Market Maker Behavior Around Liquidity Events
Execution Precision
8 min read
Learn how market makers adjust their quoting and inventory management around major liquidity events.
Loading
8 min read
Learn how market makers adjust their quoting and inventory management around major liquidity events.
The market maker is not your adversary, but they are not your ally either. Understanding how they behave when liquidity shifts is the difference between getting filled at your price and getting filled at theirs.
A market maker is a firm that earns the bid-ask spread by quoting both sides of the order book and managing inventory risk; around liquidity events, they widen quotes, shrink size, or pause quoting to compensate for adverse-selection risk. Market makers profit from the bid-ask spread. They post limit orders on both sides of the book, buying at the bid and selling at the ask. In normal conditions, they provide liquidity, tighten spreads, and keep markets functional.
But during high-impact events -- FOMC announcements, CPI releases, ETF decisions, or liquidation cascades -- their behavior changes dramatically. They do not stop being market makers — they reprice the cost of providing liquidity. Spreads widen, displayed size shrinks, and some quote streams pause entirely. The effect on you is the same: liquidity becomes expensive or absent, and execution mechanics catch retail off guard.
Understanding these shifts is not optional for anyone trading through volatile conditions. This builds on Microstructure Shifts in High-Impact Events and feeds directly into Slippage Control & No-Trade Zones.
In calm markets, BTC/USDT perpetual futures on major exchanges might have a $0.10 spread. During a CPI release or a sudden liquidation event, that spread can balloon to $5-20 or more. This is not a malfunction. It is market makers widening their quotes to price in adverse-selection risk — the cost of being filled by someone who already knows the print (Glosten-Milgrom 1985). The wider the information asymmetry, the wider the quote.
Before a scheduled event, market makers pull their resting orders. A book that normally shows 50 BTC within $50 of the mid-price might thin to 5 BTC. This creates a fragile environment where even moderate-sized market orders can move price disproportionately.
Typical BTC/USDT perp spread by session phase
| Condition | Typical Book Depth (BTC within $100) | Spread | Slippage on 1 BTC Market Order |
|---|---|---|---|
| Normal session | 80-120 BTC | $0.10-0.50 | under $1 |
| Pre-event (30 min before) | 20-40 BTC | $1-5 | $5-15 |
| During event | 5-15 BTC | $5-50 | $20-100+ |
| Post-event (5-15 min after) | 40-80 BTC | $0.50-2 | $2-10 |
This is the signature market maker move around events. In the minutes before a scheduled announcement:
The result: the initial post-event move overshoots because there is nothing to absorb it. Then price snaps back as liquidity returns. This is why the first 1-3 candles after a major event are unreliable -- they reflect thin-book mechanics, not genuine price discovery.
Roughly half of the initial post-event spike is mechanics, not repricing. Real price typically settles 30-60% back from the extreme within 15-30 minutes.
When BTC drops $1,500 in 30 seconds after a CPI print, at least half of that move is book-thinning mechanics, not fundamental repricing. The "real" post-event price often settles 30-60% of the way back from the extreme within 15-30 minutes.
Caveat first: "market makers" is a category, not a cartel. Dozens of independent firms run independent inventory books with no shared playbook. The patterns below emerge because each firm faces the same adverse-selection math — not because they coordinate. You cannot see their positions directly, but the aggregate behavior leaves footprints:
BTC/USDT during CPI release. Initial dump to $63,800 on thin book. Spread widened to $15. Waited 12 minutes for book to rebuild. Entered long at $64,200 on limit order after bid-side depth returned to 40 BTC within $100. Spread had normalized to $1.50.
The initial $1,200 dump was 70% overshoot. Price settled at $64,500 within 20 minutes, then continued higher as the CPI data was actually neutral. Traders who shorted the initial dump got caught in the snap-back.
Beyond macro events, crypto has its own liquidity catalysts:
Each follows the same CEX MM playbook: pull liquidity before, overshoot during, re-enter after. AMMs (Uniswap-style) cannot pull — their liquidity is passive code. During liquidity events, LPs absorb the toxic flow and book impermanent loss; arbitrageurs drain the pool until on-chain price catches up to CEX. If you trade on-chain during events, you are usually the toxic flow against the LPs.
It does not always work. Sometimes the print confirms expectations and there is no overshoot — price just gaps and holds. Sometimes the snap-back never comes because the underlying repricing was real. Treat pull-and-fill as a base rate, not a guarantee, and never enter without confirmation.
They price in adverse-selection risk — the cost of being filled by someone who already knows the print. The wider the information asymmetry around the event, the wider the quote.
Market makers withdraw their resting limit orders before an event, wait for the initial directional move to play out on a thin book, then re-enter aggressively on the winning side. The initial overshoot is a side-effect of the pull, not a directional signal.
Wait for the bid-ask spread to return to within 2x its normal range and for displayed depth to rebuild — typically 10-20 minutes after the event print.
No. Slippage during thin-book conditions can exceed several days of normal trading profits. Use limit orders only, and only after spread normalization.
Continue: Build on this with Scaling Into vs Fading Volatility Spikes — once you can read MM behavior, the next question is whether to trade with or against the post-event move.