Does EasyQuantTrading trade on behalf of clients?

No. EasyQuantTrading is a research and delivery OS. Live execution runs on your local account with your own keys. We do not custody funds or provide discretionary trading.

What makes backtests “honest”?

The engine is designed to avoid look-ahead bias and surfaces overfitting risk via honesty checks such as DSR/significance and deployability pipelines — including red-flagging untrustworthy strategies.

Is historical performance a promise of future returns?

No. All research output is for study and risk education. Past metrics are not investment advice and do not guarantee future results.

What membership tiers exist?

Free (research/GA limits), Premium (all engines and higher quotas), Diamond (full execution packs). See /legal for terms and /pricing for current prices.

Do all 105 fail for the same reason?

No. Some fail because the logic depends on intrabar assumptions that the terminal resolves differently. Some diverge gradually through cost modelling. Some mismatch on a small number of trades. Grouping them would hide the distinctions, so we report the categories separately.

Is 493 out of 1,307 a good rate?

It is a real rate for strategies produced by automated search rather than hand design. We publish it because a rate nobody can see is a rate nobody can check.

Could a different terminal give a different answer?

Yes, and that is the point. The check tells you the result is software-dependent. If it survives one real terminal and your own backtester, you have narrowed the uncertainty rather than eliminated it.

LEARN · EN · easyquanttrading.com

Your backtest passed. Then a real terminal disagreed.

Every strategy we keep has already passed our own backtest. That makes the next check the interesting one: does it survive a program we did not write, reading a feed we do not control? For 105 of them, the answer was no.

By the EasyQuant Research Team·Published 2026-09-26·We publish the tests our own strategies fail. Nothing here is a return promise.

A backtest and a live terminal are two different programs
Re-running on the real terminal is the cheapest lie-detector available
Divergence and mismatch are distinct failure modes, and both matter
This is a check most strategy listings never perform

What we did

We took 1,307 strategies that had already passed our in-house backtest and sent them to a real MetaTrader 5 terminal, running on its own broker feed, with the same symbol and timeframe.

Then we compared the result the terminal produced against the result our backtester had predicted, over the same window.

493 came back verified. 84 were weak. 62 failed outright. 29 diverged. 14 mismatched. The remaining 625 could not be run at all and are reported separately rather than counted as failures.

Why the two disagree at all

Your backtester and a broker terminal do not process the same bars. They may disagree on how a bar that touches both stop and target is resolved. They may model spread differently, or apply swap at a different moment.

Neither is necessarily wrong. But if your result depends on which of the two you happen to be running, then the result is a property of the software, not of the market.

This is the gap that makes a strategy look excellent in a report and behave differently the moment money is involved.

How to run it on your own list

You do not need our pipeline. Export the strategy, run it forward on the same instrument in a real terminal, and compare not just the final number but the trade list: entry times, exit times, direction, size.

Compare trade by trade rather than equity curve to equity curve. Two curves can end in similar places having taken completely different routes, and it is the route that tells you whether the divergence is harmless or structural.

Treat any mismatch as a finding about the strategy, not about the terminal.

What we do with the 105

They are not published, and they are not quietly deleted either. A strategy that fails its real-terminal check stays out of the library and the failure is recorded.

We would rather show a smaller number we can defend than a larger one we cannot.

How we measure What we publish

Current platform facts

Read live from the strategy library when this page was generated. These are the same counts published on our transparency page, and they change as strategies are added and rejected.

Strategies in the audited library	3672
Flagged by the audit	2011
Flag rate	54.8%
Checks still pending	1651
Passed the DSR overfitting check	1
Passed the significance check	504
DSR threshold used	0.90

FAQ

Do all 105 fail for the same reason?: No. Some fail because the logic depends on intrabar assumptions that the terminal resolves differently. Some diverge gradually through cost modelling. Some mismatch on a small number of trades. Grouping them would hide the distinctions, so we report the categories separately.
Is 493 out of 1,307 a good rate?: It is a real rate for strategies produced by automated search rather than hand design. We publish it because a rate nobody can see is a rate nobody can check.
Could a different terminal give a different answer?: Yes, and that is the point. The check tells you the result is software-dependent. If it survives one real terminal and your own backtester, you have narrowed the uncertainty rather than eliminated it.

More guides

← Back to home All guides Explore strategies

Not investment advice. Historical results do not guarantee future performance. EasyQuant is a research factory — you execute on accounts you control.