Slippage in Crypto Trading: What Causes It and How to Control It

You see BTC trading at $67,000, you click market buy, and you get filled at $67,015. That $15 difference is slippage. On a tiny order in a liquid market, it's noise. On a large order in a thin book, or on a DEX swap, it can be the entire trade.

What slippage actually is

Slippage = (your average fill price) − (the price you saw when you clicked).

Two distinct components:

1. Latency slippage. The market moves between when you decide and when your order arrives. Even at retail latencies (100ms+), a fast- moving market can move several ticks in that gap. This is unavoidable on a CEX and modest in size, usually 1-3 bps.

2. Impact slippage. Your order is too big to fill at the best price level alone. It walks through multiple price levels of the book, and your average fill is worse than the level at the top. This is the dominant component for any non-trivial order size.

Most "slippage" you experience is impact slippage. It's not bad luck it's mechanical, and it's predictable from the depth of the book at the moment of the trade.

How impact slippage scales

Imagine a BTC order book where the asks look like this:

$67,005  ── 0.10 BTC
$67,003  ── 0.20 BTC
$67,001  ── 0.40 BTC
$67,000  ── 0.30 BTC   ← best ask

A 0.3 BTC market buy fills entirely at $67,000. No slippage. A 0.5 BTC market buy: 0.3 at $67,000 + 0.2 at $67,001. Average ≈ $67,000.40. Tiny slippage. A 1.0 BTC market buy: 0.3 at $67,000 + 0.4 at $67,001 + 0.2 at $67,003 + 0.1 at $67,005. Average ≈ $67,001.80. Now your slippage is $1.80 per BTC, ~3 bps.

The pattern is non-linear. Doubling the order more than doubles the slippage because you walk further into the (usually thinning) book. This is why splitting a big order into smaller pieces over time can materially reduce execution cost, not always, but often.

Slippage on DEXs is a different beast

On a CEX you walk an order book of resting limits. On an AMM DEX (Uniswap, Curve), you trade against a pool whose price is set by a formula, usually x * y = k. The math means every swap moves the pool's price, and the bigger the swap relative to the pool's liquidity, the bigger the price impact.

A 1 ETH swap in a $50M pool: ~negligible slippage. A 100 ETH swap in a $50M pool: 0.5-1% slippage from price impact alone. A 1000 ETH swap in a $50M pool: don't.

DEX UIs typically show "expected slippage" before you confirm. This is the impact your trade will have on the pool, calculated from the formula. A 0.3% slippage shown means your trade will execute roughly 0.3% worse than mid. Anything over 1-2% on a routine swap is a sign the pool is too small for your size, split the order or find deeper liquidity.

Slippage tolerance, the safety valve

DEXs and some CEX advanced order forms let you set a slippage tolerance: the maximum slippage you'll accept before the trade reverts. Set too low (0.1% on a volatile alt) and your trades fail. Set too high (5% by default) and you sign up for getting front-run by MEV bots that eat your tolerance as profit.

A reasonable rule of thumb on DEXs:

• Major pairs (USDC/ETH, USDT/USDC): 0.1-0.3%
• Mid-cap tokens: 0.5-1%
• Small caps in volatile conditions: 1-3%
• During known volatility (right after a launch, in a panic): widen consciously and expect to pay it

The key word is consciously. Don't leave the default 5% and forget it. That number is a license for MEV bots to extract your edge.

A common mistake: thinking the chart price is the price you'll get

The chart shows the last trade price (or mid). Neither is what you actually pay. On any non-liquid pair, your fill is materially worse than the chart suggests. New traders who size positions off chart prices find their actual entry is 50 bps off, then their actual exit is another 50 bps off, and they're confused why "the trade should have worked."

The discipline: before placing a market order on anything that isn't a top-tier pair, look at the order book or depth chart and estimate your true fill price. On Hex37 the workspace exposes the live book; this habit transfers directly to real exchanges.

A common mistake: panic-trading during low liquidity

Liquidity is not constant. It thins out:

• Overnight (Asia low-volume hours, ~04:00-08:00 UTC for many alts)
• Weekends, especially Sunday morning
• Right before scheduled news (CPI, FOMC, ETF announcements)
• Right after a big move (MMs widen quotes to manage risk)

The worst time to use a market order is the moment when slippage is highest, and that's exactly the moment most retail traders panic into one. If you're closing a losing position on a thin book in a fast move, you'll often pay 50-200 bps of slippage on top of your loss. The discipline is to size for that scenario in advance: have your stop set, accept the take fee, but don't also expect a tight fill on a panic exit.

Mental model, slippage as paying for traffic

Imagine the order book as a road. Driving 30 mph (small order) at 3 a.m. (deep liquidity), you cruise straight through. Driving the same speed at rush hour (thin book), you crawl through traffic and arrive late. Driving 100 mph (large order) any time, you'll catch every red light. Slippage is the time-and-distance cost of pushing through the market faster or wider than the road can absorb.

You can't eliminate it, every order has some impact. But you can choose your speed (order size), your time (when you trade), and your route (which venue, limit vs market) to minimize it.

Why this matters for trading

The strategies that work in backtests often fail in live trading because the backtest assumed mid-price fills and the live execution ate slippage on every trade. A strategy with a 30 bp edge per trade disappears under 40 bps of round-trip slippage. Knowing your true execution cost, measured, not assumed, is the difference between a strategy that survives and one that quietly bleeds.

Takeaway

Slippage is the gap between the price you expected and the price you got. It's mostly impact (your order walking the book), partly latency. On thin books, large orders, and DEXs, it can dwarf fees and spreads combined. Defenses: limit orders when you can be patient, smaller order chunks when you can't, and never trade size into known low- liquidity windows without accepting the tax up front.