This article explains execution costs. The worked example is an illustration of order-book mechanics, not a measurement of any specific market. Nothing here is investment advice. Decisions are yours.
Slippage is the gap between the price that justified a trade and the price the trade actually got. It is paid on every entry and every exit, it grows with size and urgency, it explodes in stress, and it appears in no screenshot, which is why the returns people show you and the returns people earn are different numbers.
The three components
Slippage decomposes cleanly. The spread: the standing gap between best bid and best offer, paid by anyone who crosses it, tight on majors and canyon-wide on the tail. Impact: an order bigger than the best quote walks the book, consuming successively worse levels until filled. Timing: the market moves between decision and execution, a tax that grows with latency and with how predictable your order flow is. On a deep book the three sum to basis points; on a thin alt book they sum to percents, and the difference is invisible until you trade size.
Walking a thin book, worked
A stylized but honest illustration. An alt shows a last price of $1.00 and respectable printed volume. The book behind the print: $2,000 offered at $1.001, $3,000 at $1.005, $5,000 at $1.02, $10,000 at $1.05. A $20,000 market buy consumes all four levels and finishes paying about 3% above the price that motivated it; the average fill sits near $1.03. Nothing malfunctioned: the trade simply met the market's real depth, which the last-price ticker never displays. This is why printed volume and tradeable liquidity are different statistics, a gap the dollar-volume piece measures directly, and why execution costs scale non-linearly with size: the second $20,000 fills worse than the first.
Where slippage actually bites: exits and stress
The asymmetry that matters for any systematic style: entries choose their moment, exits get chosen for. A stop triggering during a cascade sells into a book that market makers have already abandoned, so the fill lands far below the stop level, which is how a liquidation cascade does its damage and why stop arithmetic and stop reality diverge. That divergence is what the ruin arithmetic has to absorb: planned loss fractions are exceeded precisely on the days that decide survival. Stress slippage is also marketwide rather than idiosyncratic, because depth evaporates everywhere at once when the one factor turns, and the sector correlations turn with it. The bad day's book is the one to size against; the good day's book is a courtesy.
What honest backtests do about it
No daily-bar backtest can simulate an order book, so the honest ones charge a disclosed toll and state its limits. Our transplant experiment charged 20 basis points per side: a defensible average for liquid names that certainly understates stressed exits on thin ones, which is one reason its results are presented as upper bounds. The general auditing rule travels: any backtest quoting returns without naming its cost assumption has embedded a zero, and any live track record on thin names that matches its backtest too exactly deserves the question of how its exits filled, one of seven checks worth putting to any provider. Costs are where paper strategies go to die; disclosure is the only difference between a limitation and a lie.
The mitigation craft
Practical, unglamorous, effective. Trade a small fraction of real depth, not of printed volume, and accept that some signals are simply not tradeable at your size. Prefer limit orders where the strategy tolerates missing fills, and know that stops cannot be limits without accepting gap-through risk. Split orders over time rather than demanding immediacy the book cannot supply. Avoid the thin hours: a market that never closes still has a depth cycle, and the same order costs more at the empty end of it. And let liquidity into the universe decision itself, which is the job of the published liquidity floor in the universe construction rules: the cheapest slippage is the trade you were never tempted to place. Decisions are yours.