Skip to article content

Execution pre-mortem · 6N

6N Trading Mistakes: A Practical Failure Checklist

The loss looked like a bad New Zealand-dollar call. The reconstruction showed something else: the trader sized from broker day margin, used an old $10 tick reference, entered before an event without checking depth, and left the wrong quarterly order working after the exit. The thesis was only one failure in a chain.

Illustrative incidentFour controls missed
BeforeStale tick value · margin-sized position
DuringEvent spread · stop slippage
AfterOld-month order left active

A post-trade label such as “bad luck” cannot close an operational control gap.

Reconstruction before explanation

Put Every Decision on a Clock Before Assigning Cause

A useful review begins with records: exchange time, contract month, quote, displayed depth, order messages, fills, event timestamp, broker notices and account changes. Memory is not a fill report.

T-30

Plan written

Direction and stop noted, but no source version, cost stress or expiry field.

T-2

Event ignored

A scheduled release was known, yet no no-trade window or depth gate existed.

T+0

Order filled

Spread widened and the marketable order crossed more levels than assumed.

T+7

Stop slipped

The trigger fired, but the realized exit was worse than the planned stop price.

T+20

Position “closed”

A stale order in the old quarter remained working and later reopened exposure.

This timeline is a constructed teaching example, not a reported trade. Its job is to show how a single P&L number can hide independent mechanics, event, execution and lifecycle failures.

Failure family 1

Wrong Contract Math Makes Every Later Decision Unreliable

The current standard 6N unit is 100,000 NZD and the current CME Globex outright increment is 0.00005 USD per NZD, worth $5. Older material may show a $10 tick. A claimed M6N Micro is not present in CME’s current product guide.

FailureObservable warningPrevention gateRecovery record
Stale $10 tickRisk sheet uses 0.0001 as one current outright tickReconcile 100,000 × 0.00005 = $5 from current Rule 258Recalculate planned and realized P&L; version the source
Side reversedPlan says stronger NZD but ticket is short 6NWrite: higher USD-per-NZD helps long, hurts shortSeparate thesis error from order-side error
Margin-sized quantityContracts equal “what the broker allows”Floor risk budget divided by stop risk, costs and stressRecord the leverage and liquidation path
Invented MicroPlatform or article claims M6N without a CME chapterRequire exchange listing, unit, tick and active chainCancel; treat standard quantity as zero if oversized

Failure family 2

A Narrative Without a Falsification Point Cannot Control Risk

“Dairy is strong,” “the dollar is weak” or “risk-on should help NZD” are mechanism fragments, not complete trades. Each can be offset by relative-rate expectations, China-sensitive growth news, positioning, U.S. data, liquidity or a different market prior.

Data mismatch

The chart uses a back-adjusted continuous series while the ticket trades one unadjusted quarter. Gate: map research and execution symbols explicitly.

Prior omitted

A release is called “good” without comparing it with consensus or market pricing. Gate: record the expected range and what would be surprising.

One-factor certainty

A plausible export or rates channel is treated as a guaranteed direction. Gate: name rival drivers and observable confirmation.

No invalidation

The position persists because the story can always be rewritten. Gate: define a price, time or evidence state that rejects the thesis.

Failure family 3

Nearly 23-Hour Access Is Not a Fill-Quality Guarantee

6N follows the regular CME FX schedule, but spread, depth and slippage vary. The relevant evidence is the live book in the exact quarter and order size, not a permanent “best time” slogan.

Observe

Measure before entry

Record spread, top-of-book size, nearby depth and recent trade flow.

Limit

Define permitted orders

Set maximum spread, slippage and market impact; define wait and reject branches.

Monitor

Watch order state

Distinguish accepted, working, partial, filled, cancelled and rejected messages.

Reconcile

Match fills to position

Verify quantity, side, average price and all working orders after every action.

A stop trigger is not a guaranteed fill.

Market gaps, thin depth, outages and rapid repricing can move the execution beyond the trigger. Include a gap stress and size the account for losses beyond the planned stop and initial margin deposit.

Failure family 4

The Calendar Check Must Cover New Zealand, the United States and Rival Drivers

RBNZ decisions and New Zealand data are obvious, but U.S. releases, Federal Reserve communication and material China-sensitive news can also reprice NZD/USD. An event plan must verify official release time and daylight-saving conversion rather than copy a stale calendar.

Verified event and acceptable book

Trade only a prewritten branch with quantity, invalidation and order type fixed.

Verified event but failed liquidity gate

Reduce or wait; a strong thesis does not repair a wide spread or shallow book.

Timestamp, prior or release package unclear

No trade. Revisions and statement detail can matter more than the headline.

Failure family 5

A Root Symbol Hides the Quarter and the Delivery Obligation

6N is physically deliverable. Trading terminates on the second business day before the third Wednesday under Rule 258, while a broker may impose an earlier cutoff. A roll is two trades, not a chart setting.

Wrong quarter

Chart, ticket and risk sheet disagree on month or year.

Silent liquidity migration

The old quarter remains in the plan after activity moves.

Partial roll

Old leg closes but new leg fills only partly, changing net exposure.

Stale orders

Working orders survive a roll or exit and later reopen risk.

Cutoff surprise

Broker closes or restricts the contract before the exchange deadline.

Delivery assumption

“Everyone rolls” replaces an actual offset and reconciliation.

After a control breaks

Stabilize the Account Before Editing the Story

The first objective is to establish the real position and stop unintended orders. Analysis comes after operational state is known.

1

Freeze

Stop new entries and automated strategies tied to the affected symbol.

2

Reconcile

Confirm every position, fill, working order, margin change and cash movement.

3

Contain

Reduce unintended exposure through an approved order path; do not improvise delivery.

4

Classify

Separate thesis, data, mechanics, execution, event and process causes.

5

Repair

Add an observable gate, owner and test; replay before live trading resumes.

Failure checklist

Pre-Trade and Post-Trade Answers Must Be Written

Before entry

Prevent

  1. Exact quarter, quote direction, unit and current $5 outright tick reconcile.
  2. Integer quantity fits stop, all costs, gap stress and funding buffer.
  3. Thesis has rivals, confirmation, invalidation and a time limit.
  4. Official event times and daylight-saving conversions are verified.
  5. Live spread and depth pass; expiry, roll and broker cutoff are recorded.

After exit

Prove

  1. Net position is zero or equals the intended remaining hedge.
  2. No stale orders remain in any quarter or linked strategy.
  3. Realized P&L reconciles from fills and fees, not chart marks.
  4. Slippage, spread and failure stress are compared with the plan.
  5. Any breach becomes a tested control change before the next live order.
Sources, methods and editorial disclosure — reviewed August 20, 2026

Sources were reviewed August 20, 2026. Incident examples are constructed and do not describe a customer or original trading study. Mechanisms and controls are educational hypotheses about failure prevention, not evidence that any strategy is profitable. Live liquidity, costs, calendars and broker rules must be checked at decision time.