Trading 101 treated slippage as a cost — the gap between expected price and filled price. That framing is useful for small trades, but it hides the more important truth: slippage is risk transferred from you to whoever fills you.
Who catches your slippage, and what they pay for it
When you submit a market order, you're not buying a good price — you're buying certainty of execution. The counterparty (an LP, a CLOB market-maker, an MEV searcher, another retail user) gets the difference between what you'd have paid in a perfect market and what you actually paid. From their angle, that difference is compensation for two real risks they bear:
- Inventory risk. They now hold the asset you sold. If the price moves against them before they can offload it, they lose money. Wide spreads (high slippage tolerance on your part) compensate for this.
- Adverse-selection risk. They don't know if you're an uninformed retail trader or a hedge fund who knows something they don't. If you're informed, they're going to lose. They widen the spread to insure against this.
Read those two sentences again. Together, they explain why the rational maker on the other side of your trade always quotes a wider spread for larger orders and during news events — both risks rise.
This is the same logic insurance underwriters use. Slippage isn't a transaction cost. It's an insurance premium. You're insuring yourself against the price moving while you wait for a better fill.
When you should pay it, and when you shouldn't
Two questions worth asking before any trade above €1,000:
Question 1: do I actually need certainty of execution?
If you're rebalancing into XRP because you think it'll outperform over six months, no — you don't need to fill in the next ten seconds. A limit order at mid-price might fill in five minutes for 80 bps less. The €40 you save on a €5,000 trade is the premium you would have paid for an insurance policy you don't need.
If you're closing a position because a security incident just hit your wallet's other tokens (a drainer is active, an exchange just halted withdrawals), yes — execution certainty is everything. Market orders exist for this.
Question 2: is the market quiet enough that the spread reflects only inventory risk?
In calm markets, AMMs and CLOB makers quote tight spreads because adverse-selection risk is low — most flow is dumb retail. In volatile markets — ECB days, US CPI prints, hack news on Twitter — spreads widen by 3-10× as makers price in the possibility you're informed. A trade that costs 50 bps on a Tuesday morning costs 300 bps thirty minutes after CPI.
The 301-tier rule: trades in volatile windows should default to limit orders. Trades in calm windows can take market orders up to ~1% of pool depth before the slippage premium becomes meaningful. Above that, slice.
What "slippage tolerance" really sets
When Gopnik asks you to set a slippage tolerance — "max 1%, 2%, 5%?" — you are setting the worst price you will accept. Higher tolerances do not give you a better trade; they give you a worse worst case. The trade fills at whatever price the curve produces at the time the transaction lands, which may be at or below your tolerance.
The mistake retail traders make is setting tolerance high enough that the trade "always fills." The mistake institutions never make is the same mistake. A trade that gets rejected because slippage tolerance was tight is a trade that would have lost money. The right reaction to a rejected swap is to take a breath, look at the chart, and decide whether the original opportunity still exists. Often it doesn't, and you've been saved.
The exit habit
At every tier from here up, build the habit of asking: what's the worst price I'm willing to fill at — not the price I expect, the price below which I'd rather walk away. Set the tolerance to that price. If the trade rejects, walk away. The 5% of trades that get rejected save you more, on average, than the 95% that fill. This is the entire risk-management discipline of professional execution compressed into one habit.