Market consensus says COSCO Shipping Holdings' (601919) profit is propped up by freight rates, so for the Q3 report on October 30, was the single-quarter gross margin higher than in Q2, and did it move in the same direction as the Shanghai Export Containerized Freight Index published every Friday?
- Broken into single quarters: Q1 17.24%, Q2 19.29%, Q3 30.96%.
- Voyage costs accounted for 22.11%, up 10.46%
- Asia-Europe revenue per container was 11,849 yuan, up 9.68%.
- Zhonggu's Q3 single-quarter revenue grew only 1.4%, and operating cost actually fell 2.4%
- Gross profit of 234.33 yi yuan divided by 756.92 yi yuan is exactly 30.96%.
Market consensus says COSCO Shipping Holdings' (601919) profit is propped up by freight rates, so for the Q3 report on October 30, was the single-quarter gross margin higher than in Q2, and did it move in the same direction as the Shanghai Export Containerized Freight Index published every Friday?
市场共识说中远海控(601919)的利润是运价撑起来的，那10月30日三季报的单季毛利率，有没有比二季度更高、并且和每周五公布的上海出口集装箱运价指数同向？
The desk answersverifiedWalnut, from public filings
Both premises hold. In the third-quarter report dated October 30, the single-quarter gross margin for Q3 was 30.96%, 11.67 percentage points higher than Q2's 19.29%; and the Shanghai Containerized Freight Index (SCFI) in Q3 was indeed overall higher than in Q2.
But the same direction does not mean synchronization: in Q2 SCFI rose seven and a half tenths, yet that quarter's gross margin rose only 2.05 points; in Q3 SCFI itself rose only one tenth, yet the gross margin rose 11.67 points. Gross margin follows the previous quarter's freight rates.
Cumulative figures must be broken apart
The cumulative gross margin for the first three quarters was 23.43%, versus 18.34% in the half-year report — seemingly only 5.09 points higher, because the cumulative figure mixes Q2 and Q3 together. Broken into single quarters: Q1 17.24%, Q2 19.29%, Q3 30.96%.
The reconciliation can verify itself: operating revenue of 1,876.14 yi yuan in the third-quarter report minus 1,119.22 yi yuan in the half-year report gives 756.92 yi yuan; the third-quarter report's "operating revenue for this reporting period" is written as 75,691,788,243.39 yuan — identical digit by digit.
Operating cost of 1,436.54 yi yuan minus 913.96 yi yuan gives 522.58 yi yuan; gross profit of 234.33 yi yuan divided by 756.92 yi yuan is exactly 30.96%.
| Quarterly gross margin (%) | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| 2024 | 19.39 | 27.16 | 40.81 | 25.94 |
| 2025 | 27.05 | 14.45 | 23.53 | 13.83 |
| 2026 | 17.24 | 19.29 | 30.96 | — |
Looking only at quarter-on-quarter, this finding does not hold
Four shipping companies all rose from Q2 to Q3 2026 (the first three operate container ships; COSCO Shipping Specialized Carriers is specialized vessels, used as a control): Jinjiang Shipping +2.31 points, Zhonggu Logistics +2.64 points, COSCO Shipping Specialized Carriers +4.15 points, COSCO Shipping Holdings +11.67 points. Peak season is in Q3 anyway; liner companies' gross margins rise at this time every year.
| 2026 single-quarter gross margin (%) | Q2 | Q3 | quarter-on-quarter pp | year-on-year pp |
|---|---|---|---|---|
| COSCO Shipping Holdings | 19.29 | 30.96 | 11.67 | 7.43 |
| Jinjiang Shipping | 29.46 | 31.77 | 2.31 | (2.15) |
| Zhonggu Logistics | 28.66 | 31.30 | 2.64 | 11.17 |
| COSCO Shipping Specialized Carriers | 26.24 | 30.39 | 4.15 | 7.07 |
What has discriminating power is the year-on-year comparison. Three of the four rose: COSCO Shipping Holdings +7.43 points, Zhonggu Logistics +11.17 points, COSCO Shipping Specialized Carriers +7.07 points; only Jinjiang Shipping fell 2.15 points.
But the mechanisms are completely different. Zhonggu's Q3 single-quarter revenue grew only 1.4%, and operating cost actually fell 2.4%; COSCO Shipping Specialized Carriers' revenue grew 3.7% and cost fell 2.2% — these two companies' gross margins were squeezed out from the cost side. COSCO Shipping Holdings is the reverse: revenue was 155.67 yi yuan higher quarter-on-quarter, while cost was only 37.29 yi yuan higher; 76% of the incremental revenue fell directly to gross profit. Gross margin curves that look alike do not mean the same chain.
There is only one path for freight rates
SCFI is the spot booking price for exports from the Port of Shanghai, published every Friday. It first affects COSCO Shipping Holdings' spot slot quotes, then long-term contracts (signed annually or repriced quarterly), and finally becomes revenue per container, recorded in operating revenue — not in investment income; investment income for the first three quarters was 45.46 yi yuan, basically flat versus the same period last year, and cannot support this change.
The cost side is quasi-fixed. In the half-year report, equipment and cargo transport costs accounted for 45.32% of operating cost, up only 1.40% year-on-year; voyage costs accounted for 22.11%, up 10.46%; vessel costs accounted for 20.57%, up 9.12%. The first item moves with cargo volume; the latter two are basically fixed. Revenue rises while cost does not follow, so the gross margin jumps.
There is also an exchange-rate layer in between: the half-year report used 6.9043 for USD translation versus 7.1793 in the same period last year — the renminbi appreciated about 4%, which eats a chunk of dollar revenue.
Verify one layer further down. Q3 single-quarter gross profit was 234.33 yi yuan versus 137.62 yi yuan in the same period last year, an increase of 96.71 yi yuan; over the same period, single-quarter net profit attributable to parent increased by 76.97 yi yuan. The gross profit increment covers the net profit increment, indicating that the profit uplift basically occurred at the gross profit layer and was not swallowed by the items below.
There is also counter-evidence: for the first three quarters, financial expenses went from -24.93 yi yuan in the same period last year to +0.43 yi yuan, and interest income fell from 42.71 yi yuan to 29.05 yi yuan; this item sits below gross profit and is a drag, not a contribution.
During this reporting period, the Group's container shipping business saw both cargo volume and revenue per container rise year-on-year, and operating efficiency rose year-on-year.
Revenue per container writes the trace of freight rates in plain sight
Q3 cargo volume was 721.15 wan TEU, up only 4.46% year-on-year; route revenue was 678.03 yi yuan, up 30.94% year-on-year. Divided out, revenue per container is 9,402 yuan versus 7,501 yuan in the same period last year, up 25.3%. Volume up four points, price up twenty-five points — the revenue growth is all in price.
This calculation can be checked first: using the same method, international route revenue per container for the first half was 1,181 USD, while the half-year report disclosed 1,180.68 USD — it matches.
By route it is even clearer who is rising. Trans-Pacific revenue per container was 16,874 yuan, up 42.68% year-on-year, with cargo volume up only 12.98%; Asia-Europe revenue per container was 11,849 yuan, up 9.68%. Q3 SCFI itself was strong on US routes and weak on Europe routes — when US East Coast rates broke ten thousand dollars at the end of August, Europe routes were still falling. The aggregate and the by-route figures point the same way.
The first half was the reverse: revenue per container was 6,827 yuan versus 7,274 yuan in the same period last year, still down 6.1%. In the first half SCFI had already risen from 1,854.96 points on April 3 to around 3,240 points at the end of June, yet revenue per container did not keep up; it only flipped in Q3. Last quarter's price, this quarter's accounts.
The strongest rebuttal
In the third quarter of 2024, the single-quarter gross margin was 40.81%, nearly 10 points higher than this year's 30.96%; no matter how much freight rates rose, they did not return to the level of two years ago, which shows that freight rates cannot explain the gross margin.
That statement is half right. That 10-point gap may be due to the cost base (voyage costs +10.46% and vessel costs +9.12% in the half-year report), or it may be due to route mix; only the annual report's segment breakdown can settle it. But it cannot lead to the conclusion that "freight rates are useless": in the same table, the third-quarter 2026 gross margin rose 7.43 points year-on-year, while costs were rising over the same period, and revenue rose much faster than costs. If freight rates had not been passed through, this difference would not have appeared.
Under what circumstances I would be wrong
If the fourth-quarter SCFI weekly average falls below 3,000 points while COSCO Shipping Holdings' fourth-quarter single-quarter gross margin still holds above 28%, then I am wrong; that would show the gross margin is propped up by the cost side, and freight rates merely rode the tailwind.
If, in the annual report's segment breakdown, the container shipping gross margin does not rise year-on-year and only the terminal business gross margin rises, then I am also wrong; the terminal business had a first-half gross margin of 27.63% and a revenue share of 5.6%, so in terms of scale it cannot support 11.67 points. If it were what supported it, this chain would be broken.
Sources: COSCO Shipping Holdings Co., Ltd. "2026 Third Quarter Report," October 30, 2026, CNINFO announcement ID 1225642417; "2026 Half-Year Report," August 29, 2026, 1225531118; Shanghai Shipping Exchange "Shanghai Containerized Freight Index" single-period query page, readings for October 30, 2026 and October 23; East Money SCFI period-by-period news flashes, July 3, July 24, August 21, September 11, September 18, and October 16, 2026; the 2026 third quarter reports of Jinjiang Shipping 601083, Zhonggu Logistics 603565, and COSCO Shipping Specialized Carriers 600428, with cumulative figures taken from the tushare financial report interface and then restored to single-quarter figures.
两个前提都成立。10月30日的三季报里，三季度单季毛利率30.96%，比二季度的19.29%高11.67个百分点；上海出口集装箱运价指数（SCFI）三季度也确实整体高于二季度。
但同向不等于同步：二季度SCFI涨了七成半，当季毛利率只抬2.05个点；三季度SCFI自己只涨一成，毛利率却抬了11.67个点。毛利跟的是上一季的运价。
累计数必须拆开
前三季累计毛利率23.43%，半年报18.34%，看着只高5.09个点，因为累计把二季度和三季度混在一起。拆成单季：Q1 17.24%、Q2 19.29%、Q3 30.96%。
还原过程可以自证：三季报营业收入1,876.14亿元减半年报的1,119.22亿元，得756.92亿元；三季报"本报告期营业收入"写的是75,691,788,243.39元，逐位相同。
营业成本1,436.54亿元减913.96亿元，得522.58亿元；毛利234.33亿元，除以756.92亿元，就是30.96%。
| 单季毛利率(%) | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| 2024 | 19.39 | 27.16 | 40.81 | 25.94 |
| 2025 | 27.05 | 14.45 | 23.53 | 13.83 |
| 2026 | 17.24 | 19.29 | 30.96 | — |
只看环比，这个发现不成立
四家航运公司，2026年二季度到三季度全部抬升（前三家做集装箱，中远海特是特种船，拿来当对照）：锦江航运+2.31个点，中谷物流+2.64个点，中远海特+4.15个点，中远海控+11.67个点。旺季本来就在三季度，班轮公司每年这个时候毛利率都往上走。
| 2026年单季毛利率(%) | Q2 | Q3 | 环比pp | 同比pp |
|---|---|---|---|---|
| 中远海控 | 19.29 | 30.96 | 11.67 | 7.43 |
| 锦江航运 | 29.46 | 31.77 | 2.31 | (2.15) |
| 中谷物流 | 28.66 | 31.30 | 2.64 | 11.17 |
| 中远海特 | 26.24 | 30.39 | 4.15 | 7.07 |
有分辨力的是同比。四家里三家抬升：中远海控+7.43个点、中谷物流+11.17个点、中远海特+7.07个点，只有锦江航运掉了2.15个点。
但机制完全不同。中谷三季度单季营收只增1.4%、营业成本还降了2.4%；中远海特营收增3.7%、成本降2.2%——这两家的毛利率是从成本端挤出来的。中远海控反过来：营收环比多出155.67亿元，成本只多37.29亿元，增量收入的76%直接落到毛利额上。毛利率曲线长得像，不等于同一条链。
运价走的路只有一条
SCFI是上海港出口的即期订舱价，每周五出。它先影响中远海控的即期舱位报价，再影响长约（一年一签或季度调价），最后变成单箱收入，记在营业收入里——不在投资收益，前三季投资收益45.46亿元、跟去年同期基本持平，撑不起这个变化。
成本那一头是准固定的。半年报里，设备及货物运输成本占营业成本45.32%，同比只增1.40%；航程成本占22.11%，增10.46%；船舶成本占20.57%，增9.12%。前一项随货量走，后两项基本固定。收入涨、成本不跟着涨，毛利率就跳起来。
中间还夹一层汇率：半年报的美元折算用6.9043，去年同期7.1793，人民币升值约4%，会吃掉一块美元收入。
再往下验一层。三季度单季毛利额234.33亿元，去年同期137.62亿元，多出96.71亿元；同期单季归母净利润多出76.97亿元。毛利增量盖住了净利增量，说明利润抬升基本发生在毛利这一层，没被下面几项吞掉。
反向的证据也有：前三季财务费用从去年同期的-24.93亿元变成+0.43亿元，利息收入由42.71亿元降到29.05亿元，这一项在毛利下面，是拖累不是贡献。
本报告期内，本集团集装箱航运业务货运量与单箱收入均同比上升，经营效益同比上升。
单箱收入把运价的痕迹写在明面上
三季度货运量721.15万标准箱，同比只增4.46%；航线收入678.03亿元，同比增30.94%。一除，单箱收入9,402元，去年同期7,501元，涨25.3%。量增四个点、价增二十五个点，收入增长全在价上。
这个算法可以先验收：用同样方法算上半年国际航线单箱收入是1,181美元，半年报披露的是1,180.68美元，对得上。
分航线更能说明是谁在涨。跨太平洋单箱16,874元，同比涨42.68%，货量只增12.98%；亚欧单箱11,849元，涨9.68%。三季度SCFI本身就是美线强、欧线弱——8月底美东运价破万美元的时候，欧线还在跌。汇总和分航线，方向一致。
上半年反过来：单箱收入6,827元，去年同期7,274元，还跌6.1%。上半年SCFI已经从4月3日的1,854.96点涨到6月末的3,240点附近，单箱收入没跟上，三季度才翻过来。上一季的价，这一季的账。
最强的一句反驳
2024年三季度单季毛利率40.81%，比今年的30.96%高近10个点；运价再怎么涨也没回到两年前，可见运价解释不了毛利率。
这句话对一半。那10个点的缺口，可能是成本基数（半年报航程成本+10.46%、船舶成本+9.12%），也可能是航线结构，要等年报分业务才能定。但它推不出"运价没用"：同一张表里，2026年三季度毛利率同比抬了7.43个点，同期成本在涨，收入涨得比成本快得多。运价没传导，这个差额不会出现。
什么情况下我错了
四季度SCFI周均跌破3,000点、而中远海控四季度单季毛利率还守在28%以上，我就错了，那说明毛利是成本端撑的，运价只是搭了顺风车。
年报分业务里集装箱航运毛利率同比没抬升、只有码头业务抬升，我也错了；码头业务上半年毛利率27.63%、收入占比5.6%，体量上撑不起11.67个点，如果是它撑的，这条链就断了。
- 01中远海运控股股份有限公司《2026年第三季度报告》，2026年10月30日，巨潮公告ID 1225642417
- 02《2026年半年度报告》，2026年8月29日，1225531118
- 03上海航运交易所"上海出口集装箱运价指数"单期查询页，2026年10月30日、10月23日读数
- 04东方财富SCFI各期快讯，2026年7月3日、7月24日、8月21日、9月11日、9月18日、10月16日
- 05锦江航运601083、中谷物流603565、中远海特600428的2026年第三季度报告，累计数取自tushare财报接口后按单季还原
## 拆开这11.67个点
以2025年三季度为基数，分两步替换：先把单价换成今年的、成本留在去年，再换单位成本。
| 步骤 | 毛利率(%) | 变化(pp) |
|---|---|---|
| 2025年三季度实际 | 23.53 | — |
| 单价换成2026年 | 38.26 | 14.73 |
| 单位成本换成2026年 | 30.96 | (7.30) |
一加一减，结果就是同比的7.43个点。反事实更直接：单价一动不动，只让今年的货量和成本上场，收入611.11亿元、成本522.58亿元，毛利率14.49%。把运价这条腿抽掉，今年三季度的毛利率比去年还低9个点。
共识说这是运价撑起来的，真实情况更极端：不是运价多给了一块，是别的什么都没给。
## 价涨了，成本也涨了
把全公司收入和营业成本摊到集装箱货运量上（收入含码头和供应链业务，这个"单箱"只看同比方向，不能拿去和纯班轮公司的单箱水平比）：单箱收入涨23.86%，单箱成本涨11.82%，单箱毛利涨62.96%。
```chart
type: bar
title: 单箱收入涨24%，单箱成本涨12%，单箱毛利涨63%
unit: %
labels: 单箱收入, 单箱成本, 单箱毛利
values: 23.86, 11.82, 62.96
```
单位成本同比涨近12%，先排掉两种流行解释：不是降本，也不是规模效应。货量只增4.46%，成本增16.82%，摊到每一箱上反而更贵。
成本为什么涨，三季报不给分项。半年报的拆分是：设备及货物运输成本占营业成本45.32%、航程成本22.11%、船舶成本20.57%，第一项随货量走，后两项更接近固定。运价涨的时候成本不跟，运价跌的时候成本也不跟——这个结构是双向的。
"单箱收入涨了，但不全是涨价，还有把货挪到高价航线"——这个质疑得用数回答。拿集团口径的分航线美元收入和分航线货量做结构分解：各航线自身涨价贡献27.17个百分点，航线结构（跨太平洋占比从17.82%升到19.27%）贡献2.92个点，交叉项1.18个点，合计31.27%。结构确实在往美线挪，但它只值3个点，占不到一成。
汇率也吃掉一截：美元口径单箱收入涨31.27%，折成人民币只有25.34%，中间5.93个点是升值吃掉的。报表上看到的涨价，比实际发生的少。
## 同行：真提价的只有两家
把二季度到三季度的收入和成本环比增量摆在一起，四家人的性质立刻分开。
| 2026Q3较Q2 | 收入环比(亿) | 成本环比(亿) | 增量毛利率(%) |
|---|---|---|---|
| 中远海控 | 155.67 | 37.29 | 76.0 |
| 锦江航运 | 0.98 | 0.21 | 78.9 |
| 中谷物流 | 0.39 | (0.49) | 225.4 |
| 中远海特 | 2.90 | (1.26) | 143.4 |
中远海控和锦江航运的增量收入里，76%、79%落到毛利上，同时成本还在环比上涨——这是真提价。中谷物流和中远海特的增量毛利率是225%和143%，超过100%只有一个原因：成本环比是降的，改善是从成本端挤出来的。
所以三季度这轮毛利率抬升在行业内并不同质。跟SCFI挂钩的两家，报表上写的是价；不挂钩的两家（内贸、特种船），报表上写的是成本。海控这11.67个点是四家里最强的那一版，也意味着运价掉头时它最没有缓冲。
## 这是结算单，不是预测
单箱收入是即期价和长约价的加权。长约按季或按年调，所以二季度涨上去的运价，要到三季度才完整进报表；三季度自己涨的那部分，要等四季度才结算。
二季度运价涨了七成半，毛利率只抬2.05个点；三季度运价只涨一成，毛利率抬了11.67个点。不是运价没用，是账期错开了一季。
按这个错位，三季报没有提供任何关于四季度的新信息。四季度对应的是三季度的运价，而三季度的运价本身涨幅有限、季末还在往下走。谁把三季度的毛利率当成四季度的起点，就是把滞后当成了同步。
## 长约是唯一的出口，但它查不到
即期运价是一次性的。能把它变成明年地基的只有长约：年底到明年初是谈判窗口，如果欧美线按今年的即期水平签，2027年的收入底盘就被抬高了。
但长协的占比和价格，三季报没有，半年报也没有。这是查不到的数，不能猜。11月17日15:00到16:00有业绩说明会，提问预征集11月16日16:00截止，这是年内唯一能问到的窗口。要问的就一句：明年欧美线长约的签约比例和价格，比今年高还是低。
## 高毛利的时候下单，2028年的供给是自己造的
8月28日订了18艘船：12艘22,000TEU、6艘3,200TEU，总价202.74亿元，交付期2028到2030年。28.32万TEU相当于现在366万TEU运力的7.7%。资金安排是70%外部融资，公告自己算了，总负债要增加141.92亿元。
在建工程从年初166.65亿元涨到267.22亿元，九个月增100.57亿元，增幅60.35%。钱已经在往船台上走。
> 通过运力梯次置换，将优势有效传导至新兴市场、第三国市场和区域市场，实现本集团运力全球均衡化布局。
公告给的理由是老船替换和布局均衡，不是单纯扩张，这个说法站得住。但净运力是增加的，7.7%不因为理由好听而变小。这批船2028到2030年交付，那时候的运价由那时候的供需定，而这批船本身就是那时候的供给。今天的高毛利，有一块正在变成三年后的低毛利。
## 最强的一条反驳
有人会说：2024年三季度毛利率40.81%，今年只有30.96%，运价再怎么涨也没回到两年前，可见运价解释不了毛利率。
这句话对，而且比它自己想的更有力。2024年三季度收入735.36亿元，今年756.92亿元，两年只多2.93%；营业成本从435.25亿元涨到522.58亿元，多20.07%。毛利率掉的那10个点，几乎全在成本这一边。
所以准确的说法是：成本是慢变量，决定毛利率的中枢；运价是快变量，决定季度的波动。跨季看运价是主变量，跨两年看成本是主变量。这个成本中枢两年前就抬上去了，这份季报没有改变它一个字。这11.67个点因此不能外推：它是在已经抬高的成本基数上，被快变量弹了一下。
## 什么情况下我错了
- 四季度单季毛利率还守在28%以上，而同期SCFI周均往下走。那说明成本不是慢变量，这套中枢与波动的分法就错了。
- 2026年年报里"集装箱航运业务"毛利率同比没抬升，只有码头业务抬升。码头上半年毛利率27.63%、收入占比5.65%，撑不起这个量级，传导链就断了。
- 2027年长约价同比明显抬升并被公司披露。那说明今年的高运价已经写进明年的地基，"一次性"这个判断错了。
- 2027年单箱成本同比转降。新船交付加上老船退出，成本中枢下移，那利润就不是只有运价一条腿。
来源：中远海运控股《2026年第三季度报告》，2026年10月30日，巨潮公告ID 1225642417（单季数据按合并利润表还原，货运量及分航线收入取自第6至8页）；《2026年半年度报告》，2026年8月29日，1225531118；《关于全资子公司订造十二艘22,000TEU型及六艘3,200TEU型集装箱船的自愿性公告》，2026年8月29日，1225531141；《关于召开2026年第三季度业绩说明会的公告》，2026年10月30日，1225642474；2024至2026年三季报累计数取自tushare income接口（20240930、20250930、20260930），单季按本期累计减上期累计还原；锦江航运601083、中谷物流603565、中远海特600428的2026年单季数据同样按累计数还原后计算。