Trading Cost Assumptions That Kill Strategies

Every backtest that ignores costs is lying to you. A strategy that "makes money" in your backtest often loses money live because you forgot to model the friction of actually trading. This is why so many great-looking strategies fail in real accounts.

Commissions are the easy part. For US stocks, most brokers now charge $0. For options, expect $0.65 per contract at most brokers, plus regulatory fees. Multi-leg strategies like iron condors cost 4x that per round trip. Add these into your backtest explicitly — they compound.

Slippage is the hidden killer. When you place a market order, you rarely get the exact price you saw on screen. For liquid stocks (SPY, AAPL), slippage might be 1 cent. For illiquid options, it can be 5-10% of the option's value. Assume you get the mid-price minus half the spread — never assume mid-price fills.

For options specifically, the bid-ask spread matters more than commissions. A wheel strategy on AAPL might have $0.02 spreads. The same strategy on a small biotech might have $0.30 spreads. Your backtest must use actual historical bid/ask, not mid-price. Otherwise you will overstate returns dramatically.

A good rule of thumb: whatever your backtest shows, cut it by 20-30% for realistic execution. If the strategy still makes money after that haircut, you have something worth trading. If it barely broke even in the backtest, it will definitely lose money live.