What Trading Costs Actually Are
Every live trade incurs three distinct costs, which a rigorous backtest models separately.
Commissions. The broker's fee per contract. For retail CME futures, round-turn commissions at typical futures brokers range from roughly $2.00 to $4.50 per contract depending on volume tier and account type.
Exchange and regulatory fees. CME levies an exchange fee per contract (around $1.14 for ES, $1.18 for NQ), plus a small NFA regulatory fee (~$0.02) and an additional pass-through fee on market-data-generating trades. Total round-turn: roughly $1.50–$2.60 in fixed fees depending on product.
Slippage. The difference between the expected fill price and the actual fill price. For a market order into a liquid futures instrument, slippage has two sources: the bid-ask spread (which you cross when taking liquidity) and market impact (the price movement caused by your own order).
Why Slippage Dominates
Commissions and fees are deterministic. You know them in advance and they can be modeled exactly. Slippage is probabilistic and instrument-specific, and for most retail strategies it is the largest component of total cost.
In a liquid CME instrument like ES, the bid-ask spread is almost always exactly one tick ($12.50 on a full-size ES contract, $1.25 on MES). A market buy order typically fills at the ask; a market sell at the bid. The implied cost per round trip for crossing the spread twice is one full tick. The available-liquidity baseline matters here — CME Q1 2026 set a record 36.2M-contract ADV with all six asset classes printing records simultaneously, which deepens the resting book and reduces the impact-cost component of slippage on size, but only for strategies that recalibrate their participation parameters to the new baseline.
In less liquid instruments or during fast market conditions, slippage can be materially worse. During the open, a macro release, or a news spike, a market order can slip two, three, or more ticks. Strategies that trade during volatile periods need to model this.
Modeling Slippage in a Backtest
Several approaches exist, in rough order of increasing sophistication:
- Fixed tick model. Assume one tick of slippage per trade, doubled for round trip. Simple, conservative for liquid retail-sized orders, and a defensible default for futures systems trading ES, NQ, CL, GC at retail size.
- Spread-based model. Sample the observed bid-ask spread at the trade timestamp and cross the full spread on entry and exit. More realistic for instruments whose spreads vary.
- Volatility-scaled model. Scale slippage by a measure of instantaneous volatility (e.g., a short-window ATR). Matches the empirical pattern that slippage is worse when volatility is high.
- Order-book replay. For institutional-grade research, simulate against historical order book depth. Out of reach for most retail strategy work because of data cost and complexity.
For the vast majority of retail futures backtests, the fixed tick model is fine. The question is less about perfecting the slippage estimate than about including one at all — many retail backtests assume zero slippage, which is a guaranteed source of optimism.
Why Turnover Is the Silent Variable
Total cost scales with trade frequency. A system that enters and exits once per day incurs cost once per day. A system that takes 30 trades per day incurs cost 30 times. For a strategy with a modest per-trade edge — say, 0.2 ticks average — a system can be gross-profitable and net-unprofitable purely because of turnover.
This is why many academic high-frequency alpha studies fail to replicate in live trading: the strategy's signal is real, but the signal is smaller than the cost of capturing it at retail execution levels. The institutional version of the same signal, with lower latency and favorable liquidity rebates, remains profitable. The retail version does not.
Conclusion
The difference between a profitable backtest and a profitable live account is almost always rooted in cost modeling. Commissions are knowable. Fees are knowable. Slippage is estimable if you are willing to be conservative. A strategy whose backtest includes all three honestly and still shows edge has a fighting chance live. A strategy whose backtest assumes any of these are zero is measuring the wrong thing.
Disclaimer: FalcoAlgo is a software product of Falco Systems LLC and is not a registered investment adviser. This article is for educational purposes only and does not constitute investment, trading, tax, or legal advice. Futures trading involves substantial risk of loss. Hypothetical performance results have inherent limitations and are not indicative of future results.
