Where does the copper-clad laminate price increase hit first on the financial statements: for Shennan Circuits (002916), in the third-quarter report disclosed on October 31, how much more raw material inventory was there than at the end of the second quarter, and was the single-quarter gross margin lower than in the second quarter?
- The operating cost ratio fell from 67.59% to 65.15%
- Single-quarter gross margin did not fall, rising from 32.41% to 34.85%
- Inventory at the three companies rose 18% to 25% quarter-on-quarter in the same direction
Where does the copper-clad laminate price increase hit first on the financial statements: for Shennan Circuits (002916), in the third-quarter report disclosed on October 31, how much more raw material inventory was there than at the end of the second quarter, and was the single-quarter gross margin lower than in the second quarter?
覆铜板涨价先压到哪张报表:深南电路(002916)10月31日披露的三季报,原材料存货比二季度末多了多少,单季毛利率有没有比二季度低?
The desk answersverifiedWalnut, from public filings
Two premises in this are invalid: Shennan Circuits' third-quarter report was disclosed on October 23, not October 31; and the third-quarter report has no "raw materials" line item, so the figure for "how much more raw materials inventory than at the end of the second quarter" cannot be found in any public document.
Three things that can be checked: total inventory is 18.82 yi more than at the end of the second quarter; single-quarter gross margin did not fall, rising from 32.41% to 34.85%; the copper-clad laminate price increase has landed on the balance sheet and has not yet landed in operating cost on the income statement.
What was posted on October 31 was a research record
The third-quarter report was disclosed on October 23, announcement no. 2026-065. What was posted on October 31 was an investor relations activity record, no. 2026-24, covering a research visit by fifteen institutions from October 27 to 30.
In the record there happens to be a question about raw materials, and the company answered: main raw materials include copper-clad laminate, prepreg, copper foil, gold salt, ink, etc.; in the third quarter, "affected by changes in commodity prices, prices of some raw materials such as gold salt rose year-on-year and also rose somewhat quarter-on-quarter."
Asked about consolidated gross margin quarter-on-quarter, the company said it "improved somewhat quarter-on-quarter," citing growth in data center and wired communications revenue, release of packaging substrate orders, and ramp-up at the Guangzhou Nantong phase-four Thailand plants — it did not mention cost.
The company itself did not attribute the gross margin rise to lower cost.
The raw materials line does not exist in the quarterly report
The third-quarter report has only one line, "inventory": 96.997 yi at end-September, 51.397 yi at the start of the year, up 88.72%; the company gave the reason for the change as "increased business scale and the impact of stocking materials."
The breakdown exists only in the interim report: on June 30, raw materials book balance 26.03 yi (book value 25.50 yi), work in progress 19.90 yi, finished goods 14.26 yi, goods shipped 20.95 yi, totaling 81.14 yi, book value 78.17 yi.
So the only thing that can be calculated is the total: from 78.17 yi to 97.00 yi, an increase of 18.82 yi, up 24.1%. The raw materials balance on September 30 will not appear in any public document — the interim report gives June 30, the annual report gives December 31.
Asking "how much more raw materials" is asking for a number that does not exist.
Single-quarter gross margin rose 2.44 points, but the improvement is narrowing
Breaking out single quarters by subtracting cumulative figures: first quarter revenue 65.96 yi, cost 46.72 yi, gross margin 29.17%; second quarter 87.01 yi, 58.80 yi, 32.41%; third quarter 104.78 yi, 68.26 yi, 34.85%, single-quarter gross profit 36.52 yi.
The operating cost ratio fell from 67.59% to 65.15%; the cost side simply was not pushed up.
Last year followed the same path: second quarter 2025 27.59%, third quarter 31.39%, up 3.80 points quarter-on-quarter. This year the same third quarter rose only 2.44 points, 1.36 points less.
This year's third-quarter revenue rose 66.29% year-on-year, with capacity utilization at a high level; both scale and mix are helping. Against that backdrop, the improvement instead narrowed; the most reasonable explanation is that a chunk of cost was absorbed, but not enough to outweigh the mix upgrade.
Inventory turnover days are also lengthening (inventory ÷ single-quarter cost × 91): end of second quarter 78.17 yi against quarterly cost 58.80 yi, 121 days; end of third quarter 96.997 yi against 68.26 yi, 129 days. The extra 8 days, converted at average daily cost of 0.75 yi, is 6.2 yi of materials.
The whole chain is stocking up, and upstream pockets the money first
| Company | Q2 single-quarter gross margin (%) | Q3 single-quarter gross margin (%) | Quarter-on-quarter (pp) | Inventory vs end-Q2 (%) |
|---|---|---|---|---|
| Shennan Circuits (PCB) | 32.41 | 34.85 | +2.44 | +24.1 |
| Wus Printed Circuit (PCB) | 41.28 | 44.02 | +2.74 | +18.2 |
| Shengyi Technology (copper-clad laminate) | 32.63 | 41.63 | +9.00 | +24.8 |
Shengyi Technology makes copper-clad laminate and is Shennan's upstream: third-quarter single-quarter revenue 141.82 yi, up 78.76% year-on-year, single-quarter gross margin up 9.00 points quarter-on-quarter.
Its exact words were "facing rapid increases in raw material prices and continued high-speed growth in demand for high-end products," and it lifted gross margin through volume ramp-up of early-stage developed products and capacity release.
The money from the price increase lands first on Shengyi's income statement. Inventory at the three companies rose 18% to 25% quarter-on-quarter in the same direction; it is not Shennan alone that is betting, it is the whole chain building stock.
Shennan's notes payable plus accounts payable went from 76.72 yi to 88.64 yi, up 11.92 yi. Of the 18.82 yi of new inventory, 11.92 yi is owed to suppliers; it advanced only 6.90 yi of its own.
The transmission chain in one sentence
Where it comes from: page 12 of the interim report — copper-clad laminate and glass fiber cloth, driven by AI computing demand, face "tightening supply and accelerating price increases," and some downstream customers are placing orders early.
Who it passes through: copper-clad laminate makers like Shengyi Technology, with third-quarter single-quarter gross margin of 41.63%.
Where it lands: Shennan's balance sheet, the September 30 column, inventory up 18.82 yi and payables up 11.92 yi; operating cost must wait until this batch of materials enters production, at the earliest the fourth quarter.
The only inventory impact that entered the income statement in the first three quarters was asset impairment loss of 3.15 yi (cumulative year-on-year increase 40.81%; the company noted the main reason as increased provision for inventory write-downs) — that is price declines on old products, not price increases on new materials.
The strongest one-line rebuttal from the bear side
The bear side would say: the gross margin rise is entirely due to the AI product mix and has nothing to do with copper-clad laminate; "not yet transmitted" is unfalsifiable nonsense, and price increases will always be covered up by mix.
There is half a point to it. But two things don't add up: the improvement is 1.36 points narrower than the same period last year; Shengyi's third-quarter revenue rose 78.76% year-on-year, and single-quarter gross margin rose 9.00 points quarter-on-quarter, with volume and price moving together — the price increase really did happen, while Shennan's own capacity utilization is still at a high level, so it is not without bargaining power over downstream.
There is also a harder rebuttal: the price increase was passed through long ago, but Shennan simultaneously raised prices to customers, which is why gross margin was not squeezed. I cannot rule out this possibility — the company only said it "maintains active communication with suppliers and customers" and did not disclose the magnitude of the price increase. If that is the case, what should be watched in the fourth quarter is whether the price increase can continue, not when costs enter the statements.
The watershed is whether the fourth quarter can pass on the upstream increase.
Under what circumstances I would be wrong
If single-quarter gross margin in the fourth quarter falls below 33% and the operating cost ratio rises more than 2 points quarter-on-quarter, it means the price increase had already penetrated into costs in the fourth quarter, and my timing — that it did not enter costs in the third quarter and would only show in the fourth quarter — would be wrong.
Another one: if the share of raw materials in inventory in the annual report clearly drops from 32%, it means this 18.82 yi is not stockpiling but accumulation, and the characterization of "active stockpiling" would also be wrong.
Sources: Shennan Circuits 2026 Third Quarter Report, 2026-10-23, Announcement 2026-065; Shennan Circuits 2026 Semi-Annual Report, 2026-08-27; Shennan Circuits Investor Relations Activity Record, 2026-10-31, No. 2026-24; Shennan Circuits 2025 First Quarter Report, 2025-04-24, Announcement 2025-017; Wus Printed Circuit 2026 Third Quarter Report, 2026-10-28, Announcement 2026-076; Wus Printed Circuit 2026 Semi-Annual Report, 2026-08-26; Wus Printed Circuit 2026 First Quarter Report, 2026-04-23; Shengyi Technology 2026 Third Quarter Report, 2026-10-30; Shengyi Technology 2026 Semi-Annual Report, 2026-08-15; Shengyi Technology 2026 First Quarter Report, 2026-04-29.
这条的前提有两处不成立:深南电路三季报是10月23日披露的,不是10月31日;三季报里也没有"原材料"这个科目,所以"原材料存货比二季度末多了多少"这个数,公开文件里查不到。
能查到的三件事:存货合计比二季度末多18.82亿;单季毛利率没降,从32.41%升到34.85%;覆铜板涨价落在资产负债表,还没落到利润表的营业成本。
10月31日挂的是调研记录
10月23日披露三季报,公告编号2026-065。10月31日挂出来的是投资者关系活动记录表,编号2026-24,记的是10月27日至30日十五家机构调研。
记录里正好有一问问原材料,公司答:主要原材料包括覆铜板、半固化片、铜箔、金盐、油墨等,三季度"受大宗商品价格变化影响,金盐等部分原材料价格同比提升、环比亦有一定涨幅"。
问综合毛利率环比,公司说"环比有一定提升",理由写的是数据中心和有线通信收入增长、封装基板订单释放、广州南通四期泰国工厂爬坡——没提成本。
公司自己也没把毛利率上升归给成本下降。
原材料这一栏,季报里没有
三季报只有一行"存货":9月末96.997亿,年初51.397亿,增88.72%,公司写的变动原因是"业务规模增加以及备料影响"。
明细只在半年报:6月30日原材料账面余额26.03亿(账面价值25.50亿)、在产品19.90亿、库存商品14.26亿、发出商品20.95亿,合计81.14亿,账面价值78.17亿。
所以能算的只有合计:78.17亿到97.00亿,多18.82亿,增24.1%。9月30日的原材料余额不会出现在任何公开文件里——半年报给6月30日,年报给12月31日。
问"原材料多了多少",问的是一个不存在的数。
单季毛利率升了2.44个点,但改善在变窄
拆单季,累计相减:一季度收入65.96亿、成本46.72亿,毛利率29.17%;二季度87.01亿、58.80亿,32.41%;三季度104.78亿、68.26亿,34.85%,单季毛利额36.52亿。
营业成本率从67.59%降到65.15%,成本这一头根本没被顶上来。
去年也是这么走的:2025年二季度27.59%,三季度31.39%,环比升3.80个点。今年同一个三季度只升2.44个点,少了1.36个点。
今年三季度收入同比增66.29%、产能利用率高位,规模和结构都在帮忙。这种背景下改善反而变窄,最合理的解释是有一块成本被吃掉,还不够大,没盖过结构升级。
存货周转天数也在拉长(存货÷单季成本×91):二季度末78.17亿对当季成本58.80亿,121天;三季度末96.997亿对68.26亿,129天。多出的8天按日均成本0.75亿折,是6.2亿的料。
整条链都在囤,上游先把钱赚走
| 公司 | Q2单季毛利率(%) | Q3单季毛利率(%) | 环比(pp) | 存货比Q2末(%) |
|---|---|---|---|---|
| 深南电路(PCB) | 32.41 | 34.85 | +2.44 | +24.1 |
| 沪电股份(PCB) | 41.28 | 44.02 | +2.74 | +18.2 |
| 生益科技(覆铜板) | 32.63 | 41.63 | +9.00 | +24.8 |
生益科技做覆铜板,是深南的上游:三季度单季收入141.82亿、同比增78.76%,单季毛利率环比抬了9.00个点。
它的原话是"面对原材料价格急速上涨、高端产品市场需求持续高速增长",靠早期研发产品放量和产能释放把毛利做上去。
涨价的钱,先落在生益的利润表上。三家存货环比同向增18%到25%,不是深南一家在赌,是整条链在备库。
深南应付票据加应付账款从76.72亿到88.64亿,多11.92亿。18.82亿新增存货里,11.92亿是欠供应商的,自己只多垫6.90亿。
传导链一句话
从哪来:半年报第12页,覆铜板、玻纤布在AI算力需求拉动下"供给趋紧与价格加速上行",部分下游客户提前铺单。
经过谁:生益科技这类覆铜板厂,三季度单季毛利率41.63%。
落到哪:深南的资产负债表,9月30日那一列,存货加18.82亿、应付加11.92亿;营业成本要等这批料投产,最快四季度。
前三季度唯一进利润表的存货影响是资产减值损失3.15亿(累计同比增40.81%,公司注明主因计提存货跌价准备增加)——那是老产品跌价,不是新料涨价。
反方最强的一句话
反方会说:毛利率升全是AI产品结构的功劳,跟覆铜板无关,"还没传导"是不可证伪的废话,涨价永远被结构盖住。
有一半道理。但两点说不通:改善比去年同期窄1.36个点;生益三季度收入同比增78.76%、单季毛利率环比抬9.00个点,量价一起走,涨价是真发生过的,而深南自己产能利用率还在高位,对下游不是没有议价空间。
还有一种更硬的反驳:涨价早就传过来了,只是深南同步对客户提了价,所以毛利率才没被压。这种可能我排除不掉——公司只说"与供应商及客户保持积极沟通",没披露提价幅度。真要是这样,四季度该看的是提价能不能延续,而不是成本什么时候进表。
分水岭是四季度能不能把上游的涨幅传下去。
什么情况下我错了
四季度单季毛利率跌破33%、营业成本率环比抬升超过2个点,说明涨价四季度已经穿透到成本,我说"三季度没进成本、四季度才见"的时点就错了。
另一条:年报里原材料占存货的比重从32%明显往下掉,说明这18.82亿不是备料是积压,"主动备料"的定性也错。
- 01深南电路2026年三季度报告,2026-10-23,公告2026-065
- 02深南电路2026年半年度报告,2026-08-27
- 03深南电路投资者关系活动记录表,2026-10-31,编号2026-24
- 04深南电路2025年一季度报告,2025-04-24,公告2025-017
- 05沪电股份2026年三季度报告,2026-10-28,公告2026-076
- 06沪电股份2026年半年度报告,2026-08-26
- 07沪电股份2026年一季度报告,2026-04-23
- 08生益科技2026年三季度报告,2026-10-30
- 09生益科技2026年半年度报告,2026-08-15
- 10生益科技2026年一季度报告,2026-04-29
先把两处前提纠正掉:三季报是10月23日披露的,不是10月31日;季报里没有"原材料"这一栏。
## 现金先塌了
拆单季:一季度经营现金流2.47亿、净利8.50亿,比值0.29;二季度12.55亿对14.01亿,0.90;三季度4.57亿对18.78亿,0.24。收入越大,现金越收不回来。
存货从78.17亿涨到97.00亿,多18.82亿。同期合同负债从27.02亿到36.30亿,多9.28亿;应付票据加应付账款从76.72亿到88.64亿,多11.92亿。两头一共挪进来21.2亿,比18.82亿还多出2.4亿。
钱还是不够。缺口在另一头:应收账款从72.92亿涨到90.58亿,多17.65亿,应收票据多0.46亿。存货加应收一共占用36.93亿,合同负债加应付补回21.2亿,净占用15.73亿,跟经营现金流对净利的14.2亿缺口基本对得上。
这不是季节性。去年三季度经营现金流8.18亿、净利9.66亿,比值0.85;今年利润接近翻倍到18.78亿,现金反倒少了一半。
这就是涨价的第一个动作——不是抬营业成本,是让深南在收入最好的一个季度里,一边向客户预收、一边向供应商赊账、一边把货堆在仓库,还是没留住现金。
## 差额停在覆铜板那一层
六家摊开看单季毛利率,第一次分配很清楚。
| 环节 | 公司 | 二季度 | 三季度 | 环比pp |
|---|---|---|---|---|
| 覆铜板 | 生益科技 | 32.63 | 41.61 | +8.98 |
| 覆铜板 | 南亚新材 | 21.68 | 38.08 | +16.40 |
| 覆铜板 | 华正新材 | 18.72 | 36.06 | +17.34 |
| PCB | 深南电路 | 32.41 | 34.85 | +2.44 |
| PCB | 沪电股份 | 41.28 | 44.02 | +2.74 |
| PCB | 胜宏科技 | 32.10 | 34.16 | +2.06 |
上游抬9到17个点,下游只抬2到3个点。中间6到15个点的落差,就是涨价在链上留下的比例。
不是深南一家弱。同行沪电、胜宏同样是2个点出头,三家板厂的动作几乎一样齐——这说明压下来的不是某家公司的经营问题,是环节位置。
上游存货也不是低价库存:三家覆铜板厂存货环比至少增了24.8%,一边涨价一边补库,说明毛利率跳升来自真提价,不是库存重估。
反过来看下游:胜宏存货环比增57.0%,比深南还猛,毛利率只抬了2.06个点。囤多少料跟能不能涨价没关系,站在哪一层才有关系。
但环节不是全部答案。沪电三季度单季毛利率44.02%,比生益还高两个多点,收入同比增90.0%的同时毛利率还升2.74个点。它做的是更高端的高多层和HDI,客户换供应商的成本高。
更准的说法:环节决定默认结果,产品结构决定能抵抗多少。深南两头都不占优。
还有一层:深南三季度单季毛利额36.52亿,去年同期19.78亿,增84.6%,快过收入增速66.3%。结构确实在帮忙,不能把2.44个点的改善全归给"涨价还没到"。我的判断只是:结构帮了忙,但没能像上游那样跳起来。
## 两条线在三季度交叉
深南和生益的单季毛利率差,2026年一季度是+1.07个点,二季度翻成-0.22,三季度-6.76——两个季度反转近8个点。
```chart
type: group
title: 深南与生益单季毛利率:三季度上游反超
unit: %
labels: 2026Q1, 2026Q2, 2026Q3
values: 29.17, 32.41, 34.85
values2: 28.10, 32.63, 41.61
legend: 深南电路|生益科技
```
方向别搞反:上游毛利率上去了,下游还没下来。这个"还没"不是好消息,是账没到期。
深南三季末存货周转129天(97.00÷68.26×91),比二季末的121天多8天。按先进先出,9月底进的料要到4.3个月后才进营业成本,2027年1到2月。
所以时间表比"最快四季度"更窄:四季度只吃一部分,2027年一季度吃全额,二季度才开始消化。
## 净利率少涨的那0.6个点
毛利率环比升2.44个点,净利率只升1.82个点(17.92%对16.10%)。差出来的0.62个点,全在费用里。
三费加研发占收入比重从12.16%升到13.04%,多0.88个点;资产减值从1.33%降到0.85%,回补0.48个点,两头基本对冲。拆开看,财务费用从0.82亿涨到1.48亿,接近翻倍;管理费用率从4.72%到5.13%。
财务费用翻倍是囤料的第二张账单。18.82亿存货里,11.92亿是欠供应商的,剩下6.90亿得自己垫,垫的钱要么是借款要么是自有资金,利息落在财务费用,不落在营业成本。
## 反方最强的一条
反方最有杀伤力的不是"结构升级盖过涨价",而是这个:涨价早就传过来了,深南只是同步对客户提了价,所以毛利率才没被压。
这条我排除不掉。三季报和调研记录里,公司只说"与供应商及客户保持积极沟通",没披露提价幅度。合同负债环比增34.4%、快过收入环比20.4%,可以说客户在预付,但不能证明是提价。
还有一条:覆铜板价格可能已经在三季度见顶。若如此,这18.82亿就不是"未来成本压力",而是"高价库存等着计提跌价"——方向对,原因全错。
两条剧本互相排斥,但前瞻指标是同一个:生益四季度的单季毛利率。掉到36%以下、南亚掉到38%以下,是见顶;还在40%以上,是涨价延续。
## 什么数字出来我错了
证伪条件五条:
- 四季度单季毛利率不低于34.85%:涨价根本没进成本,"四季度开始吃"就是错的。
- 四季度落在33%到34.85%之间:方向对,时点后移一个季度。
- 年报里原材料占存货比重跌破30%:18.82亿不是备料是积压,定性错。
- 四季度单季经营现金流对净利回到0.6以上:现金压力是一次性的,不是结构性的。
- 生益四季度单季毛利率跌破36%、南亚跌破38%:换成"涨价见顶、库存减值"剧本,前面整套成本传导失效。
## 一句话的传导链
铜箔和玻纤布涨,传到覆铜板,传到PCB,最后传到服务器主板报价。落到报表上分三处:上游落毛利率,中游落存货和应付票据,下游客户落预付和合同负债。现在钱停在第二层,第三层还没开始付。
压力已经在往最上端推。覆铜板厂一边涨价,一边也在赊它的上游的钱:
| 公司 | 应付票据环比(%) | 存货环比(%) |
|---|---|---|
| 生益科技 | +47.5 | +24.8 |
| 华正新材 | +51.0 | +37.5 |
| 南亚新材 | +0.3 | +54.7 |
南亚最直接:存货增了54.7%,应付票据几乎没动,等于现款抢料。玻纤布和铜箔这一层,比覆铜板更紧。
所以别看深南四季度的营业成本,先看生益四季度毛利率。上游的毛利率,就是下游三个月后的成本。
来源:深南电路2026年三季度报告,2026-10-23,公告2026-065;深南电路2026年半年度报告,2026-08-27;深南电路2026年一季度报告,2026-04-24;深南电路2025年三季度报告,2025-10-30;深南电路投资者关系活动记录表,2026-10-31,编号2026-24;沪电股份2026年三季度报告,2026-10-28,公告2026-076;沪电股份2026年半年度报告,2026-08-26;生益科技2026年三季度报告,2026-10-30;生益科技2026年半年度报告,2026-08-15;生益科技2026年一季度报告,2026-04-29;南亚新材2026年三季度报告,2026-10-28;华正新材2026年三季度报告,2026-10-23;胜宏科技2026年三季度报告,2026-10-30。