In early April 2025, the surprise announcement of a new U.S. tariff regime precipitated an increase in equity market volatility. This period saw the highest levels of market volatility since the COVID-19 pandemic leading to a bear market, where the E-mini S&P 500 futures declined by over 20% from its peak in February 2025. This downturn was subsequently followed by a substantial recovery in both asset prices and market liquidity.

During such periods of heightened volatility or market stress, a common observation among market participants is a perceived decline in liquidity. The conclusion is often drawn from a reduction in displayed order book depth. While a decrease in this metric is an expected response to increased risk during volatile episodes, relying solely on it can lead to an incomplete assessment of market liquidity.

A more comprehensive analysis requires utilization of a number of complementary metrics. For example, evaluating trading volume in isolation would likely lead to a different conclusion regarding the market’s liquidity, highlighting the limitations of focusing on a  single indicator. Therefore, a holistic understanding of market liquidity necessitates the analysis of several indicators beyond the top of the order book.

Figure 1 shows the period from April 3 to April 10, when the E-mini S&P 500 futures market experienced a substantial increase in trading volume. Volume on April 7 was over 99% greater than the average observed in the first quarter of 2025. This finding suggests a significant increase in market liquidity when assessed solely on the basis of trading volume.

However, this conclusion appears to be inconsistent with a narrow focus on order book depth, which often leads to a perception of declining liquidity during volatile periods.  This discrepancy highlights the limitations of using single metrics to evaluate market conditions and underscores the need for a more comprehensive, contextualized approach.

A more robust measure of market liquidity is an asset’s ability to absorb an order with minimal market impact, often referred to as fill quality. This metric accounts for how the execution price of a transaction relates to its arrival price.

The central limit order book (CLOB) serves as a repository for unmatched orders. Passive market participants, who supply two-way orders, contribute liquidity, while aggressing participants, who submit single-sided orders, demand it. In periods of high trading velocity, the rate at which market makers refresh their quotes at the top of the order book often accelerates to meet the increased demand from aggressing orders.

Therefore, a low level of resting volume at the top of the order book does not necessarily indicate a lack of liquidity. A high quote refresh rate can allow incoming buy and sell orders to be filled without a significant price change, even if the order book depth is lower than in a calmer trading environment.

The fundamental definition of a liquid market is one where a large volume of transactions can be executed without substantial impact on the price. Given this, it is not always accurate to conclude that a market is illiquid because fewer idle orders are present in the book. A more nuanced assessment requires considering the fill quality – the relationship between execution prices and arrival prices before making a judgment about illiquidity.

In conclusion, while liquidity in the E-mini S&P 500 futures market experienced a decline in April 2025, a holistic analysis incorporating fill quality suggests this decline was more moderate and understandable than a sole focus on order book depth would imply. When contextualized with multiple metrics and market conditions, it becomes evident that liquidity was still available but had been rationally repriced to reflect the prevailing risk conditions.

Figure 2 shows the U-shaped volume curve for E-mini S&P 500 futures – i.e., with heavy trading around the cash open (9:30 a.m. ET) and the close (16:00 ET) and less activity during the middle of the day. The surge in volume at the close is particularly notable and can be attributed to specific market participant behaviors, like month-end flows and trading mechanisms, such as Trade Market at Close (TMAC) and Basis Trade at Index Close (BTIC). Explore more information on these trading functionalities and the recent launch of S&P 500 Month-End futures.

Figure 3 provides a weekly analysis of average book depth for the same period.  The data reveals a decline in book depth at the top three levels in anticipation of the April 2nd “Liberation Day” tariff volatility.  Specifically, book depth during the 10:00 a.m. to 10:15 a.m. ET window decreased by approximately 27% by March 31st compared to the previous week.  This decline intensified during the peak of volatility, with a further 68% drop by April 7th

Interestingly, this effect was less pronounced near the close of the trading day. Book depth in the 15:45 p.m to 16:00 p.m. ET window was less impacted and recovered more quickly than in the morning, indicating a differentiated market response to volatility depending on the time of day. The relationship between liquidity, volatility and market impact costs will be contextualized later in this paper using the square-root estimated market impact model.

^{Source: CME Group}

To measure fill quality, we propose to first look at the price range of transactions that occur in close temporal proximity to each other. In a liquid market, the range would be narrow, whereas the range would be wide in an illiquid market. Since E-mini S&P 500 futures are continuously traded, one could group all the trades consummating within the same second and find out the trading range within each one-second interval.Since the prices only move in fixed minimum price increments, the number of prices traded within that one second interval signifies the trading range.

Figure 4 provides a comparison of price dispersion in relation to the rate of trading in a certain week. These weekly periods were taken at one-week intervals starting from March 17, 2025, through the week starting April 21, 2025. For example, if all the trades within the one-second interval have an identical price, the number of prices would naturally be one. If trades bounce between the top bid and offers without exhausting either, the number of prices would be two. Thus, the lower bound for the measure would be one. When the trading volume is light, this measure tends to average close to one. As the trading volume that occurs within a single second increases, the number of different prices consummated within that time frame also increases, as the top of the order book together with the refreshing of quotes at the top of the order book are insufficient to cater for the volumes being transacted.

Figure 4 illustrates where the x-axis denotes the rate of trading, i.e., number of contracts traded per second, while the y-axis denotes the level of price dispersion. At a low rate of trading, the price dispersion (i.e., the average number of traded prices) is low (one or two prices incurred). As the rate of trading increases, the top level of the order book gets depleted, and the price moves onto the next level. Thus, the average number of prices will increase. Below it applies this methodology to four different 15-minute time slices of the trading day.

^{Source: CME Group}

The y-axis serves as a proxy to the trading range whilst the x-axis denotes the trading volume per second. For example, in the week starting March 17, during the 15-minute window of each day for 9:30 a.m. – 9:45 a.m. Eastern Standard Time (EST) (blue line in the first chart of Figure 4), a trading rate of 206 contracts per second would on average have approximately a trading range of 4.9 prices (or ticks). In basis points this equates to 2.1bps of implicit execution cost for 206^[1] contracts ($59mm of delta), given 4.9 * the minimum price increment (MPI) of 25c for E-mini S&P 500 futures (1.2 index points) relative to E-mini price of 5,700 on that day during the window around the open.

Each line represents the estimated average price range within a second for each week during the time frame indicated in the heading of each chart. A wider dispersion of prices, after controlling for trading volume, indicates a less liquid market.

In a low volatility environment, like the week starting March 24, 2025, (honey colored line), the increase in trading rate is associated with an increase in the trading range. In a volatile market environment such as in the week starting April 7, (red line), the same increase in trading rate will be associated with a more significant increase in the trading range. This can be seen from the four panels in Figure 4, and this in isolation could be taken as an indication that the market in the first half of April 2025 was less liquid than in mid to late March 2025. By the end of April 2025 (brown line) the price dispersion calms down and reverts towards the late March environment.

However, this analysis is still incomplete and needs to be contextualized together with volume. The extent of the fill quality degradation is an interesting aspect to be examined. A set of complementary graphs to show the trading volume pattern will be introduced below to put the comparison in proper context.

^{Source: CME Group}

In each graph above in Figure 5, the y-axis represents the cumulative probability and X-axis represents trading volume per second. For example, during the 15-minute window each day of 3:45 p.m. – 4 p.m. EST in the week starting on March 17 (blue line, last chart of Figure 5), trading at approximately 33 contracts per second or less has a cumulative density of 0.50. In other words, the median or 50th percentile of trading volume per second is approximately 33 contracts per second. The same 50th percentile of per second trading rate is over 104 contracts for the week starting on April 7, 2025. The cumulative density function for the week of April 7 lies to the right of that of the week of March 17, meaning the rate of trading for April 7 week was much higher than that of March 17 week. Once volatility started to subside later in April 2025, we can see the rate of trading for the 50th percentile revert to 63 for the week of April 21, 2025, approaching levels seen in mid March. All these data points are detailed in the Appendix 1 for future analysis.

Figure 5 shows the range of trading for the weeks from March 27, 2025, to April 21, 2025. Referencing Figure 2, trading concentrates during the cash market open (9:30 a.m. EST) and the cash close (4:00 p.m. EST). The rate of trading is typically lower during the middle of the U.S. trading day.

For the week of March 17, the 50th percentiles of trading rate were approximately 41 (contracts per second) around the cash market open and 33 contracts around the cash close but 15 contracts at 11:30 and 12 contracts at 13:45, in the middle of the U.S. trading day. If we contrast this with the rate of trading in the week of April 7, the same 50th percentiles were on average over 213% greater than those for the week of March 17.

Equipped with the set of graphs in Figure 5 and the data points in Appendix 1, one observes for the April 7 that 90% of the time (i.e. the 90th percentile), the trading rate was approximately at 235 contracts per second for the cash market open and 364 contracts during the cash close window. The rate is at or below approximately 111 contracts per second around the middle of the trading day at 11:30 a.m. EST 90% of the time.

By using the trading rates found in the previous section for different points in the day and applying these quantities to the price dispersion charts shown in Figure 4, the magnitude of the increase in price dispersion between March 17, 2025 and April 7, 2025 can be identified.

For the open, the 90th percentile of average number of traded prices increased from approximately 4.1 found on March 17 for 140 contracts to 10.8 on April 7 for 235 contracts. In other words, 90% of the time the price dispersion for the week of April 7 (the most volatile week) was at most 6.7 ticks wider than the week of March 17 (a far less volatile week).  

Alternatively, for $59mm, using the 5,000 index price during the cash open window, we see 5.4bps impact costs for April 7.  Figure 6 details the comparable market impact for the equity market open window on March 17 for different percentiles, notionals and number of contracts. It is notable that for the same $59mm notional (96th percentile), market impact cost was 2.1bps on March 17 (i.e. 3.3bps lower than the same notional on April 7).

^{Source: CME Group}

The middle of-the-day sample (11:30 a.m. EST) produces similar results. At 62 contracts per second, the price dispersion for 90% of the time increased from 2.8 (11:30 a.m., March 17) to 111 contracts with a price dispersion for 90% of 7.9 ticks (11:30 a.m., April 7), an increase of 5.1 ticks at most.

By April 21, 2025, the price dispersion for 90% during the cash close for a trading rate of 225 had fallen back to 4.5 ticks, respectively, which is less than one tick more than found in the lower volatility week of March 17, 2025. This shows how the market adjusted to price risk more expensively, when volatility was at its peak in early April and how quickly the market repriced as volatility subsided. The relationship between volatility and cost to trade appears to hold as one would expect. All data points referenced are shared in Appendix 1 and will be compared to an estimated market impact in a later section.

It is also worth noting that the retracement in market impact costs for April is at odds with the conclusions from the book depth charts found in Figure 3, suggesting “fill quality” is a superior way to assess “liquidity.”

This section will compare two interesting points of time, specifically the volatility height of COVID on March 16, 2020 and tariff-palooza on April 7th 2025.  Both periods are defined by heavy trading volumes and increased volatility. Between the end of 2019 and 2025, the trading volume of E-mini S&P 500 futures and options contracts experienced a significant increase of 126%, as illustrated in Figure 7. This substantial growth, particularly when comparing the 2025 volatility to the COVID-19 period, highlights a new paradigm in risk management and trading activity. Concurrently, the notional of the E-mini S&P 500 futures contracts also doubled, rising from approximately 3,200 at the end of 2019 to over 6,400 at time of writing.

A comparison of the 90% price dispersion during the U.S. market open on April 7, 2025, to the peak volatility data points from the 2020 Assessing Liquidity report shows only a marginal increase in price dispersion, even with a significantly higher delta value. The observed market impact metrics are summarized below in Figure 8.

^{Source: CME Group}

When normalized for the increase in price and notional value, the market impact is even more telling. Using opening window prices of 2,400 on March 16, 2020 and 5,000 on April 7, 2025, the basis point impact for a $33 million notional trade in 2020 was 10 basis points, whereas a $59mm notional trade in 2025 resulted in 5.4 basis points. This demonstrates that despite the near doubling of contract prices or notional values, the market impact, measured in basis points, was approximately halved during the April 2025 volatility event compared to the COVID-19 period.

These differences in realized impact costs are primarily driven by the variations in underlying market volatility between the two periods. The following section will utilize a simple market impact model to compare estimated impact costs with these realized data points.

A foundational model for estimating market impact, widely employed by practitioners, is the square-root formula. This model, which has been extensively referenced by academics, such as Jim Gatheral of Baruch College, and can be expressed in a simplified form:

Estimated Impact = Spread cost + Factor * Daily vol * Square Root (Order Quantity / ADTV)  

This section evaluates the efficiency of the square-root model against empirical data from the E-mini S&P 500 futures market. The model’s parameters were calibrated using the following data:

• Daily volatility: The expected daily move, derived from options on E-mini S&P 500 futures, was averaged across the same four distinct time periods each day.
• Spread cost: The average bid-ask spread was utilized and averaged from each window.
• Order quantity: This metric represents the daily average of the 90th percentile order size per second, expressed in dollar notional terms.
• ADTV: A five-day average of daily traded volume was employed.
• Factor: The factor was held constant at one, given the E-mini S&P 500 futures’ significant liquidity. 

^{Source: CME Group}

Figure 9 compares the estimated market impact from the square-root model to the actual realized impact. The model demonstrates a reasonable performance as a “rule of thumb” predictor, suggesting that market participants rationally reprice market impact costs in response to heightened volatility.

However, the analysis also reveals potential limitations. Given the small sample size, further research is warranted. Preliminary observations suggest the model may underestimate risk during highly volatile periods (e.g. April 7) and overestimate it during calmer periods (e.g. March 24, April 14 and April 21). Future research could explore additional higher order risk metrics to improve the model’s predictive behavior. Furthermore, it would be valuable to investigate how trader behavior – specifically, the tradeoff between time risk and market impact costs – varies across different volatility regimes.

Within the square-root market impact model, the factor term is a critical component that accounts for differences in market microstructure across related products. A compelling case study for this factor involves a comparison between E-mini S&P 500 futures and the adjacent ETF market. Although these products share the same underlying benchmark, variations in market participants, trading venues and regulatory frameworks create distinct liquidity profiles.

As shown in Figure 10, the E-mini S&P 500 futures market has consistently traded at an average daily value of approximately 11 times the combined value of the three largest S&P 500 ETFs since 2020. This indicates a profound difference in liquidity.

Source: CME Group

The factor term can be used to align the market impact costs of the S&P 500 ETFs with those of E-mini S&P 500 futures. Using the square-root model in Equation 1, the relative volume metrics of 11 would point to S&P 500 ETFs estimated impact costs being 3.3x (the square root of 11) that of E-mini S&P 500 futures. However, this is likely too high when considering the premiums charged by liquidity providers for managing the full lifecycle of a trade, such as Exchange for Physicals (EFPs), creation/redemptions and Basis Trade at Index Close (BTIC), to name a few.

Figure 11 provides a detailed comparison of execution risk for both State Street’s S&P 500 ETF (SPY) and E-mini S&P 500 futures (ESM5) during U.S. cash trading hours only (9:30 a.m. to 16:00 p.m. ET). The data reveals that during these volatile periods, the trading volume ratio between ESM5 and SPY was 10, a figure consistent with the long-term trend presented in Figure 10.

^{Source: CME Group and Bloomberg}

Further analysis indicates that the average standard deviation of prices for SPY was approximately 2 basis points greater than that of E-mini S&P 500 futures. Although trading in SPY realized higher impact costs, the square-root model output of 3.3 times E-mini S&P 500 futures estimated impact cost is too high. This is where the factor term can be utilized to align estimated impact costs to realized impact costs.

The liquidity and capital efficiency of E-mini S&P 500 futures substantially benefits the adjacent ETF market. Likewise, futures are a beneficiary to the vast amounts of assets tracking the S&P 500 in products like ETFs. Future research could more precisely quantify the estimated versus realized impact costs across these markets to derive the factor in the appropriate factor.

This analysis provides a comprehensive framework for evaluating market liquidity during periods of heightened volatility, moving beyond traditional, single metric-assessments.  The findings demonstrate that relying solely on order book depth can be misleading, as this often declines during volatile episodes, leading to a simplistic conclusion of market illiquidity. A more robust approach, which incorporates metrics such as trading volume, fill quality and price dispersion reveals a more nuanced and rational market response.

Our examination of the April 2025 volatility event in the E-mini S&P 500 futures market illustrates several key points:

1. Divergent metrics: While order book depth decreased, trading volume increased substantially, highlighting a fundamental inconsistency between these two liquidity indicators.
2. Fill quality as a superior metric: The analysis of price dispersion and fill quality provide a more accurate measure of market impact. Our data shows that while fill quality did degrade during the peak of volatility, the degree was more moderate and directly related to the rate of trading and the rational repricing of risk.
3. Market consistency: Despite near-doubling of notional value since 2020, the market impact cost in basis points approximately halved during the April 2025 event compared to the COVID-19 period. This suggests a consistent relationship between market repricing of risk and different volatility regimes.
4. Model performance: The square-root market impact model serves as a useful predictive tool, though it exhibits some limitations, such as possibly underestimating risk during extreme volatility. Further research is needed to refine this model and explore additional actors that influence execution costs.
5. Inter-market dynamics: The study of the adjacent markets demonstrates that the E-mini S&P 500 futures market’s robust liquidity and capital efficiency significantly benefit the S&P 500 ETF market. The factor term in the square-root model, when appropriately calibrated, can effectively bridge the liquidity profiles of these arbitrageable products.

In summary, this study argues that liquidity in the E-mini S&P 500 futures market remained robust during the April 2025 volatility event, albeit with a rational repricing of risk. These findings underscore the importance of a multi-faceted approach to liquidity analysis and highlighted evolving resilience and efficiency of the futures market in managing risk.

The price dispersion charts are constructed in the following manner: for each second of trading, the volume traded in the lead month of the E-mini S&P 500 futures as well as the range of traded prices, expressed in the form of number of traded prices, are recorded. These data points are sorted into the corresponding 15-minute windows.

Using the data points collected in the 15-minute windows, the curve representing the average price range in relation to the volume traded per second was estimated by applying kernel regression techniques. It is akin to finding the weighted “local” average (i.e., the average price dispersion at a given level of trading volume rate) without imposing too many assumptions on the underlying probability distributions of the data.

Similarly, the cumulated density functions are produced based on kernel density estimation methodology.

^{Source: CME Group}

[1] 206 contracts per second on 17 March is chosen to compare to the results in Figure 6

1. CME Group
2. Three models of market impact, Jim Gatheral
3. Source: Quikstrike for ’30-Day ATM Constant Maturity’
4. Note that “calendar spreads” trades, or rollover volume, are excluded from the data, as those are position management trades that were separate entering and exiting positions in E-mini S&P 500 futures per se, the inclusion of which would obscure the “illiquidity” picture
5. The graphs are the cumulative limit order book quantity for E-mini S&P 500 index futures by month at the top three price levels. Average of the bid/ask quantities are sampled once per second for each of the fifteen-minute intervals from the lead-month futures outright order book
6. E-mini S&P 500 futures averaged 5,074 during the April 7 period