TradingView Backtester: test a strategy before it costs you money.

TradingView runs two kinds of Pine Script. An indicator draws lines, labels and alerts - it never simulates a trade, so there is nothing to backtest. A strategy declares itself with strategy() instead of indicator() and places simulated orders with strategy.entry and strategy.exit. Only strategies produce results.

To run the test: open your symbol and timeframe, add the strategy from the Pine Editor (or the community library), then open the Strategy Tester tab at the bottom of the chart. You get the equity curve, net profit, profit factor, max drawdown and the full trade list. If you wrote the rules as an indicator, you are not stuck - the entries and exits translate into a strategy almost line for line, or you can go straight to alerts and validate in real time instead (more on that at the end).

A backtest with zero costs is marketing, not research. In the strategy settings set a commission that matches your broker (a per-lot figure for forex, or a percentage for crypto) and a few ticks of slippage on every fill. These two fields alone routinely turn a 60%-win-rate wonder into breakeven - better to learn that now than from your broker statement.

Two more honest-makers. The bar magnifier (paid plans) uses lower-timeframe data to guess what happened inside each bar, so stops and targets within one candle resolve in the right order - without it, TradingView assumes a path that may flatter your strategy. And deep backtesting on higher paid plans unlocks the full symbol history instead of a capped bar window, which matters because a strategy tested only on the last trending year has not really been tested. Free plan limits change, so check TradingView's current plan page for exact bar caps.

The most dangerous backtest bug is not in your logic - it is in your timing. A script repaints when its historical signals were computed with information that was not available at that bar. The equity curve looks supernatural because, in a sense, it was.

A backtest, even an honest one, is still a simulation on historical candles. It cannot show you live spread widening at the news, the half-second between your alert and the fill, or how your strategy behaves when three signals fire in the same hour. The cheapest reality check is a forward test on an MT5 demo account: same rules, live market, fake money, real fills.

The practical way is to convert your strategy rules into TradingView alerts - one message per entry and exit - and let them execute automatically on the demo. That is exactly what SignalForge does: it routes your TradingView webhook alerts to MetaTrader 5 in milliseconds, with stop-loss, take-profit, multi-TP, break-even and trailing handled by the EA. You can start with the 14-day free trial, no card required, and watch a full cycle of signals fill on demo before a single real dollar is at risk. If you are choosing between bridges, our honest TradingView to MT5 bridge comparison lays out the options.

Treat the forward test as part of the strategy, not a formality: two to four weeks on demo, compare live fills against the backtest's assumptions, and only then decide. If you want the alert JSON right the first time, our guide on Pine Script alert webhooks covers the message format, and copying TradingView signals to MT5 walks the full pipeline end to end.