Market data can reach a trading platform in a different sequence from the one in which prices were recorded. Network delays, feed recovery, and timestamp differences may cause an older observation to appear after a newer one. The result is not necessarily a visibly broken chart: an automated strategy can misinterpret the order of events even when the latest displayed quote looks plausible.
In mt5, developers can inspect tick timestamps and request historical tick records, but they must distinguish a late observation from a genuine reversal. Five problems arise when a CFD strategy treats arrival order as market chronology.
Late Quotes Can Manufacture a Reversal Signal
A momentum algorithm may compare consecutive received prices without checking their timestamps. If an older, lower quote arrives after a newer high, the program can register a sudden decline that never occurred in that order. The problem is more likely during reconnection or a burst of updates from an intermediary feed.
A timestamp check is more useful than a simple price-change threshold. Compare tick times, including millisecond precision where available, and define how the strategy handles duplicates or records older than its last accepted observation.
Reconstructed Candles Can Give Indicators a False Sequence
Indicators built from locally accumulated ticks depend on assigning observations to the correct time interval. A delayed tick from the previous minute can contaminate the current candle if the program groups records by receipt time. Its high, low, or closing value may then diverge from the broker’s official bar history.
The discrepancy matters for breakout systems that trade at candle boundaries. Check whether the calculation uses server timestamps and whether historical bars are refreshed after connectivity interruptions. A visually ordinary candle does not establish that a custom indicator processed its inputs chronologically.
Stale Bid-Ask Pairs Can Distort the Expected Exit Cost
Imagine a hypothetical long cocoa CFD covering five metric tons at $7,800 per ton. A locally coded exit routine expects to close if the bid falls to $7,770, implying a $150 loss. During a feed interruption, its last cached bid remains $7,785 while a newer executable bid has moved to $7,735. An older update arriving late reinforces the misleading impression that the exit threshold has not been reached.
When the routine finally refreshes and closes at $7,735, the loss is $325, or $175 more than planned, before fees. Out-of-sequence data did not force the broker to execute at the old quote; it delayed the local decision. Inspect bid and ask freshness separately, and require a current executable quote before submitting a market order.
Cross-Symbol Signals Can Combine Prices From Different Moments
A spread strategy may compare two related index CFDs using the most recently received tick for each. If one feed lags, the calculated relationship can suggest an arbitrage opportunity that disappears once both instruments are measured at comparable times. More frequent updates for one symbol can actually make the combined signal less reliable.
For mt5 algorithms using multiple instruments, impose a maximum acceptable timestamp difference and suspend entry when either leg is stale. Also account for different trading sessions and scheduled quote interruptions rather than treating every missing update as a price discrepancy.
Feed Recovery Can Trigger Repeated Signals From Old Records
After a connection resumes, a program may process a backlog of observations as though each represented a fresh trading opportunity. Replaying threshold crossings can create several CFD entries even when the live market has moved beyond the original setup.
Recovery logic should separate historical reconstruction from live execution. Before opening a CFD position, replay delayed, duplicated, and reordered ticks in a test environment, verify that timestamps govern signal eligibility, and confirm that trading remains disabled until the latest bid and ask pass a defined freshness check.
