Build a strategy and put it through the same statistical rigor checks our own research has to survive — walk-forward validation, placebo controls, a penalty that rises the more you've searched. Most ideas don't survive it. That's the tool working, not failing.
Then one verdict: PASS or REJECT.
Try enough variations and something will look profitable on past data, even when there is no edge in it at all. Below are twenty ideas with a true edge of exactly zero, tested the way most people test: once, on the data they built them on.
Two of them come out looking like winners. Nothing about them is real — and without a rigor check there is no way to tell them apart from an idea that is.
Two "winners" from pure luck. That is what a rigor check is for.
Every strategy faces the same five checks, in the same order, whether you built it or we did. There is no setting to turn one off.
part of the history is held back and never used to build or tune the strategy, so there is always data it has not seen.
the strategy is rebuilt and retested across successive windows, checking the result survives forward in time rather than only in the window it was fitted to.
compared against random entries with the same risk, costs, and exits, to check your entry signal is actually doing the work.
the bar you need to clear rises the more variations have been tried, so trying enough things until one looks good doesn't count as finding an edge.
concentration and dispersion checks make sure the result isn't being carried by one or two lucky trades.
PASSorREJECT
Nothing can be armed for live trading without a pass, and a pass is a statement about the evidence — never a promise about the future.
See what each check doesFour screens, in this order. Nothing can skip a stage, and nothing can be armed for live trading without a verdict.
Pick entry conditions, exit rules and optional filters from a guided library of blocks. No code, and every block explains in plain English what it does.
Illustration — figures are examples, not results.
Run it over years of real historical data and see the shape of the result: equity curve, trade count, expectancy per trade, drawdown.
Illustration — figures are examples, not results.
The full harness runs: out-of-sample, walk-forward, placebo controls, a penalty scaled to how much has actually been searched, and robustness checks.
Illustration — figures are examples, not results.
One outcome, stated plainly. A rejection tells you not to risk money on the idea — which is the whole point of running it.
VERDICT
REJECTED
Entry signal did not beat random entries with the same risk. Nothing can be armed from this result.
Illustration — figures are examples, not results.
Building and backtesting needs no broker at all. If a strategy passes and you decide to automate it, you connect your own IG account — we never hold your money, and your credentials are encrypted at rest, never visible to us in plaintext.
Nothing trades until you explicitly arm it, at the moment real capital is at risk. You can disarm at any time.
Build a strategy from entry, exit, and filter blocks — no code required.
Run it through the rigor harness against years of real historical data.
Get an honest verdict — and if it passes, choose whether to arm it for live automation.
An independent daily-loss circuit breaker sits beneath every armed strategy, regardless of that strategy's own settings. It can only make your risk tighter, never looser — and it disarms the strategy automatically if it trips.
Atlas won't let a strategy override the limits you set for yourself.
HMRC treats spread betting as betting rather than investing, so for most UK retail traders there is no capital gains tax, no income tax, and no stamp duty on spread bet profits. That is not an Atlas Edge feature — it is how the product is taxed, and it applies whether you trade manually or automate a strategy you built here.
It comes with real trade-offs, including the fact that losses cannot be offset against other gains, and it only applies to UK residents.
Read the full spread betting tax guideTax treatment depends on your individual circumstances and can change. This is not tax advice. The exemption applies to UK residents trading on their own account — it does not extend to anyone whose spread betting forms part of an existing trade, and losses cannot be offset against other gains the way CFD losses can. The 2025 Budget restructured gambling duty broadly but explicitly left spread betting's treatment untouched. If you're unsure how this applies to you, speak to a qualified tax adviser rather than relying on this summary.
Automated execution for UK accounts is in development and opening to a limited first group after launch. The backtest and strategy builder above are live worldwide right now.
No pitch, no guaranteed returns — just an honest test.
