How to Read the Backtest Report

The giant percentage is TOTAL RETURN — what the strategy made or lost, compounded across every trade, over the whole test window (the "over N days" line right under it tells you how long that was). Green for a long strategy, red for a short.

One thing to know up front: this number is already net of costs. Trading fees, slippage, and funding payments were all deducted inside the simulation — the cost chip at the bottom of the return card is the receipt.

The grid beside the return holds twelve cells. The first four just echo your picks — Symbol, Side, TF, Leverage. The other eight are the strategy's vital signs:

CAGR — the return restated as a steady yearly pace. · Sharpe — return earned per unit of shaking along the way. · Sortino — the same idea, but only counting the downward shakes that actually hurt. · PF (profit factor) — dollars won for every dollar lost; above 1 means profitable.

MDD (max drawdown) — the deepest valley measured from a peak; the number that tells you how it would have felt to hold. · Win Rate — the share of trades that closed green. · Trades — how many chances the stats are built on; more trades, more trustworthy numbers. · Avg Hold — how many bars a typical trade stays open.

If you only read three: PF, MDD, and Trades. Together they say "did it win, how much did it hurt, and is there enough evidence."

Under the return sits a one-sentence verdict — a plain-English description of how the strategy actually trades, written automatically from its rules ("Enters on moving-average signals, then rides the trend…"). It's the card introducing itself.

Below that sits the rarity seal, grading the card from Normal to Legendary — earned by score, not rolled — and under it the four verification badges: Walk-Forward, Out-of-Sample, Robustness, Monte-Carlo. Every card in the vault has already cleared all four — hover any badge for a one-line explanation of what it tested, or read The Four Gates.

Everything above the fold is the backtest — how the strategy did on the history it was built and graded on (its in-sample life). The Since Launch panel underneath asks the harder question: how has the same frozen strategy done on the candles that came after it was created — the stretch it never saw during the backtest (out-of-sample)?

It's a simulation replayed on real candles since launch, with no re-fitting — not live trading, no real orders. A backtest that looked brilliant but faded once the strategy went live is exactly what this panel exposes, so you can judge "great on paper — but does it hold up?" for yourself. The start → end · N days label tells you how long the out-of-sample window is — and the Trades cell how much evidence it rests on; a handful of trades over a couple of weeks is a hint, not a verdict.

The Since Launch panel refreshes once a day. Every strategy's out-of-sample record is recomputed from the latest candles each day, so the window keeps stretching and the numbers move as new trades close — it's a living track record that grows for as long as the strategy runs, not a one-off snapshot.

Under the Return figure sits a small vs hold line — what simply buying the coin and holding it would have returned over the exact same window, spot, with no leverage. It's deliberately unleveraged even when the strategy runs 3× or 5×: a "leveraged hold" isn't a real benchmark — it can be liquidated on the way down and pays funding the whole time, so multiplying the hold line would be comparing you to a fantasy. Unleveraged holding is the one alternative anyone can actually do. The line is there because a young record is easy to misread: +2% in three weeks looks timid on its own, but next to a coin that fell 4% over the same stretch it reads very differently. Holding is the benchmark every strategy has to beat — and on the days holding wins, the line says so just as plainly. If the line is missing, we simply don't have the candles to measure it.

If a strategy stops updating, a reopened card can show a Retired or Stopped chip — its final Since Launch record is kept, not hidden. In that case vs hold is measured over the same frozen window, up to the day the record stopped. Managing Your Saved Strategies covers the shelf, the toggle, and both labels.

The chart shows the real price candles the strategy traded, with every trade stamped on top. A gold LAUNCH line splits it in two: to the left, the greyed-out backtest window; to the right, full-colour candles from since launch. Entry arrows follow the same code — gold for backtest trades, teal for since-launch trades — and each exit is labeled with its profit or loss. Scroll and zoom to walk through it trade by trade; click a row in the trade log to jump straight to that trade.

The green FILTER toggle shades the background wherever the strategy's entry filters were all satisfied — the zones where it was allowed to buy. The shading is computed with no look-ahead: the green zone marks the bars where an entry could actually fill, because the filters said yes on the bar before — exactly like the backtest itself.

Below the chart, the Trade Log lists every single trade — entry time, in and out prices, P&L, and the reason it exited (stop-loss, take-profit, trailing stop…). Backtest and since-launch trades share the table in time order: since-launch trades are marked in teal, backtest ones sit quietly greyed. Next to it, the Blueprint panel opens the strategy itself: its entry signals and exit rules — each with a plain-English tooltip — plus its parameters.

Want the raw data? Export Trades · CSV downloads the full log — with a phase column tagging each row backtest or live. Like the card? Save Strategy names it into My strategies so you can reopen the full report anytime — or take it live later.