Performance

We publish verified results only

There are no performance figures on this page yet. We hold every strategy to the standard below, and we will publish a result only when one meets it - with the method and the data it was measured on.

Every module we tested is in the evidence register - the tests it passed or failed and the numbers behind them, failures included. Live forward signals are measured in the Telegram record.

Out of sample
A result counts only if it was measured on data the model never saw while it was being built.
After costs
Spread, commission and financing are included. A result that only exists before costs is not shown.
Rules written first
The test and its pass mark are written down before the data is read, so a result cannot be tuned after the fact.
Against chance
Every result is compared with random entries at the same times. Beating the market is not enough; it has to beat luck.
What we are measuring now

Every Copilot signal is recorded before its result is known - entry, stop, target and its Consensus Score - and then its outcome on the real price path, wins and losses alike. IEB signal modules keep the same kind of record, including the version of the module that made each signal, and publish it live on Telegram. We keep the Consensus Score (how strongly the analysts agree) apart from the Historical Outcome Rate (how often comparable signals reached their target before the stop); a band with too few signals is reported as “insufficient sample”. What a signal is worth is measured separately from what it costs to trade, because costs depend on your broker. Aggregate figures will appear here, with their definitions, once there is enough data to mean something.

Follow the live signal record on Telegram →

Trading involves risk of loss. Our tools help you analyse markets and manage risk; they do not promise profits.

See the products →