Original title: 432 Hours of Hyperliquid Oil Market Data: A Microstructure Comparison with CME WTI Futures

Original source: Castle Labs

Original translation by Golem, Odaily Planet Daily

Editor's Note: This article systematically studies the differences in crude oil contract trading data between Hyperliquid and CME during weekday and weekend sessions, and draws some important conclusions. Currently, Hyperliquid cannot match CME in absolute indicators such as liquidity depth or slippage, with overall liquidity below 1%. This is related to the fact that the main users of the RWA trading platform are still crypto-native retail investors.

What sets Hyperliquid apart is the significant increase in trading volume for crude oil contracts over the weekend. This indicates that in addition to retail traders with speculative needs, traders looking to gain exposure to crude oil and hedge their positions before Monday are also trading on Hyperliquid. This trend is becoming increasingly apparent, demonstrating that Hyperliquid has developed price discovery capabilities for commodities.

However, for institutional investors, Hyperliquid's high transaction costs compared to CME remain a major obstacle to its expansion in the commodities trading sector. If Hyperliquid doesn't develop the capability to handle institutional-level orders soon, it will remain merely a temporary trading venue for traditional traders on weekends, ultimately becoming a small supplement to the traditional financial landscape. (The following is the original text:)

Research Methodology and Data Sources

This analysis assesses the microstructure of the crude oil market through two studies, covering both weekday and weekend markets, and utilizes tick-by-tick data from two exchanges: Hyperliquid's xyz:CL perpetual contract and the Chicago Mercantile Exchange (CME) CLJ6 (April 2026 NYMEX WTI crude oil futures) contract.

CME data originates from the Databento trading data source, which captures tick-by-tick trades rather than order book snapshots. Therefore, all CME depth and slippage estimates are based on actual trading volume, not quote depth. Hyperliquid data comes from Hyperliquid's publicly available S3 database, which contains complete on-chain transaction records.

Therefore, the analysis for both trading venues is based on actual trading volume. All depth data represents explicit liquidity, i.e., trading volume within a specific basis point range within a 5-minute window of the VWAP midpoint, rather than the full resting depth on the order book.

Research period and market background

The research period is from February 27 to March 16, 2026, a period that coincides with the geopolitical turmoil following Iran's attack on February 28, 2026.

• Market closing before the attack: The last CME trading day before the attack.

Monday's opening: Market reopening faced significant pressure, with CME shares opening sharply higher, while the Hyperliquid xyz:CL market was constrained by discovery boundaries.

• Over the following weekends: High oil prices and market volatility led to high crude oil trading volumes on the Hyperliquid platform.

xyz: CL launched in early 2026, meaning this three-week observation period covered the early maturation phase of the Hyperliquid market. The observed trends, including improved liquidity depth, increased trading volume, and user growth, partly reflect market maturity. However, we believe that on-chain trading platforms currently cannot compete with traditional trading platforms in absolute metrics such as liquidity depth or slippage.

Our research aims to track directional trends: whether the price difference is narrowing, how fast it is narrowing, and under what conditions it is narrowing.

Data Analysis

The data analysis is divided into two parts based on the time period:

• Weekday Trading Hours: Covering a full three-week period, this analysis compares the market depth, slippage, and premium/discount for Hyperliquid and CME trading during weekday trading hours. For Hyperliquid, we also analyzed its funding rates throughout the entire trading period.

• Weekend Period: Within a given timeframe encompassing three weekends, we analyzed price discovery and the price gap deviation of Hyperliquid relative to the CME opening price.

Weekday Time Period Data Analysis

This analysis covers a full three-week period, with a focus on the periods when both trading platforms are active simultaneously.

Liquidity depth is measured by USD trading volume within ±2, ±3, and ±5 basis points of the VWAP midpoint per 5-minute interval, and aggregated as the median across all business day intervals. As mentioned above, this reflects volume within the interval, not resting quote depth. This approach may underestimate liquidity depth at CME and Hyperliquid.

Slippage is estimated using a synthetic order book sorted by execution price. Within each 5-minute timeframe, observed take orders are sorted in ascending price order (simulating sell orders) and executed sequentially until the target order size is reached. The arrival price is set to the lowest execution price within that timeframe (representing the best selling price when the order arrives). Slippage is calculated as the difference between the execution volume-weighted average price (VWAP) and the arrival price, expressed in basis points. This method is applied to incremental order sizes from $10,000 to $1,000,000.

Hyperliquid-CME Basis for Weekday Sessions: Tracks the signed price difference between the Hyperliquid midpoint and the latest CME price within a 5-minute window for all weekday sessions. This reflects any structural premium or discount of Hyperliquid relative to the CME reference price during active trading hours. The Hyperliquid midpoint is derived from the volume-weighted average price (VWAP) of trades executed within each 5-minute trading session, rather than from real-time order book quotes.

Hyperliquid financing rates are calculated hourly, expressed in hourly basis points.

Weekend time period data analysis

This analysis focuses on the three different weekend closing times of the CME:

• W1: February 28 to March 1, 2026

• W2: March 7 to March 8, 2026

• W3: March 14 to March 15, 2026

In W1 and W2, Hyperliquid perpetual contracts are restricted so that the marked price cannot exceed the "range limit boundary" (DB). When oracle prices are frozen (e.g., when the main reference market (CME) is closed and external price data sources stop updating), the protocol effectively confines the price to a narrow range.

For each weekend trading window, we will report key data metrics for Hyperliquid xyz:CL, including price, volume, and number of trades. To measure the Monday opening spread deviation, each weekend we will measure the price difference between Hyperliquid and CME at three reference points:

CME Reopening 3 Hours Before Reopening; CME Reopening 1 Hour Before Reopening; CME Opening (T=0)

All spreads are expressed in basis points; a positive value indicates that Hyperliquid is above the CME opening price, and a negative value indicates a discount.

Quantitative analysis

This section first analyzes and compares the liquidity of the Hyperliquid xyz:CL HIP-3 crude oil market and NYMEX CLJ6 during overlapping weekday periods.

Liquidity depth: Hyperliquid is less than 1% of CME's.

Undoubtedly, the liquidity situation on on-chain trading platforms is drastically different from that of CME. The average liquidity depth of CL on Hyperliquid is less than 1% of that of CLJ6, and the liquidity depth is consistent across all price ranges (109 times at ±5 bps). Within a range of ±2 bps from the midpoint price, CME's executable depth is $19 million, while Hyperliquid's is only $152,000, a difference of 125 times.

Given the novelty of the CL marketplace on Hyperliquid and its different target user group, this result is not surprising. The main value of on-chain trading platforms lies in providing permissionless trading channels for users traditionally excluded from institutions like CME.

However, as weekend trading volumes on DEXs like Hyperliquid have increased, perceptions of these platforms have begun to shift, with institutional investors showing growing interest in hedging positions outside of trading hours. Therefore, creating a market environment suitable for both traditional and retail investors has become increasingly important for Hyperliquid.

For retail traders with a trading volume of $10,000, this cost difference is negligible. However, for institutional investors with a trading volume exceeding $1 million, the on-chain transaction costs on CL (and most other markets) remain prohibitive.

In fact, the inherent differences among user groups are reflected in the median transaction size during these overlapping market periods.

The 166-fold difference in median trade size ($90,450 vs. $543) most clearly demonstrates the fundamental difference in the user groups served by these exchanges. CLJ6's median trade size is equivalent to a standard crude oil futures contract (with a notional value of approximately $94,000 at current prices), while Hyperliquid's median trade size of $543 reflects the leveraged directional betting by native cryptocurrency retail traders.

We anticipate that the median transaction size of the Hyperliquid commodity market will reach an inflection point as these markets become increasingly legitimate in the eyes of more traditional investors and they move funds on-chain.

To further differentiate between different transaction sizes, we conducted order simulations with order sizes ranging from $10,000 to $1 million.

For orders of $10,000, CLJ6 traders experienced no slippage, as expected, while the median execution slippage for Hyperliquid users was less than 1 basis point, at 0.77 basis points. The difference arises with orders of $100,000, where Hyperliquid users experienced slippage of 4.33 basis points, approaching the 5 basis point threshold, while CME CLJ6 traders experienced no slippage.

It is worth noting that this is higher than the median transaction size in the CLJ6 market ($90,450).

With a trading volume of $1 million, Hyperliquid's 15.4 basis points is approximately 20 times that of CME's 0.79 basis points, confirming that the trading venue currently lacks the capacity to handle institutional-level orders. Considering Hyperliquid's average trading size, the platform could easily provide the same high-quality service to users without incurring slippage.

CLJ6 orders only start to show significant slippage when the transaction size is around $500,000, affecting their execution.

When we extended the order size analysis to the weekend, slippage decreased across all order sizes, particularly for orders of $100,000 and $1 million, indicating market maturity. The decrease in slippage for simulated orders over the three weeks of analysis is as follows:

• $10,000: -16%

• $100,000: -75%

• $1 million: -86.9%

financing interest rate

CL funding rates fluctuate significantly during the CME closing session but less so during the settlement session. This helps us understand the market's internal pricing dynamics outside of trading hours. The weekend opening means the CL market can utilize internal price discovery mechanisms (supported by DB and other risk-reduction mechanisms). Therefore, funding rates are expected to be more volatile, as highlighted below.

During active trading hours, Hyperliquid's xyz:CL closely correlated with CME's CLJ6, but as oil prices rose, a structural discount emerged and widened, likely due to funding rate pressures from accumulated long positions. Over the weekend, with the CME closed, Hyperliquid's price discovery was further constrained by its price range mechanism (DB), which limits the volatility of the mark price in the absence of a real-time reference market.

Weekend session analysis: Hyperliquid has developed price discovery capabilities.

These three weekends showcased the rapid maturation of the Hyperliquid market:

W1: February 28 to March 1, 2026 (Iranian attacks)

The price on Hyperliquid rose from around $67.29 on the CME to around $70.80, representing about 45% (+1146 basis points) of Monday's final gap up to $75.

It's important to note that price discovery was limited this weekend due to trade.xyz's ±5% price range limit (DB) mechanism mentioned above. This explains the relatively flat curve in the chart and the gap up on Monday. Nevertheless, in the first second after the paired data was released, the gap between Hyperliquid xyz:CL ($73.89) and CME CLJ6 ($75) was within 1.5%.

This is not a "mistake" or a "failure," but rather a risk protection achieved through market design. Therefore, from a data perspective, the first weekend has the lowest correlation, but it highlights xyz:CL's reaction to the initial shock of the Iranian airstrikes, while also recognizing the importance of the DB as a weekend price discovery mechanism, especially for emerging markets.

W2: March 7 to March 8, 2026

The second weekend proved to be the real test, as xyz:CL touched the upper limit of its trading range in the final minutes of trading. CLJ6 opened at $98 (up 737 basis points from the closing price of $91.27), while xyz:CL reached a high of approximately $95.83, capturing only a 68% gain.

Over the second weekend, xyz:CL captured market movements better and came closer to the CME opening price than the previous weekend.

W3: March 14 to March 15, 2026

Data from the third weekend suggests that, in a relatively calm market environment, Hyperliquid can more reliably predict the final opening direction of the CME.

The convergence of xyz:CL and CLJ6 was at its best this weekend: up 226 basis points from the CME closing price, slightly higher than Monday's opening price by 62 basis points. CLJ6 closed at $99.31 on Friday, opening at $100.93 (up 163 basis points), while xyz:CL opened at $101.56.

In summary, these three snapshots reveal a structural shift in the xyz:CL market on the Hyperliquid platform, transitioning from an emerging market constrained by DB price discovery (weekend 1 and weekend 2) to one with increasingly free price discovery, exhibiting overshoot and pullbacks (weekend 3).

Analyzing price deviation errors at different times (3 hours, 1 hour, and 0 hours) before the CME market opens on different weekends reveals that the data from Week 3 is the most reliable, as the xyz:CL market was influenced by DB on the previous two weekends. In Week 3, the errors for xyz:CL in the 3 hours and 1 hour before the CME market opens are approximately +70 and -139 basis points, respectively, indicating that its price discovery ability is better than the previously analyzed weekends.

Other indicators

We also provide other metrics for our weekend summary analysis, including trading volume, total number of trades, and average trade size. These metrics vary between weekends and show a consistent upward trend over several consecutive weekends.

The total transaction volume of the xyz:CL market grew from $31 million to over $1 billion in three weeks, reflecting the increase in the number of users and the eventual maturity of the market.

In addition, the total number of transactions increased from 26,000 at the first weekend to more than 700,000 at the third weekend.

It's worth noting that the average trade size over the weekend actually increased from the median we mentioned earlier to $534. The same upward trend was observed across all three weekends, which likely indicates more institutional funds flowing into the market.

The average trade size was $1,199 over the first weekend, growing to over $1,500 by the third weekend.

This may indicate that the user base using the platform differs over the weekend, with fewer retail users and more traders needing to gain exposure to crude oil trading before Monday. Therefore, weekend trading is more driven by hedging needs than speculation .

A user deposited 1 million USDC in Hyperliquid and used leverage to buy ETH, BTC, PUMP, and other assets.

Popular Articles