Nearly every trading result you will see advertised is a backtest. Understanding exactly what that word means is the single most useful piece of due diligence available to a retail buyer, and it takes about five minutes to learn.
The difference in one line
- A backtest runs today's rules over yesterday's prices. The rules were written *after* the prices were known.
- A forward test publishes each trade *before* anyone knows how it ends, then keeps score.
Both can be honest. Only one of them is hard to fake.
Why backtests flatter
Not usually through fraud - through ordinary, well-intentioned iteration. Some of the common ways a backtest ends up better than reality:
- 1.Curve fitting. Test enough parameter combinations and one of them will look excellent purely by chance. Nothing was learned; a coincidence was found.
- 2.Survivorship in the rules. Each time a rule is added to avoid a loss that already happened, the test improves and the strategy does not.
- 3.Optimistic fills. Backtests fill at the price on the chart. Live orders pay a spread, and in fast markets - a news release, an opening gap - they can fill far away from it.
- 4.Ignored costs. Commission, swap and slippage are small per trade and enormous across a few thousand of them.
- 5.No gaps. Weekend gaps jump straight past a stop level. Depending on how the test is built, it may not model that.
What to ask about any published record
- Is it a backtest or live? If the page does not say, treat it as a backtest.
- Does it show the losses? A record with no drawdown page is a marketing page.
- What is the worst drawdown, and how long did it last? Depth tells you what you might lose. Duration tells you how long you would have to keep believing.
- Is every trade there, or a selection? The word to look for is *every*.
- How many trades? A hundred trades is noise. A few thousand at least starts to be evidence.
- Can it be checked? Entry and exit prices with timestamps can be verified against any chart. Pip totals alone cannot.
How we handle it here
The long record on this site is a backtest, and it says so on every page it appears on. It covers 2012-08-22 to 2026-08-24-979 closed legs for +150,481.9p at a profit factor of 1.97.
The worst drawdown inside it is +10,115p, from 2024-04-12 to 2024-08-12, recovered by 2025-04-06. That number is published as prominently as the profit, on the worst of it, because a results page that shows only the good half is not a results page.
Separately, the public forward test publishes each live call's direction while the trade is still open and keeps a running tally. It is much smaller, and it will stay much smaller for a long time, because that is how forward tests work. It is the honest one.