Skip to signal lessons
Trading Signals ExplainedMessages / Versions / Evidence

Signal timezones, delivery and stale alerts

Measure a declared validity window without treating a timely alert as a profitable trade.

Original edition: 9 October 2026. Expanded edition: 13 October 2026.

Educational explanations, fictional scenarios and personal worksheets. No live calls, paid-account test or provider performance certification.

Identify the start of the window

Valid for two minutes is incomplete until the start event is stated. It might begin at source publication, device receipt or entry activation. Those conventions can produce different eligibility for the same reader. Our worksheet begins at publication and subtracts declared transport, decision and acknowledgement delays. It is an invented arithmetic model, not a measurement of any messaging platform. The real service should declare its own validity convention in advance.

Record timestamps with a timezone or offset and enough precision for the actual use case. A calendar date alone cannot establish an intraday sequence. A local device clock can drift, while a provider database timestamp can be edited or assigned after processing. Keep clock source and observation method with the timestamp instead of treating every displayed time as equally reliable.

Walk through the fictional case

Assume publication at the start of a 120-second window. Transport takes 15 seconds, reading and deciding takes 30 seconds, and acknowledgement takes another five. Total delay is 50 seconds, leaving 70 at acknowledgement. That answer does not say price remained suitable or a fill happened. Acknowledgement merely marks the simplified endpoint chosen for the arithmetic; a further execution delay can consume additional time.

Change decision delay to 105 seconds while preserving the other inputs. Total delay becomes 125 and the worksheet marks arrival outside the assumed window. It does not recommend chasing the quoted price or restarting validity when the message is opened. Arrival exactly at the deadline is treated as expired under this exercise's declared boundary rule. A real service could use another boundary, but it must state it.

Time availability is not price availability

An alert can arrive inside its window after the entry opportunity has vanished. Price, spread, depth, risk limits and existing exposure need their own checks. A positive time remainder cannot override those constraints. Conversely, a later favourable price does not prove a reader could have entered earlier. Reconstruct information available at the decision, not a chart viewed after the outcome.

Session calendars matter when instructions refer to market open, session close or the next bar. Holiday schedules and daylight-saving changes can move the relevant instant. Do not infer a venue calendar from a reader's local date. For a historical review, retain the calendar and timezone mapping actually used, then identify ambiguity rather than silently selecting the mapping which produces the best trade.

Keep a delivery record without making a guarantee

Separate source publication, delivery service processing, device receipt, reading, submission, acknowledgement and fill where records support them. A receipt proves that one system recorded a message, not that a human acted on it. A hash and independent anchor can establish a limited existence claim for certain bytes, but neither measures delivery latency or ensures every subscriber had access before price moved.

The final consumer question is operational: does the stated validity and execution policy fit the reader's documented workflow? That question requires observed timestamps and terms, not a universal speed claim. Keep a small reproducible sample with all eligible messages, including late and expired ones. Publish unresolved clock differences and avoid a service-quality score when the sample is incomplete. The worksheet can organise assumptions while the evidence notebook holds source requests; neither is an external certification of timely delivery.

Reconcile two clocks

Create a fictional alert with source publication in UTC and device receipt in a local offset. Convert both to an explicit common instant before subtracting them. Introduce an unknown device-clock offset and explain why exact transport delay is now unresolved. Do not report negative transport delay as proven fast delivery or silently reset a clock.

Keep publication, uncertain receipt and later acknowledgement as separate observations. Test a delay equal to validity and a delay greater than validity to inspect the declared expiry boundary. Restore original assumptions afterward. These checks verify the lesson and arithmetic, not actual service speed. An operational recommendation still needs a complete observed sample and the applicable policy, including late alerts and clock uncertainty rather than only convenient timely examples.

Source documents

Documentation establishes the stated mechanism, not tested trading performance. Checked 9 October 2026; a later version can differ.

Continue the investigation

Signal validity and latency worksheet

Comparable UTC fields still need authentic clock sources

The teaching journal requires complete UTC seconds with Z and calendar validity. It rejects ambiguous local strings and receipt-before-event ordering under its declared comparable-clock contract. That is a format and assumption check, not independent clock authentication. A real negative apparent delay can indicate incomparable sources; it should not be repaired by quietly shifting timestamps or turned into a provider accusation.

Three receipts reach or pass entry expiry in the reference, including Gamma's exact boundary. A later close after entry expiry is still a separate exit event. Fourteen events are not fourteen trade decisions, and source chronology differs from delivery chronology twice. The original latency tool remains a four-assumption calculation rather than an observed network benchmark. No arbitrary remote feed, local publisher or workstation fallback is added to capture real messages.