Executive Summary

This report systematically analyzes coke market profit by combining industrial logic, historical pattern validation, and quantitative data. The core conclusion is that coke market profit fundamentally follows industry capacity cycles, adhering to the classic loop: "capacity expansion leads to inventory accumulation, which drives price declines and profit compression, triggering capacity reduction, which leads to inventory drawdowns, price recovery, and profit restoration." In most cycles, coke market profit moves in tandem with rebar market profit, reflecting the consistency of profit transmission along the ferrous supply chain. However, the profit divergence from late 2020 to 2021 was a typical supply shock anomaly: dual-carbon policies caused a rigid contraction in coal supply, sharply raising raw material costs and squeezing coke profits, creating a deviation during the broader ferrous bull market. Only after policy corrections and supply normalization did the profit logic return to the industry cycle's main track.

A very strong positive correlation between coking coal and coke market prices is the core basis for profit games. Adjustments to coking coal delivery standards directly revalue the market profit center via warrant costs. A transmission bias exists between fundamentals and the market, with market profit often showing a leading indicator function.

Macro Dynamics: Profit Compression During Industry Downturns

Similar to steel capacity, coke capacity is also in an overcapacity cycle. Different stages of the capacity cycle bring predictable changes in inventory, profit, and prices. Capacity expansion or contraction influences supply, which, combined with demand, determines inventory levels, price trends, and profit margins. These three factors are interconnected, forming the cyclical pattern: "capacity expansion, inventory accumulation, price decline, profit decline, capacity contraction, inventory reduction, price rise, and profit improvement."

Historical data shows that coke market profit changes generally align with rebar, but a divergence occurred from late 2020 to 2021. During that period, the ferrous industry was in a bull market, yet coke profit was compressed. This was primarily because, in the first half of 2021, carbon peak and neutrality policies created a supply-demand gap in coal supply, leading to a coal price surge. In the second half of 2021, policy overcorrection and supply normalization caused prices to fall as the market traded on supply recovery.

Coke capacity has moved from "expansion, overcapacity, capacity reduction, optimization" to a "stock clearance plus integration" phase. Future total capacity is unlikely to increase. As a midstream processing product, coke is squeezed by upstream and downstream prices, lacking independent pricing power. Its profitability is the most unstable, with industry profits depending more on the game between coking coal costs and steel mill margins.

Micro Level: The Game with Upstream Raw Materials

1. Single-Leg Correlation Between Coking Coal and Coke

The Pearson correlation coefficient between the closing prices of coking coal and coke main contracts is 0.9581, with a Spearman rank correlation of 0.9648, indicating a very strong positive linear relationship. The 60-day rolling correlation average is 0.801, with a historical low of -0.3662 (brief negative correlation), and the latest reading is 0.7729. Under extreme market conditions (e.g., supply-side reform, pandemic shocks), the linkage may temporarily decouple.

2. Impact of Delivery Standard Modifications Over 100 Yuan/Ton on Market Profit

On April 13, 2022, the DCE made its second revision to coking coal delivery standards (effective from JM2304 contract, with the main contract JM2305 beginning in December 2022). This revision raised the upper limit of ash content, implemented a fixed-price reward and penalty system for indicators, relaxed sulfur content requirements, added a penalty system for volatile matter, changed the heat strength index pricing from penalty to premium, and adjusted the moisture deduction method. The benchmark delivery area was moved to the main Shanxi production area, with port areas becoming non-benchmark areas, subject to a 170 yuan/ton premium. Based on the revised standards, Shanxi coal warrant costs dropped by about 340 yuan/ton, and Mongolian coal warrant costs fell by about 290 yuan/ton. The market profit increased by approximately 385 yuan/ton. From January to November 2022, the main contract market profit ranged from -662 to 20, with an average of -340. From December 2022 to the present, the range is -407 to 157, with an average of -76.

On December 19, 2025, the DCE made its third revision to coking coal delivery standards (effective from JM2701 contract). This revision reduced the penalties for sulfur content, adjusted the heat strength index from a premium for high strength to a penalty for low strength, and revised the delivery warehouse premiums in Tangshan and Tianjin from 170 yuan/ton to 140 yuan/ton. Based on the revised standards, Shanxi coal warrant costs increase by about 80 yuan/ton, and Mongolian coal warrant costs increase by about 140 yuan/ton. The resulting market profit change still requires close monitoring of the warrant cost increase for the optimal deliverable product at that time.

3. Guidance of Physical Fundamentals on Market Profit Fluctuations

The correlation between physical profit and market profit is weak. Using 466 data points from 2017 to the present, the Pearson correlation coefficient is r = 0.4076 (p < 0.001), classified as a weak correlation. A positive relationship exists between coking plant spot profit and coke market profit: higher spot profit tends to indicate higher market profit. The coefficient of determination R² = 0.1661, meaning that about 16.6% of the variation in market profit can be explained by the linear relationship with spot profit. Furthermore, the correlation fluctuates significantly over different periods, indicating instability. Some periods show strong positive correlation, while others show weakened or even negative correlation, possibly due to changes in coke supply and demand, policy adjustments, or futures market sentiment.

Market profit leads spot profit, with the strongest lag being a 16-week lead. This suggests that the futures market has a price discovery function for profit changes, with market profit movements providing early signals for future spot profit trends.

Coke supply and demand fundamentals show a moderate positive correlation with market profit, but this relationship is unstable. In years with prominent supply-demand contradictions (e.g., 2020, 2024), the correlation is strong. In years with complex market environments (e.g., 2022-2023), it nearly disappears. Profit changes offer a weak leading indicator for production changes and should not be used solely as a trading signal.

The correlation between hot metal production and coke production profit is overall weak but positive. A one-way leading effect from profit to hot metal production exists (profit leads by about 4-8 weeks for the strongest predictive power). The intra-year relationship is highly unstable, being significant only in 2018, 2020, and 2024, so it should not be used as a standalone trading signal.

Coke inventory shows a moderate positive correlation with market profit, but this involves a Simpson's paradox: 5 out of 9 years show a negative correlation (2018: r=-0.62, 2019: r=-0.65, 2020: r=-0.79, 2022: r=-0.50, 2024: r=-0.45), and 4 years show a positive correlation (2021: r=+0.59, 2023: r=+0.66, 2025: r=+0.65, 2026: not significant). The alternating pattern of yearly correlation coefficients clearly reflects the cyclical nature of supply and demand in the coke industry. The source of the overall positive correlation is the long-term trend: average inventory fell from 12.73 million tons in 2018 to 8.6 million tons in 2024, while average profit dropped from 221 yuan to -303 yuan, driving the overall positive correlation. Within each year, inventory and profit tend to move inversely (short-term logic of "inventory accumulation puts pressure on prices"). Lag effects show that r is highly stable (0.40~0.42) within a ±8-week range, with the highest r (0.4225) when profit leads by 8 weeks, but the difference is minimal, suggesting they are more synchronous indicators than one-way leading. The weekly change rate shows a weak negative correlation (r=-0.109, p=0.034), meaning that in a week when inventory increases, profit tends to decrease slightly, consistent with basic supply-demand logic.

Micro Level: The Game with Downstream Steel Mill Margins

The Pearson correlation coefficient between market profit for rebar and coke is only 0.2221, classified as a weak positive correlation. However, the Spearman rank correlation is 0.2916, higher than the Pearson, indicating a non-linear monotonic relationship. Correlation analysis of the first difference (daily change) yields a coefficient of -0.0953 (p=1.01e-07), a very weak negative correlation. This suggests that on a daily basis, the two tend to move in opposite directions slightly more often than in the same direction—intraday, an increase in rebar profit is associated with a slight decrease in coke profit, reflecting short-term profit distribution game dynamics.

In 2020 and 2021, the correlation dropped to -0.73, a strong negative correlation. This period coincided with the deepening of supply-side reform and the implementation of dual-carbon policies, where steel production restrictions pushed up rebar profit while suppressing coke profit (or soaring coke prices eroded steel mill margins). In contrast, the correlation in 2023 was as high as +0.86, a strong positive correlation, indicating that in some years, both are driven by common factors like changes in end-user demand or policies. The correlation fluctuates wildly between years (from -0.73 to +0.86), suggesting the relationship is not simply linear but dynamically switches based on the supply-demand structure of the industry chain. This aligns with the profit distribution game between steel mills and coking plants in the ferrous supply chain.

The combination of positive rebar profit and negative coke profit accounts for 47.6% of observations, making it the most common scenario. This reflects the core profit distribution nature of the ferrous chain—when steel mills are profitable, it often implies that raw material (coke) is being squeezed, and vice versa. The scenario where both are positive accounts for only 29.5%, indicating that win-win situations for both upstream and downstream are relatively rare.

Summary of Market Profit Trading Rules for Coke

Reviewing the historical structure of coke market profit, we find that profit tends to expand during market downturns and contract during market upturns. The core reason for this characteristic is that price volatility differences in upstream raw materials are primarily driven by supply-side factors. Coking coal supply has stronger narrative-driven volatility, while coke supply is mainly adjusted by profit. Consequently, coking coal experiences greater daily volatility than coke. When a bullish market emerges, coking coal rises more, leading to profit contraction. When a bearish market emerges, coking coal falls more, leading to profit expansion.

Supply-demand fundamentals, as well as upstream and downstream costs and profits, show some correlation or lead-lag relationship with coke market profit. However, no single factor is sufficient to predict market profit direction. Therefore, when positioning for a coke market profit spread trade, a comprehensive assessment of the upstream and downstream industry fundamentals is required. The core premise for forming a final trading decision should be the judgment of the ferrous supply chain's single-leg price trend.