TradingView Backtesting: Why the Tester, Other Engines, and Live Fills Disagree

Run the same strategy on TradingView's strategy tester, on an independent backtest engine, and on a live demo account, and you will usually get three different answers. That is not a scandal — it is the nature of simulation — but the size and direction of the differences tell you whether your strategy is real. This page explains where the gaps come from, based on our experience building trade-for-trade parity between the ForexEdge engine and exported Pine Script strategies.

Where the differences come from

• Data feeds differ. TradingView's forex candles come from its data partners; your broker's candles come from its own liquidity. Daily candles are the sharpest example: TradingView forex dailies open at New York close, while many data files use UTC midnight — the same strategy on the "same" daily chart can see different candles entirely.
• Intrabar assumptions differ. When a candle spans both your stop and your target, which was hit first? A tester must guess the path of price inside the bar. Different engines guess differently, and every guess is sometimes wrong.
• Native trailing stops are erratic. TradingView's built-in trailing (trail_points) simulates intrabar behavior that neither other engines nor most brokers reproduce. Close-based stop logic is far more portable across engines.
• Repainting indicators. Anything computed with future knowledge (unconfirmed pivots, higher-timeframe values that repaint) makes the tester's history quietly better than any live run could have been.
• Costs. Spread, commission and slippage settings rarely match between platforms — and live spreads vary by hour in ways no flat setting captures.

How to verify a strategy across engines

The honest procedure is trade-level comparison, not eyeballing two profit numbers. Export both trade lists, align them by entry time and direction, and measure: how many trades matched, how far apart the fills were in pips, and whether any trades flipped direction. ForexEdge automates this: build a strategy, export it as Pine Script, run it in TradingView, upload TradingView's "List of trades" CSV, and the comparison runs trade by trade. A strategy earns a parity badge only when at least 95% of matched trades agree within 2 pips, with zero direction mismatches, across at least 100 trades. Below that bar, the tool shows you which exit mechanism diverged — because "the numbers differ" is useless until you know why.

The three-stage reality check

Backtest → independent tester → live demo. Each stage strips away one layer of simulation. Watching the win rate move across the stages (say 65% → 61% → 58%) tells you how much of your edge was real and how much was the simulator being kind. This is the entire reason ForexEdge shows all three side by side per strategy. Related: what backtesting can and cannot tell you and how to automate a TradingView strategy safely.