The gap between a beautiful backtest and a painful live account almost never comes from the strategy being wrong. It comes from costs and assumptions the backtest quietly ignored. The money didn't vanish — it was always going to be spent; the test just didn't charge you for it.
Spread. Backtests often fill at the mid-price or the close. Live, you buy at the ask and sell at the bid. On a strategy that trades often, paying the spread every round trip can erase the whole edge by itself.
Slippage. Your test assumes you got the exact price on the candle. Live, between your signal and your fill, the market moves — and it tends to move against you precisely when you most want in, because everyone else wants the same thing at the same moment. Stops are worse: they trigger into fast, thin conditions.
Commission and financing. Per-trade commissions and overnight swap are small line items that compound. A test that omits them reports a curve you were never going to earn.
Fill assumptions. The deadliest one. Did your backtest assume it bought the exact low of the candle? That it always got filled on a limit order that only touched? That it never gapped through the stop? These aren't small errors — they're the difference between a system and a fantasy.
The tell is simple: the higher a strategy's frequency and the tighter its edge per trade, the more of its "profit" lives inside these assumptions. A slow strategy making 80 pips a trade barely notices a 1-pip spread. A scalper making 3 pips a trade is mostly paying costs — and only the live account sends the invoice.
The first-principles takeaway
A backtest is a measurement of a strategy under assumptions. Change the assumptions to match reality — real spread, modeled slippage, commissions, honest fills — and most edges shrink, and many disappear. That's not a bug in your strategy; it's the test finally telling the truth.