Your Bot Didn't Get That Price

Slippage is the difference between the price a bot's order targets and the price it actually fills at. It comes from order book depth, not from the exchange cheating anyone — a market order eats through whatever liquidity sits at each price level, and the average fill drifts away from the top-of-book quote as size goes up. Bitsgap's own backtesting data puts normal slippage at 0.05%–0.30% per trade, widening sharply in thin books and during volatility spikes. For a bot running dozens of cycles a week, that gap compounds into a real chunk of the return a backtest promised.

TL;DR

• Slippage = quoted price minus filled price, and it exists on every exchange, not just DEXs
• Order size relative to book depth is the main driver — the same order slips more on a thin pair than a liquid one
• Limit orders cap slippage but can miss the fill; market orders guarantee the fill but not the price
• GRID, DCA, and COMBO bots slip differently because they place different order types at different moments
• Backtests routinely underestimate slippage unless the fee/slippage assumptions are set realistically

What Slippage Actually Is

Slippage is the gap between the price a trader — or a bot — expects to pay and the price the order actually fills at. It shows up because an order book isn't one price, it's a ladder of prices with limited size sitting at each rung. A market order doesn't buy at the top of the book; it buys through the book, level by level, until the full order size is filled. The bigger the order relative to what's resting at each level, the further the average fill price drifts from the first number on screen.

This isn't unique to crypto or to bots. Any market with a finite order book produces slippage on size. What changes between assets and exchanges is how much liquidity sits close to the current price — and that's the number that actually determines the cost.

Where the Gap Comes From

Three things move the size of the gap on a given trade:

Order book depth. A pair with thin liquidity has few orders resting near the current price. A modest order can walk through several price levels before it's filled, and the average execution price ends up well past the quote. A deep book absorbs the same order with barely a dent.

Order size relative to that depth. The same $500 order that's invisible on a top-10 BTC pair can move the price meaningfully on a low-cap altcoin. Slippage scales with the ratio of order size to available liquidity, not with the dollar amount alone.

Volatility and timing. During fast moves, the book empties and refills constantly. An order sent a second too late can land against a book that's already thinner than it looked when the strategy checked it. This is why slippage widens during news events, liquidation cascades, and the first minutes after a major data release — the same order that costs 0.05% on a calm afternoon can cost several times that during a spike.

Bitsgap's backtesting data reflects this range directly: real slippage commonly runs 0.05%–0.30% in normal market conditions, and widens well beyond that during stress events and in thin order books — which is also why a backtest that ignores it can look profitable on paper and come out flat or negative live.

Market Orders vs Limit Orders: Where the Trade-Off Sits

The choice between order types is really a choice about which risk a trader accepts.

A market order guarantees the fill but not the price. It executes immediately against whatever liquidity is available, which is exactly why it's the order type most exposed to slippage — it takes the price the book offers, level by level, with no ceiling.

A limit order guarantees the price but not the fill. It sits on the book and only executes at the specified price or better. If the market never reaches that price, the order simply doesn't fill — which means zero slippage on that trade, but also a missed entry or exit if the setup depended on timing.

Most bot strategies use both, at different moments. A GRID bot places limit orders across its range by design, which is part of why grid strategies tend to slip less on entries — the price is fixed before the order goes out. A DCA bot's initial and averaging orders are typically limit-based too, but its stop-loss or emergency exit, if triggered, often needs a market order to guarantee the position actually closes. COMBO bots inherit both patterns, since they layer grid logic and DCA averaging inside one futures position, and add leverage on top — which raises the stakes on any market-order exit, since the same slippage percentage now applies to a larger notional position.

What This Costs in Practice

Slippage isn't a one-time cost — it applies to every fill, on both sides of a trade. A strategy that opens and closes a position slips twice: once going in, once going out. For a bot cycling through dozens of trades a week, that's dozens of small gaps between the expected price and the executed one, and they don't cancel out — they accumulate in the same direction the strategy is paying, not the direction it's earning.

This is the same reason a backtest and a live run of the identical configuration rarely produce identical numbers. A backtest that assumes zero or unrealistic slippage is testing a strategy against conditions that don't exist on a live order book. The fix isn't avoiding market orders altogether — it's knowing which parts of a strategy actually need one, and sizing positions so that the order itself doesn't become the thing moving the price.

Reducing Slippage on a Live Bot

• Trade liquid pairs. Top-20 by market cap on a major venue will consistently slip less than a low-cap pair, regardless of strategy.
• Size orders against depth, not against account balance. A position sized for a thin order book needs to be smaller than the same percentage allocation on a deep one.
• Prefer limit orders where the strategy allows it. Grid and DCA entries don't need to be market orders — the price is already defined.
• Read backtest slippage assumptions before trusting the return number. A backtest with slippage set to zero is not a forecast — it's a best case that live execution won't match.
• Expect wider slippage during volatility, not just during illiquidity. A liquid pair in a calm market and the same pair during a liquidation cascade are two different execution environments.

FAQ

What is slippage in crypto bot trading? Slippage is the gap between the price a bot's order targets and the price it actually fills at. It happens because an order book holds limited size at each price level, so a large enough order fills across several levels instead of one fixed price.

Does slippage only happen with market orders? Mostly, yes. A market order executes immediately against whatever liquidity is available, which is why it carries the slippage risk. A limit order only fills at the specified price or better — it avoids slippage entirely, but can miss the trade if price never reaches it.

How much slippage is normal in crypto trading? On liquid pairs in calm conditions, slippage typically runs from a few basis points up to a few tenths of a percent per trade. Thin or low-cap pairs, and any high-volatility event, push that range considerably wider.

Do GRID, DCA, and COMBO bots slip the same amount? No. GRID and DCA place most orders as limits, which caps slippage on entries by design. COMBO layers both on a leveraged futures position, so slippage on any market-order exit lands on a larger notional size.

How can I reduce slippage on a live bot? Trade liquid, high-volume pairs, size positions against actual order book depth rather than account balance, and use limit orders wherever the strategy allows. Expect wider slippage during volatility even on liquid pairs — depth thins fast when price moves quickly.

Why does a backtest show different results than live trading? A backtest without realistic slippage and fee assumptions tests a strategy against conditions no live order book actually offers. Real execution always includes the gap between quoted and filled price, which a backtest can estimate but never guarantee.