesis
Observable condition, rival explanation and invalidation.
Execution
Order type, arrival benchmark, spread/depth gate, timeout and partial-fill rule.
Risk
Stop, stressed slippage, gap reserve, fees and dollar budget.
Events
SARB, U.S. data, central-bank and South African calendar restrictions.
Lifecycle
Time stop, overnight permission, roll date and broker delivery cutoff.
  • Mechanics pass. Quote direction, $12.50 tick and physical delivery match the 6Z specification authority.
  • Size passes. Whole-contract risk, portfolio cap and margin cap all permit the same quantity.
  • Market quality passes. Intended quantity fits the declared spread and depth thresholds now.
  • Technology passes. Data freshness, clock, connection, order acknowledgements and backup exit route are working.

State changes, rules do not

Manage the Position Through Prewritten Branches

1

Pending

Cancel if the signal expires, spread/depth fails or the event buffer begins before a fill.

2

Partial

Accept, cancel remainder or reprice only according to the frozen partial-fill rule; recompute risk on actual quantity.

3

Open

Attach or verify protection, record fill benchmark and monitor only declared state variables.

4

Impaired

Data, connectivity, liquidity or order-status uncertainty triggers the interruption protocol.

5

Resolved

Confirm zero position and open orders, then preserve the audit record.

Observed statePermitted actionForbidden improvisation
Thesis intact; quality passesHold under the original stop, target and time ruleAdding size because price moved favorably
Thesis invalidatedExecute the declared exit branchWidening the stop or switching timeframe
Spread/depth breachReduce, cancel unfilled quantity or exit under the impairment branchAssuming displayed liquidity will return
Unexpected informationApply the unscheduled-shock rule; flatten if no rule existsInventing a directional interpretation in real time
Platform ambiguityStop new orders, verify position at the authoritative broker/exchange stateSending duplicate exits without reconciling acknowledgements

The event package can change the entire price path

Use an Override That Starts Before the Release

An event policy must name the source, timezone, buffer, allowed position state and restart condition. A consensus-matching headline does not guarantee a quiet response; details, guidance, revisions, liquidity and prior positioning can dominate.

Flat

No event exposure

Cancel pending orders before the buffer and remain flat until spread, depth and data normalize.

Reduced

Predeclared smaller risk

Recalculate using event gap and execution stress; no discretionary add-back during the release.

Held

Explicit event thesis

Requires tested logic, enough loss liquidity and an order path that acknowledges stop gaps.

Blocked

Unknown calendar or stale data

If the source, timestamp or release status is uncertain, the position state is flat.

A trigger is not a fill

Design Stops as Order Processes

DecisionRequired declarationRisk to preserve
InvalidationExact price or observable state and when it becomes knowableChanging it after adverse movement
TriggerExchange/broker trigger basis and session applicabilityTrigger may differ from chart price
Order typeStop-market, stop-limit or monitored exit with a backup branchMarket order has price risk; limit order has non-fill risk
QuantityFull, staged or partial behavior and cancellation logicPartials can leave residual risk
ImpairmentResponse to halt, gap, stale data, reject or disconnectLoss can exceed the planned stop
A tighter stop is not automatically safer.

If ordinary spread and price noise repeatedly touch the trigger, the stop can increase churn without reducing tail risk. The analytical invalidation comes first; quantity must adapt to it.

Every open position needs a terminal state

Define Profit, Time and Scale Branches Without Hindsight

A target is not justified because it produces an attractive reward-to-risk ratio on the ticket. It needs a market or thesis basis and an executable order policy. A time stop asks a different question: how long may the thesis remain unresolved before capital, event or lifecycle risk outweighs the original premise?

BranchPretrade declarationLive control
Price targetExact level or observable state, quantity and order typeDo not move it farther solely because price approaches
Partial exitQuantity, trigger and treatment of the remaining stopRecompute residual dollar risk after every fill
Trailing exitPast-only update rule, activation state and minimum tick roundingNo discretionary loosening after activation
Time stopTimestamp, completed-bar count or session boundaryUse a timezone-aware clock and event calendar
End-of-dayHold or flatten permission and broker day-margin transitionDo not let a day trade become overnight by omission
Roll boundaryLast eligible entry and mandatory exit/roll dateThe broker's earlier delivery cutoff overrides strategy preference

Scaling out changes the remaining position but does not create “free contracts.” Open risk still includes stop distance, execution, gap, fees and portfolio exposure. Moving a stop to entry can reduce price risk, yet an adverse gap can still produce a loss. Record each child order and confirm that cancelled/replaced orders did not leave duplicate protection.

When target and invalidation arrive together

Bar data may show that both levels traded without revealing which came first. Classify the historical outcome as ambiguous unless higher-resolution authoritative data resolve sequence. In live management, use actual order acknowledgements and fills rather than chart reconstruction.

Price branch
Observable
Time branch
Timezone aware
Partials
Risk recomputed
Ambiguity
Preserved

Separate decision quality from market outcome

Capture Evidence at Every Exit

Decision record

Why the exit began

Timestamp, rule version, thesis state, event state, market-quality readings, open quantity and intended order.

Execution record

How the exit completed

Trigger, decision, arrival, submission, acknowledgement, partial and final-fill timestamps plus prices and fees.

(actual exit price − decision benchmark) × side sign × 500,000 × contractsfees

For an exit, use +1 for a buy and −1 for a sell; a worse execution should produce positive shortfall.

One trade cannot validate a method

Run an After-Action Review Without Rewriting the Plan

QuestionEvidenceClassification
Did every gate pass?Saved preflight inputs and source timestampsProcess pass/fail
Were branches followed?Order and decision logRule adherence/deviation
Was cost error material?Expected versus actual spread, slippage, partials and feesModel calibration issue
Was the thesis wrong?Predeclared observables and rival explanationsResearch question, not emotional label
What changes?Versioned rule proposal applied prospectivelyNo retroactive rescue

Control principle

A profitable rule breach is still a process failure

Outcome and adherence are recorded separately. Otherwise luck teaches the system to abandon its own controls.

Sources, methods and editorial disclosure — reviewed August 25, 2026

This page prescribes a control workflow, not profitable entries, target multiples, event directions or empirical 6Z performance. All thresholds must be defined and validated for the user's data, strategy, broker and quantity.