Direct answer
Session liquidity can change quickly during volatile markets. The most common drivers are (1) wider price gaps when fewer orders are available near the current price, (2) higher latency so orders arrive later than the market state assumed at submission, (3) liquidity withdrawal when market participants reduce or stop providing quotes, and (4) order handling effects such as partial fills or rejected orders under stress conditions. These changes affect both visible spreads and the probability of getting filled at the expected price.
Definition and simple model of session liquidity
Session liquidity is the availability of buy and sell interest for a financial instrument during a particular trading session. A simple way to think about it is as “how many executable prices exist close to where the price is now,” and how quickly new executable prices appear when conditions change.
In a calm market, many participants are willing to quote or place resting orders. That creates a dense “order book” around the current price (even if you cannot see it directly). In volatile conditions, the same participants may adjust their behavior: they widen spreads, pull resting orders, or become less willing to quote because the risk of adverse price movement rises.
Separating stable mechanics from variable conditions helps: the mechanics of execution (latency, order arrival, matching, and partial fills) are stable ideas, while volatility and participant behavior are variable.
Why volatility changes it: gaps, latency, and liquidity withdrawal
1) Gaps: fewer nearby orders, larger price jumps
A price gap is a jump in observed price over a short time where intermediate executable prices are scarce. If volatility increases, the number of resting orders close to the current price can fall. Then, when the market needs an immediate match, it may “walk” to the next available executable prices farther away, producing a gap.
Material implication: your order may be filled, but not at the price level you expected when you submitted it. The fill can be at a worse level because the nearest available liquidity is now farther.
2) Latency: orders reach the market later than you think
Latency is the time delay between when an order is sent and when it reaches the matching process (directly or indirectly). During volatile markets, prices can move quickly, so even small delays matter.
Assumption for examples: suppose the mid-price moves by an amount that would cross multiple price levels within a fraction of a second. If your order arrives after that movement, it is evaluated against a different set of available orders. The market state at “send time” may differ from “arrival time,” changing fill probability and price.
3) Liquidity withdrawal: fewer quotes when risk rises
Liquidity withdrawal means participants reduce or stop providing the resting interest that supports tight spreads and fast execution. This can happen because of risk limits, inventory constraints, or operational throttling during stressful conditions.
Simple model: if quoting depth decreases at multiple price levels, the probability of immediate matches falls. Even if new quotes appear later, there can be a window where executable liquidity is temporarily thin.
4) Order handling: partial fills and rejections
Even without market “fraud” or unusual behavior, volatile conditions can trigger different order handling outcomes. For example:
- Partial fill: your order size may exceed the currently available executable quantity at the time of matching.
- Order rejection: an order can be canceled or rejected if it violates constraints enforced by the execution venue or intermediary (for example, due to rapidly changing conditions).
- Slippage: the executed price differs from an intended reference because the matching happens against the order set at arrival time.
Material limitation: the exact behavior depends on the execution pathway and venue rules, which can vary and are not identical for all setups.
Evidence or example you can test without live predictions
You can independently check whether session liquidity “changed” by using time-aligned data you already have (quotes and executions) and applying a consistent method.
Assumption for the method: you record (a) quote timestamps, (b) executed trade timestamps, and (c) spread or an equivalent measure of the distance between executable buy and sell prices.