Abstract
Options markets are commonly analyzed through prices, implied volatility, trading volume, open interest, and bid/ask spreads. Yet these indicators provide only a partial view of market structure. An equally important dimension is the willingness of liquidity providers to maintain executable quotes.
When an option quote disappears, the event can reflect routine repricing, high-frequency order updates, inventory management, or an increase in adverse selection risk. These mechanisms have different economic implications even when their visible effects on the order book appear similar.
This article examines the economics of quote withdrawal, the importance of liquidity distribution across strikes and expirations, and the limits of interpreting order book changes as signals of future market risk.
The central argument is that liquidity is not merely a quantity. Its distribution, persistence, and underlying economic mechanisms also matter.
The discussion synthesizes established academic research and theoretical concepts. It does not present new empirical findings or claim that quote withdrawal has been established as a predictive trading signal.
1. An Option Quote Is More Than a Price
An executable quote expresses a participant's willingness to trade at a particular price and quantity.
For a buyer, the best ask represents the lowest currently displayed selling price. For a seller, the best bid represents the highest currently displayed buying price.
However, these prices do not fully describe the economics of providing liquidity.
An option market maker must consider several factors simultaneously:
- The probability that a quote will be executed.
- The impact of execution on the existing portfolio.
- The cost and availability of hedging instruments.
- Exposure to volatility changes, nonlinear price movements, and jumps.
- The risk of trading against a better-informed counterparty.
Consequently, a quote is more than an estimate of theoretical option value. It also reflects the price at which a participant is willing to accept additional exposure.
A market maker may consider an option fairly valued while becoming unwilling to provide the same quantity at the previously displayed price.
Key distinction
Fair value concerns valuation. Displayed liquidity concerns the current willingness to transact. These concepts are related, but they are not interchangeable.
2. Why Do Option Quotes Disappear?
The disappearance of an option quote is sometimes interpreted as evidence that market conditions are deteriorating.
That interpretation may be correct in some circumstances, but it is not generally valid.
At least four different mechanisms deserve consideration.
2.1. Routine repricing
When the underlying price changes, an option's theoretical value changes.
A market maker may cancel an existing quote and quickly replace it at a different price.
The original order disappears, but the market maker may continue providing similar liquidity at the new price.
Therefore, the disappearance of an individual order does not necessarily represent a persistent reduction in available market depth.
2.2. High-frequency quote updates
Modern electronic markets contain substantial numbers of short-lived orders.
Hasbrouck and Saar (2009) documented extensive rapid cancellation activity in electronic equity markets, consistent with dynamic strategies involving price adjustment and liquidity seeking.
In the options market, Ulze, Stadler, and Rathgeber (2026) investigated flickering quotes using high-frequency derivatives data.
Their findings associate these rapidly changing quotes with liquid trading environments and price-discovery activity in the markets examined.
The distinction between association and causation is essential. The results do not establish that rapidly cancelled orders always improve market quality or that the findings apply uniformly across markets and trading regimes.
Nevertheless, they demonstrate why a high cancellation rate should not automatically be interpreted as a sign of stress.
2.3. Inventory management
Suppose a market maker has accumulated substantial exposure to a particular option or group of contracts.
Further executions may increase unwanted portfolio risk.
The dealer may respond by reducing quoted size, widening spreads, or withdrawing liquidity on the side that would worsen the position.
Such behavior does not require a significant movement in the underlying price.
2.4. Adverse-selection risk
Liquidity providers face the possibility of executing trades against participants with superior information.
When the expected cost of adverse selection rises, maintaining narrow spreads may become less attractive.
A market maker may require additional compensation through wider spreads, smaller sizes, or selective withdrawal.
However, this does not necessarily imply knowledge of the market's future direction.
The dealer may simply be responding to greater uncertainty about the quality of incoming order flow.
Research distinction
Order cancellation, quote replacement, displayed-depth reduction, and persistent liquidity withdrawal should not be treated as equivalent events. Their economic meanings depend on execution, price changes, subsequent liquidity replenishment, and market conditions.
3. Measuring Liquidity Beyond the Bid–Ask Spread
The quoted spread of option i is
s_i(t)=A_i(t)-B_i(t)where A_i(t) and B_i(t) are its best ask and bid.
Its relative spread is
s_i^{rel}(t)= \frac{A_i(t)-B_i(t)} {\left[A_i(t)+B_i(t)\right]/2}.These measures are useful, but they do not fully describe executable liquidity.
Two contracts may display the same quoted spread while offering substantially different quantities at the best prices.
A broader measure therefore includes order-book depth.
Define displayed bid depth within a price distance \delta>0 of the best bid:
D_i^b(t;\delta)= \sum_{B_i(t)-\delta\leq p\leq B_i(t)} q_i^b(p,t),where q_i^b(p,t) is the displayed bid quantity at price level p.
Similarly, displayed ask depth can be defined as
D_i^a(t;\delta)= \sum_{A_i(t)\leq p\leq A_i(t)+\delta} q_i^a(p,t).If the previous bid depth is positive, a normalized depth-reduction measure is
W_i^b(t)= \max\left( 0,\, 1-\frac{D_i^b(t;\delta)} {D_i^b(t-\Delta;\delta)} \right).This measure ranges from zero to one and describes the proportional reduction in displayed depth over the observation interval.
It does not identify the cause of that reduction.
The same observation may result from cancellations, executions, repricing, or changes in the set of contracts being monitored.
Furthermore, depth within a fixed price band may change mechanically when the best price moves. That effect must be distinguished from a genuine change in willingness to supply liquidity.
For comparisons across contracts, one must also consider differing tick sizes, contract specifications, option prices, and market activity.
A measured reduction in depth is an observation—not an explanation.
4. Why the Distribution of Liquidity Across the Options Surface Matters
Options written on the same underlying asset share important sources of risk.
However, their sensitivities differ across strikes, expirations, and contract types.
Short-dated at-the-money options can exhibit substantial Gamma sensitivity. Longer-dated options may carry greater Vega exposure. Out-of-the-money puts may be particularly relevant to the pricing of downside protection.
These differences affect the economics of quoting.
Cao and Wei (2010) documented significant commonality in options-market liquidity, even after accounting for underlying-market liquidity and other determinants.
Their results support the existence of interconnected liquidity conditions across options.
They do not, by themselves, establish that the spatial distribution of liquidity predicts future asset returns or volatility.
That is a separate research question.
A conceptual example
Consider two groups of options:
Group A: short-dated out-of-the-money puts.
Group B: longer-dated near-the-money options.
Suppose displayed depth changes as follows:
| State A | State B | |
|---|---|---|
| Depth reduction in Group A | 80 units | 20 units |
| Depth reduction in Group B | 20 units | 80 units |
| Total reduction | 100 units | 100 units |
Illustrative quantities only; not observed market data.
An aggregate measure of total depth reduction assigns the same value to both states.
Yet the distribution across contracts differs.
The states may reflect different portfolio constraints, hedging costs, or liquidity-provision incentives.
The example does not demonstrate that either state predicts a particular market outcome. It shows only that aggregation can conceal differences between underlying configurations.
Central question
When two markets experience the same aggregate loss of displayed depth, does the location of that loss across strikes and expirations contain additional information about market risk?
The answer depends on the underlying economic mechanism and cannot be inferred from the aggregate measure alone.
5. Dealer Gamma, Vega, and the Cost of Providing Liquidity
A market maker typically manages portfolio-level exposure rather than evaluating each quote independently.
For small changes in the underlying price, volatility, and time, a local approximation of option value is
\Delta V\approx \Delta\,\Delta S +\frac12\Gamma(\Delta S)^2 +\mathcal V\,\Delta\sigma +\Theta\,\Delta t.Here, the Greek \Delta represents first-order price sensitivity, while \Delta S denotes the change in the underlying price.
This simplified expansion omits higher-order terms, cross-Greeks, and other effects. It is not a complete P&L identity.
Portfolio exposures may be expressed as
\Gamma_P=\sum_i n_i\Gamma_iand
\mathcal V_P=\sum_i n_i\mathcal V_i,where n_i represents the signed position in contract i, with consistent contract multipliers and units.
A new execution changes the portfolio's risk exposure.
Therefore, the decision to provide liquidity in one contract can depend on positions held in other contracts.
Stoikov and Sağlam (2009) examined how inventory risk, market incompleteness, and hedging liquidity influence optimal option quotes.
Baldacci, Bergault, and Guéant (2021) developed a model of algorithmic market making for portfolios of options, using volatility-related portfolio exposure to make the optimization problem tractable.
These studies provide theoretical foundations for cross-contract interactions in liquidity provision.
But publicly observable option quotes do not generally reveal every market maker's inventory, hedging constraints, or direction of exposure.
An observed reduction in depth therefore cannot be mechanically converted into a reliable estimate of aggregate dealer Gamma or Vega.
6. Can Quote Withdrawal Reveal Future Market Risk?
Consider a reduction in displayed liquidity across part of the options surface.
Several explanations may fit the same observation:
Information-related withdrawal: liquidity providers anticipate a higher cost of trading with informed counterparties.
Reactive withdrawal: dealers respond to changes already occurring in the underlying or related derivatives markets.
Inventory-related withdrawal: internal portfolio constraints make additional executions unattractive.
Common-factor response: both option liquidity and underlying-market volatility respond to the same information.
Without additional assumptions or evidence, the observation alone does not distinguish these explanations.
6.1. The information-theoretic perspective
Let
\mathbf W_t=(W_{1,t},\ldots,W_{G,t})^\toprepresent liquidity changes across option segments.
Define an aggregate indicator
L_t=\sum_g\omega_gW_{g,t},with fixed weights \omega_g.
Let R_{t+h} represent a future market-risk outcome and X_t the information already available at time t.
Because L_t is a deterministic function of \mathbf W_t, the conditional data-processing inequality gives
\boxed{ I(R_{t+h};L_t\mid X_t) \leq I(R_{t+h};\mathbf W_t\mid X_t). }Here, I denotes conditional mutual information.
The interpretation is precise:
Compressing a detailed liquidity vector into a single statistic cannot increase its information about future risk when the joint probability distribution is fixed.
But the inequality may hold with equality.
The detailed vector contains additional information only when the aggregate fails to preserve some risk-relevant variation.
6.2. Information content is not forecasting performance
A second distinction is equally important.
The information inequality does not imply that a forecasting model using every component of the liquidity vector will always outperform a model using the aggregate.
With finite observations, a simpler aggregate indicator may reduce estimation variance, model complexity, and overfitting.
A high-dimensional model may have greater theoretical access to information while producing less accurate out-of-sample forecasts.
Therefore, three questions must remain separate:
- Does the full liquidity distribution contain information absent from its aggregate?
- Is any additional information related to future market risk?
- Can a particular forecasting method use that information reliably after accounting for estimation error and trading costs?
The first is a mathematical question under a specified model.
The second concerns the joint distribution of liquidity and future outcomes.
The third concerns statistical and economic implementation.
None of these should be treated as automatically answered by the others.
7. Implications for Options Analytics
Implied volatility, open interest, volume, bid–ask spreads, and Greeks remain essential analytical tools.
However, none is a complete description of market liquidity.
A more careful analysis distinguishes:
- Changes in quoted prices from reductions in executable depth.
- Short-lived quote updates from persistent withdrawal.
- Contract-specific liquidity changes from cross-sectional effects.
- Dealer inventory constraints from market-wide changes in risk expectations.
- Statistical association from causal mechanisms.
- Theoretical information content from practical predictive performance.
The existing academic literature provides strong reasons to examine these distinctions.
It does not establish a universal rule that disappearing options quotes predict the next movement in the underlying asset.
Conclusion: The Market Structure Behind the Price
An option quote reflects both an economic valuation and a willingness to accept additional trading risk.
Its disappearance may represent routine repricing, algorithmic order updates, portfolio-risk management, or a change in the expected cost of providing liquidity.
These mechanisms cannot be reliably distinguished by counting cancellations alone.
Likewise, aggregate measures of market depth may conceal meaningful differences in how liquidity is distributed across contracts.
The essential analytical question is not simply how much liquidity disappears, but where it disappears, how long the change persists, and which economic mechanism produced it.
Whether those differences contain incremental information about future market risk remains a question requiring further investigation.
Understanding the distinction is a necessary foundation for analyzing options market microstructure without reducing complex market behavior to simplistic trading signals.
References
Hasbrouck, J., & Saar, G. (2009). Technology and Liquidity Provision: The Blurring of Traditional Definitions. Journal of Financial Markets, 12(2), 143–172. https://doi.org/10.1016/j.finmar.2008.06.002
Cao, M., & Wei, J. (2010). Option Market Liquidity: Commonality and Other Characteristics. Journal of Financial Markets, 13(1), 20–48. https://doi.org/10.1016/j.finmar.2009.09.004
Stoikov, S., & Sağlam, M. (2009). Option Market Making under Inventory Risk. Review of Derivatives Research, 12(1), 55–79. https://doi.org/10.1007/s11147-009-9036-3
Baldacci, B., Bergault, P., & Guéant, O. (2021). Algorithmic Market Making for Options. Quantitative Finance, 21(1), 85–97. https://doi.org/10.1080/14697688.2020.1766099
Ulze, M., Stadler, J., & Rathgeber, A. W. (2026). The Case of Fleeting Orders and Flickering Quotes. Journal of Futures Markets, 46(4), 629–652. First published online December 19, 2025. https://doi.org/10.1002/fut.70076
Mikhail Sukhov is a fintech product owner, market analyst, and founder of OptionsHub. His writing focuses on options, volatility, market microstructure, and the analytical foundations of financial decision-making.
AI tools were used to assist with editorial refinement, text structuring, and the creation of charts and illustrations for this article.
Research note: This article is a literature-based review with conceptual and mathematical illustrations. It contains no proprietary empirical results and does not establish a new predictive relationship.