Consensus says Zhongji Innolight (300308) fourth-quarter revenue can still grow quarter-on-quarter. In the annual report before April 30, 2027, will fourth-quarter single-quarter revenue be lower than third-quarter?

Consensus says Zhongji Innolight (300308) fourth-quarter revenue can still grow quarter-on-quarter. In the annual report before April 30, 2027, will fourth-quarter single-quarter revenue be lower than third-quarter?

Let me put the conclusion first: single-quarter revenue for Q4 2026 is highly likely to still be above 277.69 yi, I give it a 75% probability, with the landing point seen at 295 yi to 315 yi, quarter-on-quarter +6% to +13%, and full-year revenue roughly 990 yi to 1,010 yi.

But the second half of that sentence is more valuable—the quarter-on-quarter growth rate will come down. Q2 quarter-on-quarter +14.3%, Q3 +24.6%, Q4 is highly likely to be single-digit to low double-digit. The slope of the first three quarters cannot be extrapolated.

The basis for judgment is not "the industry is still rising," but that in Q4 the plus items in this business outnumber the minus items, and every plus item can be traced to the financial statements.

277.69 yi was pushed out by capacity and materials together

Q3 single-quarter revenue was 277.69 yi, year-on-year +171.83%; the first three quarters cumulative 695.47 yi, year-on-year +178.13%. Looking at the two numbers together, the cumulative growth rate is even a bit higher than the single-quarter growth rate, indicating that the acceleration in Q3 is part of the full-year rhythm, not something that popped out in a single quarter.

Single-quarter gross margin was 46.06% in Q1, 46.42% in Q2, 46.61% in Q3 (Q3 single-quarter cost of 148.26 yi equals the first three quarters' 372.81 yi minus the first half's 224.55 yi, then divided by 277.69 yi revenue to calculate). Revenue is accelerating while gross margin is still rising; there is only one explanation: the share of high-end products is increasing, and prices have not collapsed.

The company's October 29 conference call put it very directly: Q3 prices were generally stable compared with the first half, industry quotations were rational, and no vicious competition appeared.

The annual price reduction range is limited, and some product prices have even not seen price cuts.

This is the original wording from November 3 when answering about 2027 prices. Annual price reduction negotiations generally fall at the beginning of the year for re-quoting, and Q4 does not get the new price cuts; this item is a net positive for Q4.

There is an easily misread point on the profit side. Q3 net profit attributable to the parent was 85.87 yi, consolidated net profit 93.62 yi, and the 7.75 yi difference is minority interest income/loss. Attributable to the parent is even lower than the 90.18 yi excluding non-recurring items, separated by 6.2 yi of fair value loss on upstream supplier stocks and 3.9 yi of exchange loss on Hong Kong dollar raised funds. These two items do not enter revenue, yet they muddy the statement that "profit rose only 8.5% quarter-on-quarter."

Inventory was 238.78 yi, up 88.30% from 126.81 yi at the beginning of the year, which looks scary. Divide it by single-quarter operating cost: at the beginning of the year it was 1.73 quarters' worth, and at the end of Q3 it was 1.61 quarters. Stockpiling is rising slower than shipments; it is material stocking, not accumulation.

Accounts receivable were 173.09 yi, +175.76% from the beginning of the year, basically in sync with the first three quarters' revenue +178.13%, and there is no problem in proportion. But it is already 62% of single-quarter revenue; collection is slower than revenue recognition, and this item needs to be watched when looking at cash flow, though it does not affect the direction of Q4 revenue.

Capacity is also being laid out in sync. Fixed assets were 124.55 yi, +75.91% from the beginning of the year; construction in progress 54.95 yi, +286.46%; cash paid for purchases and construction of long-term assets in the first three quarters was 91.61 yi, versus 16.15 yi in the same period last year.

The first plus item for Q4 is a timing difference

Revenue has to pass through five stages from customer order placement to entering the income statement: cloud vendor capital expenditure, customer PO, materials (DSP, EML, CW light source, silicon photonics chips), capacity (Thailand, Tongling), shipment confirmation. If any link gets stuck, the number falling on the "operating revenue" line is reduced by a chunk, and the missing chunk usually does not disappear; it is moved later.

Q3 got stuck at the last stage—goods were made but stuck in logistics. Rain in Thailand affected logistics, some finished goods shipments were delayed, and semi-finished and finished goods were pressed into inventory.

After entering Q4, the relevant impact has been eliminated, and shipments had returned to normal in early October.

In the same Q&A, the company said this part "is mainly reflected in revenue as a timing difference." In plain terms: what was under-recognized in Q3 is made up in Q4.

How much can be made up? The company said the flood impact "accounts for a not-high proportion of the inventory increase." Q3 inventory increased by 40.52 yi versus Q2; using no more than 20% as the upper bound gives 8.1 yi of cost, which at a 46.6% gross margin converts to about 15 yi of revenue, around 5% of Q3 single-quarter revenue. It is icing on the cake, not the main force.

The main force is materials. When asked on November 3 about core materials for Q4, the answer was:

Material supply is gradually improving, and it is expected that some core materials will improve marginally in the next few quarters, and the company's shipment volume is also expected to maintain quarter-on-quarter increases for several consecutive quarters.

"Shipment volume maintaining quarter-on-quarter increases for several consecutive quarters" is management's most direct statement on Q4 and next year's Q1. On the same day, they added that data center customers' pickup pace is normal and has not slowed.

Orders are not lacking either. The October 23 announcement said "major customer orders have been placed through 2027," and on November 3 they said most 2027 orders already cover the full year, and clarified that these are definite POs given by customers, not demand guidance. With POs in hand, there is no year-end order gap in Q4.

Three minus items. Base step-up: 277.69 yi is already 1.42 times Q1. Materials: DSP remains "supply relatively tight, prices relatively firm," and 1.6T ramp-up requires more DSP. Customers: on November 3 they acknowledged "individual customers have product structural adjustments"; that sentence refers to 2027, but the same adjustments can also fall on Q4 pickup pace.

There is another layer that is easy to overlook. The bulk of revenue is denominated in US dollars. In the third quarter, exchange losses already appeared from the weakening of the US dollar and Hong Kong dollar against the renminbi. If the renminbi continues to strengthen in the fourth quarter, the same US dollar shipments converted into renminbi will be worth less, shaving off one or two points.

Let me also calculate the implied numbers in the consensus. The consensus 2026 earnings per share forecast from 26 institutions is 29.45 yuan, which at 11.78 yi shares translates to about 347 yi yuan of net profit attributable to shareholders; subtracting the 222.38 yi yuan from the first three quarters, the fourth quarter needs to deliver 124.5 yi yuan, up 45% from the third quarter. Quarter-on-quarter revenue growth can only explain part of that; the rest must come from expense ratio, tax rate, and reversal of special items. The truly tight part of that consensus statement is profit, not revenue.

In the peer corridor, someone has already fallen 13% quarter-on-quarter

Revenue unit is yi yuan; single-quarter figures are restored by subtracting the prior period cumulative from the current period cumulative.

Four of the five companies saw quarter-on-quarter growth in the third quarter; one fell 13.0%, and that was New Easy Sheng, the second-largest by volume. In the same period, its inventory was +140.70% versus the beginning of the year, and prepayments were +7695.01%. Even after stocking up to that extent, revenue still fell, which shows that what was stuck was not orders but delivery or customer pickup.

This sentence is the key point of the whole piece: the quarter-on-quarter direction of a single quarter has little to do with total demand and a lot to do with the rhythm of pickup. So "consensus expects fourth-quarter growth" and "will the fourth quarter be lower than the third quarter" are two different questions. The former is about the industry; the latter is about one company's shipping schedule in a given quarter.

Last year's fourth-quarter column is also worth looking at. Zhongji +29.6%, New Easy Sheng +37.4%, Tianfu -14.9%, Taichenguang -13.8%. In the same quarter, some rose 30% and some fell 15%; industry prosperity does not resolve that difference.

Six upward moves in seven years, but no regularity in magnitude

Median +17.2%; the only negative was 2020 (-5.9%). There is a directional bias, but no regularity in magnitude. 2022 was +5.4%, and 2024 was +0.5%, both occurring in years of upward prosperity.

So "the past four quarters all rose" cannot be used as an argument; the 2024 instance was almost flat. Conversely, "it has fallen before" also cannot be used to infer that it must fall this year.

To replicate last year's +29.6%, the fourth quarter would have to deliver 359.9 yi yuan, which means raising another 30% on top of the 277.69 yi yuan step, while the company's wording on materials is only "marginal improvement." That is why I place the landing point at 295 yi to 315 yi yuan, rather than above 350 yi yuan. Last year's slope was given by a special year, not the norm.

In this company's quarter-on-quarter changes, price is almost a constant—annual price reductions are negotiated at the beginning of the year, and prices are stable within the quarter; the company said the same on October 29. So whether fourth-quarter revenue goes up or down almost leaves shipment volume as the only variable. That is also why I bet the judgment on materials and shipping rhythm, not on demand guidance.

The strongest counterattack from the opposing side

The smartest opponent will not say demand is about to collapse. He will say three things.

First, 2024. That year's fourth quarter was 65.50 yi yuan, only 0.5% higher than the third quarter's 65.14 yi yuan, almost flat. For a company whose full-year revenue nearly doubled, quarter-on-quarter growth in the fourth quarter can be almost zero; this is a fact that has already happened.

Second, New Easy Sheng. In the same quarter it fell 13%, and stocking was still increasing. Zhongji itself admits that "individual customers have product structural adjustments." If these two things are the same action by the same batch of customers, replicating it in the fourth quarter is entirely possible.

Third, materials. The company's words are "gradual improvement" and "marginal improvement," not "already resolved." Read the statement that DSP prices remain firm in reverse, and it means supply is still insufficient.

I accept the first two of these three points; the third can only count as half. Materials indeed have not been resolved, but third-quarter revenue of 277.69 yi yuan was achieved while materials were still relatively tight; in the fourth quarter materials will only be looser, not tighter.

I do not intend to explain away the 2024 flat quarter; it is precisely the reason I leave that 25% probability. The only difference I can state is this: this year's third quarter had clear shipment delays, and the fourth quarter has catch-up, while 2024 did not have this variable. As for possible reasons such as customers waiting until year-end and new-year price negotiations, I cannot prove them from the documents, so I treat them only as assumptions, not as arguments.

What numbers would make me admit I am wrong

• If full-year operating revenue is below 973.16 yi yuan (695.47 plus 277.69), that means the fourth quarter is below the third quarter, and this judgment is wrong. Watch the annual report at the end of April 2027; if a performance express report comes out before the end of February, it will be seen earlier.
• If the fourth-quarter single-quarter gross margin falls below 44.50% (the level in the fourth quarter of 2025), it means the premise of "stable prices, improving structure" is wrong, and revenue quality must be re-evaluated.
• A signal that does not require waiting for the annual report: public disclosure during the fourth quarter of customer order delays, or the next investor relations record changing its wording to say the pickup rhythm is slowing.

One more reminder. Even if the direction is right, if the fourth quarter only grows slightly (below +3%), it should also be treated as a warning for 2027—the company's exact words are "shipment volume has increased quarter-on-quarter for several consecutive quarters"; if full-year POs are covered and only slight growth is achieved, then the problem is not in orders.

Sources: Zhongji Innolight "2026 Third Quarter Report", October 30, 2026, announcement no. 2026-101; "Announcement on the Company's Recent Production and Operation Status", October 23, 2026, announcement no. 2026-100; "Investor Relations Activity Record Form" no. 2026-009, October 29, 2026; "Investor Relations Activity Record Form" no. 2026-010, November 3, 2026; "2026 Semi-Annual Report", August 22, 2026; Eoptolink "2026 Third Quarter Report", October 30, 2026, announcement no. 2026-065; TFC Optical Communication "2026 Third Quarter Report", October 29, 2026, announcement no. 2026-064; Accelink Technologies "2026 Third Quarter Report", October 26, 2026; T&S Communications "2026 Third Quarter Report", October 30, 2026. Consensus expectations are taken from East Money F10 earnings forecasts, November 4, 2026; single-quarter gross margin, single-quarter operating cost, and inventory coverage in quarters are calculated from the data in the above reports according to the methods specified in the text.