We may earn a commission through these partner links. Demo funds are virtual; access and country eligibility depend on the platform terms. Disclosure
The moment a user clicks, the moment an order is accepted and the moment the screen updates can differ. A fair execution review records these events rather than treating a countdown animation as the whole timeline.
Map the events in order
A user action travels through the device, network and service before an accepted order record is produced. The interface then receives and displays an update. Each stage can have its own timing.
A slow screen does not by itself prove that the server accepted the order late. Conversely, a responsive animation does not prove which quote was accepted. The transaction record is essential.
An illustrative timeline
| Event | Example local time |
|---|---|
| Button clicked | 10:00:00.100 |
| Request reaches service | 10:00:00.180 |
| Order accepted | 10:00:00.210 |
| Display updated | 10:00:00.320 |
These are invented values to illustrate distinct events. They are not measurements of any provider’s infrastructure.
Clock alignment is a prerequisite
A device clock and a server clock can differ. Record timezone, offset and the source of each timestamp. A screen recording may show local elapsed time while the contract record uses another clock.
Do not subtract timestamps from unsynchronised clocks and call the difference measured network latency. The resulting number can combine clock offset and real delay.
Design a useful observation exercise
Use demo mode for a narrow interface investigation. Record device, connection, full symbol, chart interval, click observation and accepted contract record. Keep failed submissions and missing updates as separate outcomes.
Repeated observations can describe that setup. They do not establish a universal execution-speed average without a defined sample and measurement method.
Separate delay from a price complaint
If the concern is a different accepted quote, preserve the quote and timing evidence. If the concern is an order rejected or never accepted, preserve the error and request history.
These cases need different explanations. A single cropped image of the final chart cannot establish all stages of the process.
Ask support for the relevant record
Provide the contract or request ID, approximate time, timezone and observed discrepancy. Ask which server acceptance and settlement timestamps apply and which terms define them.
Avoid presenting an estimated local delay as proof of intentional manipulation. Equally, do not dismiss a documented discrepancy without an explanation. State the observed facts and unresolved measurement limits.
Calculate only intervals the evidence supports
Using the fictional timeline above, click-to-acceptance is 110 milliseconds and click-to-display is 220 milliseconds. Acceptance-to-display is another 110 milliseconds. These calculations assume aligned clocks and supported event timestamps; they are not measurements collected from a broker.
If a recording shows only a click and the later interface response, it supports an observed interface interval. It does not reveal the request’s arrival time or the internal processing interval. Do not invent intermediate events simply because the final animation appeared late.
Ping is not the complete order lifecycle
MetaTrader 5’s virtual-platform documentation describes a ping field measuring network delay between the virtual server and the broker’s trade server. That metric concerns a specific connection. It does not separately measure a user’s device processing, every order’s internal handling or the time needed to render the response on a browser screen.
A low ping therefore does not guarantee a particular execution price or instant acceptance. Likewise, a faster internet download plan does not by itself establish lower delay on the relevant network path. Keep throughput, connection delay and full transaction timing as different observations.
| Observation | What it can describe | What remains unknown |
|---|---|---|
| Screen recording | Visible sequence and frame timing | Hidden server events without corroborating records |
| Ping indicator | Network-delay metric for a stated connection | Full order-processing and display delay |
| Accepted transaction record | Recorded acceptance details | Unreported stages before acceptance |
| Settlement record | Recorded expiry or closing event | Whether a separate local animation was timely |
A clock-offset example
Suppose a device clock is 300 milliseconds ahead of a server clock. Comparing a local click timestamp directly with a server acceptance timestamp could produce an apparent negative interval even though the server accepted the request afterwards. The arithmetic would be combining clock offset with real elapsed time.
Record how the clocks were aligned and any uncertainty. If alignment cannot be established, compare events measured by one reliable clock or report the cross-clock interval as unresolved. A timezone label alone does not establish millisecond synchronisation.
Design a repeatable demo observation
- Choose one interface question, such as how long a confirmation takes to appear.
- Confirm virtual account mode and identify the product and symbol.
- Record device, browser version, connection type and the measurement method.
- Capture the relevant visible events and available transaction records.
- Retain failed requests, missing confirmations and interruptions as separate outcomes.
- Report the number of observations and what each measured interval includes.
Use demo to investigate controls without creating real-money exposure. Demo observations do not establish live execution behaviour. Do not repeat a submission blindly when its status is uncertain; first check the account record to avoid confusing duplicate requests with delayed responses.
Explain variability without hiding outliers
If a fictional sample contains confirmation intervals of 100, 110, 120, 130 and 540 milliseconds, its median is 120 milliseconds while its mean is 200 milliseconds. Publishing only the minimum of 100 would hide the slower observation.
These invented values illustrate reporting choices, not provider performance. State the measurement boundary, sample size and observed range. Keep interruptions and missing data visible instead of silently removing everything that makes the result look worse. A small selected sample is not a universal service benchmark.
Separate timing from execution rules
MetaTrader documentation distinguishes execution modes and order results. For a specific order, examine the applicable execution settings and accepted record rather than assuming the visible click price was guaranteed. For a fixed-outcome contract, separately check when its duration begins and which event defines expiry.
A chart interval, order duration and screen countdown need not share a start time. A late-looking animation cannot determine the accepted settlement rule. Our execution research academy covers the related price-source questions.
Frequently asked questions
Does a screen freeze prove late server acceptance?
No. The display can stop updating while other events occur. Preserve the transaction record and distinguish what was visible from what remains unknown.
Does a VPS guarantee better trading outcomes?
No. A different hosting location may affect a particular network path, but it does not guarantee fills, predictions or profitability. This guide makes no hosting recommendation.
What should a support ticket contain?
Include the request or contract ID, product, approximate time, timezone, recorded discrepancy and available error message. Do not include passwords or one-time codes.
Can latency alone prove manipulation?
No. Report the observed events and measurement limits. A disputed quote or rejection needs its own supporting records and applicable terms.
Original conceptual teaching visual. Illustrative values and schedules below are not current broker quotes, platform screenshots or verified provider behaviour.
A. Same-clock observation: 800 ms total
Measured end-to-end interval = 800 ms
Only the click and visible confirmation are timestamped on one observation clock. This interval can include network, queuing, processing and display work; the bar does not assign durations to those stages.
B. Internal acceptance is not observed
A separate server timestamp needs a known clock relationship before subtracting it from client time. Without clock-offset evidence, do not call the 800 ms server processing time or accepted-entry latency.
The values are invented. Use original records with clock identity, resolution and measurement method. A rejected order, delayed display and disputed accepted quote are different events. The diagram does not infer wrongdoing from a delay.
Sources and research scope
- TradingView: order execution and chart-price differences
- MetaTrader 5: virtual-platform ping metric
- MetaTrader 5: execution modes and order records
Public-source review: 25 October 2026. Examples are illustrative. Documentary research does not represent funded-account testing. Rules must be checked against the current product, entity and customer location.
Related research
Execution research academy · Research methodology · Calculation tools
Risk: Fixed-outcome contracts can lose the entire stake. No chart pattern, calculator or demo result guarantees future returns.