Quant Memo
Core

Market Making In Stressed Markets

Why market makers widen spreads, shrink size, or pull quotes entirely when volatility spikes, and why that behavior — while individually rational — makes stress worse for everyone else.

Prerequisites: Choosing How Wide to Quote, Inventory Limits And Position Caps

Market makers are supposed to be the steady hand — always quoting, always ready to trade. In calm markets that's true. In a sudden crash or a violent spike, the opposite tends to happen: spreads blow out, quoted size shrinks to a fraction of normal, and some market makers stop quoting altogether right when liquidity is needed most. This isn't a failure of the system so much as every individual desk doing the rational thing at the same time.

Why quoting gets dangerous in stress

The economics of market making depend on being able to estimate fair value and manage inventory reliably. In a fast, volatile market, both break down at once: fair value is genuinely uncertain because prices are gapping, and any position taken on is much harder to hedge or unwind before it moves further, because the market itself is thinner and faster-moving. A market maker that keeps quoting a normal, tight spread into a crash is effectively offering cheap insurance to everyone trying to get out — and will very likely be the counterparty stuck holding the position, at a bad price, once the dust settles.

Worked example: a spread response to a shock

In normal trading, a desk quotes a stock with a 2-cent spread and 5,000 shares of size on each side. News breaks that sharply changes the outlook for the company, and the price gaps down 8% in seconds. The desk's own fair-value estimate is now much less certain — is this an overreaction or the new correct level? — and inventory it picks up is harder to hedge because other market participants are also pulling back. Rationally, the desk widens its spread to 40 cents and cuts quoted size to 500 shares, or pulls quotes for a few seconds entirely while it re-establishes a fair value estimate. Multiply this by every market maker in the name doing something similar at once, and the visible order book — which might have shown $500,000 of depth within a few cents of the price a minute earlier — can shrink to a small fraction of that, right as the largest number of participants are trying to trade.

Path explorer
13055time →
end (bold path) 100.38spread of ends 58.966 independent paths, same settings

The explorer above lets you compare simulated price paths at different volatility settings; raise volatility and imagine every market maker along that path simultaneously deciding its normal spread is no longer safe to quote — that collective retreat is what produces a liquidity crisis on top of a price shock.

What this means in practice

Because this withdrawal happens simultaneously across many independent desks, it amplifies the very volatility that triggered it — a phenomenon sometimes called a liquidity spiral, where thinning quotes cause bigger price moves, which cause quotes to thin further. Exchanges build in mechanical circuit breakers and volatility auctions partly to interrupt this loop, giving market makers a pause to reassess fair value with a fresh, orderly price discovery process instead of quoting continuously into a freefall.

When volatility spikes, market makers widen spreads and shrink size because fair value becomes uncertain and hedging becomes harder — a rational individual response that, applied simultaneously across many desks, thins visible liquidity right when the market needs it most.

It's tempting to read a market maker pulling its quotes during a crash as abandoning its obligation to provide liquidity. Beyond a handful of formally "designated" market makers with regulatory obligations to keep quoting through stress, most market makers have no such duty, and pulling quotes when fair value is unknowable is a rational risk decision, not a moral failure — though it is exactly why liquidity evaporates precisely when it's needed most.

Related concepts

Practice in interviews

Further reading

  • Cartea, Jaimungal & Penalva, Algorithmic and High-Frequency Trading, ch. 10
  • Brunnermeier & Pedersen, Market Liquidity and Funding Liquidity
ShareTwitterLinkedIn