Opportunistic Intraday Liquidity Provision

A designated market maker has an obligation to post two-sided quotes continuously, in good conditions and bad. Plenty of other participants have no such obligation — they can post limit orders when they judge the odds of profiting from the spread are favorable, and simply not participate the rest of the time. That selective, condition-dependent approach is opportunistic liquidity provision.

Choosing when to quote

An opportunistic provider watches signals correlated with near-term adverse selection risk: has the order-flow imbalance in the book been building steadily in one direction (a warning sign, as in Order-Flow Imbalance)? Is realized volatility over the last few minutes unusually elevated relative to the recent average? Is a scheduled news release imminent? When these signals look calm, the strategy posts resting limit orders on both sides of the book, collecting the spread on the (hopefully) larger share of uninformed order flow that crosses. When signals look risky, it pulls its quotes rather than risk being run over by someone who knows something it doesn't.

Worked example

A stock's bid-ask spread is normally 2 cents wide with balanced order flow. A strategy quotes 1 cent inside each side of that spread, expecting to earn roughly 1 cent per round trip on the fraction of flow that isn't informed. When the order book's buy-side depth suddenly triples relative to the sell side over a few seconds — a sign, per Order-Flow Imbalance, that a large buyer may be working an order — the strategy pulls its offer (the side most at risk of being picked off by that buyer) while leaving its bid resting, or withdraws entirely until the imbalance normalizes. Once the imbalance signal fades back to its typical range, it resumes quoting both sides.

What this means in practice

The core trade-off is a version of the market maker's classic problem: quoting more often captures more spread on the calm days but exposes the strategy to occasional large losses on the days it guesses wrong about which periods are safe; quoting too conservatively protects capital but earns too little spread income to be worthwhile. Calibrating the signals that trigger pulling quotes — and how quickly to re-quote once they normalize — is most of the design work in this style of strategy.