Fees, Slippage and Leverage: The Silent Account Killers

2026-07-06

Plenty of strategies are profitable in a world without costs — and that world does not exist. Fees, slippage and leverage are not footnotes; for active strategies they routinely decide the sign of the final return. If your backtest ignores them, it is measuring a fantasy.

Fees compound against you

A crypto futures taker fee of roughly 0.04–0.06% per side sounds negligible. Round-trip, that is about 0.08–0.12% per trade. Now multiply by frequency: a strategy trading three times a day pays over 1,000 round trips a year — more than 100% of the account in annual fees at 0.1% each. Such a system must clear a huge hurdle before earning its first honest dollar.

The brutal arithmetic: if your average profit per trade is +0.15% before costs and the round trip costs 0.12%, your edge is not 0.15% — it is 0.03%, five times smaller. A small estimation error and the strategy is a slow-motion donation. This is why high-frequency signals with thin per-trade edges almost never survive costs, and why this site defaults to a conservative 0.12% round-trip fee you can adjust but should not ignore.

Slippage: the fee you did not agree to

A market order fills at whatever the order book offers. In calm markets on major pairs, slippage is small. But strategies that trade volatile moments — breakouts, panic dips, news candles — pay the most slippage exactly when they trade. The candle that triggered your signal is often the candle where the book is thinnest.

Assume slippage at least equal to your fee for market orders on volatile signals.
Limit orders avoid slippage but introduce missed fills — a different, subtler cost.
Small altcoins have wider spreads: the same strategy can be viable on BTC and hopeless on a microcap.

Leverage does not create edge — it multiplies whatever you have

Leverage scales profits and losses symmetrically, with one asymmetric addition: liquidation. At 10x, a roughly 10% adverse move wipes the position's margin (a bit less after maintenance margin). It does not matter that price recovered afterwards; the account did not live to see it.

The variable that decides survival is maximum drawdown, not average return. A strategy with a 30% historical drawdown at 1x becomes a liquidation candidate at 3x — and historical drawdown is a floor, not a ceiling: the future usually finds a deeper one. Size leverage from the worst single excursion your backtest ever saw, then leave a wide margin below the liquidation threshold.

Rule of thumb: decide leverage from the worst case, not the average case. If your backtest's single worst trade at 1x would exceed roughly half the distance to liquidation at your chosen leverage, the leverage is too high.

Position sizing: the Kelly connection

Once you know a strategy's win rate and its average win/loss ratio, the Kelly criterion gives the bankroll fraction that maximizes long-term growth: f* = W − (1−W)/R. Two things matter in practice:

Full Kelly is violently volatile, and your W and R are estimates from limited data. Most practitioners use half Kelly or less.
If Kelly comes out negative, the math is telling you the system has no edge at that win rate and ratio — no position size fixes that.

You can compute both on this site's calculators page, using the win rate and profit factor from your backtest results.

Before trusting any backtest: set fees to your exchange's real taker rate, assume some slippage on volatile entries, and check that the period table survives. A strategy that only works at zero cost is not a strategy — it is a fee-generation machine for your exchange.