Frequently Asked Questions

No. The numbers will differ, because the price data is not the same.

A backtest is arithmetic over bars. Two sources rarely have identical bars: a crypto index aggregated across venues differs from any single exchange’s prices, typically by a few tenths of a percent, and equity feeds differ in how they handle adjustments and late prints.

Differences that small still move which bar a trade opens on, and per-trade differences compound. We measured it on a representative strategy (SMA crossover with an ATR stop and target; BTCUSD daily; about eleven years), running the same strategy over two sources:

• About 90% of trades opened on the same bar, and about two thirds opened and closed on the same bars.

• Win rate and profit factor stayed close (47.9% vs 46.6%, and 1.16 vs 1.12).

• Compounded profit differed by nearly 2×, because small per-trade differences compound.

So the character of a strategy carries across sources; the exact equity curve does not. Treat a backtest as evidence about the strategy, not as a prediction of a specific number.

Two other differences: costs are off unless you add a Trading Costs node — commission is modeled when you do, slippage is not modeled at all — and a small number of indicators are still being verified against TradingView.

Commission: yes, if you ask for it. Add a Trading Costs node to your strategy and set the rate your broker charges — percent of each fill, a fixed amount per contract, or a flat fee per order. It is charged on both the entry and the exit, and every figure you see afterwards is net of it, including the equity curve, the profit factor, and the win rate. A trade that made money before fees and lost money after them counts as a loss, which is the point.

By default there is no such node, so results are gross of fees. TradingView’s Pine strategy default is also zero commission unless a non-zero commission is specified — and the exported Pine states which of the two it is in a comment, so you can tell at a glance. Export reflects the SAVED strategy, so save before you export if you have just changed the rate.

Slippage: no, not yet. Your real fills will be worse than the modeled ones, especially on thin markets and at the open. Nothing in Pinary currently accounts for that.

Pinary computes backtests on its own OHLCV (price) data.

For US stocks and ETFs, that is consolidated exchange data — prices aggregated across the exchanges where the symbol trades. Charting tools vary here: some show a single venue by default, which is one reason the bars can differ from ours.

For crypto, it is a composite index aggregated across venues, not the prices of any single exchange. Many TradingView crypto symbols represent a specific exchange, so the bars themselves differ — typically by a few tenths of a percent, which is enough to change some trades. There is no exchange you can pick that makes them identical, so we would rather say that plainly than have you hunt for one.

Letting you choose the data source is on our roadmap.