[Hong Kong Stock IPO] Zhongji Innolight Subscription Status and New Share Issuance Analysis

Company Name: Zhongji Innolight Corporation Ltd. (03308.HK, hereinafter referred to as "the Company")

Sponsors: Goldman Sachs, CICC, Morgan Stanley, GF Securities

Greeshoe (Stabilization Support): Yes (Goldman Sachs)

Cornerstone Investors: 49.05%

Subscription Period: July 22 - July 27 (Next Monday)

Main Business: Global Leader in Optical Modules

I. Sponsors, Greeshoe, and Cornerstones

Joint sponsors by Goldman Sachs, CICC, Morgan Stanley, and GF Securities, with Goldman Sachs providing the greeshoe support. The cornerstone lineup is absolutely insane: 35 top-tier domestic and international cornerstones, including BlackRock, JPMorgan, Tencent, and multiple sovereign wealth funds. The luxury of this lineup has surpassed that of Zijin Gold International from before,making it the most badass company for Hong Kong IPO cornerstones so far, without a doubt..

New Stock Subscription Index: ★★★★★

II. Share Volume and Margin Status

Mechanism B, 10%, public allocation of 109,000 lots. It doesn't look like much at first glance, but actually, with one share priced at 1010 HKD, the entry fee for one lot exceeds 50,000 HKD.

This really isn't the company's fault. According to the new regulations on July 2, the number of shares per lot can only be chosen from 1, 50, 100, 500, 1,000, 2,000, 5,000, or 10,000 shares, and the value of each lot must be greater than or equal to 1,000 HKD and less than or equal to 50,000 HKD (applicable to stocks with more than 100 shares per lot).

So the company basically had no choice. One share per lot is definitely out of the question. Not only would the shares be too fragmented, causing the chip structure to collapse, but even a small drop would require reverse stock splits, which is too troublesome.

Going higher, the situation of 500 shares or more does not comply with the regulation that "stocks with more than 100 shares per lot must have a value less than or equal to 50,000 HKD." Therefore, there are only two choices: 50 or 100.

Fortunately, the company chose 50. If they had chosen 100, the value per lot would exceed 100,000 HKD, making it even harder to win an allocation.

Currently, margin subscription is at 1.8 times, with a total margin amount of 10.3 billion HKD. Given the sharp decline in the A-share market today, achieving this result is indeed quite decent.

New Stock Subscription Index: ★★★★

III. Valuation Assessment and Fundamental Analysis

The company's main business is optical modules, which convert electrical signals sent by servers, switches, GPUs, and other devices into optical signals, and then transmit them to other devices via fiber optics.

Currently, AI data centers require tens of thousands of GPUs. These GPUs must continuously exchange massive amounts of data. Therefore, the larger the scale of the AI computing cluster, the more high-speed optical modules are needed, and the transmission rates must constantly upgrade. This gives rise to what we call optical modules with transmission rates such as 400G, 800G, 1.6T, and 3.2T.

Optical modules of 800G and below are already very mature. 1.6T is exclusive to a few global top players, while 3.2T is still in the R&D stage.

Zhongji Innolight is outstanding because its global market share ranks first; 1.6T has completed mass production shipments, with high order visibility covering up to 2027; meanwhile, it also showcased the future technology path for 3.2T at the 2026 OFC, potentially being one of the first batch of players to achieve 3.2T.

Except for Coherent, which stands shoulder-to-shoulder with Zhongji Innolight, the optical module technology and commercialization progress of other comparable companies generally lag behind Zhongji Innolight by one to two quarters or more.

Benefiting from AI data centers elevating optical modules from "network equipment accessories" to "core infrastructure for computing clusters," Zhongji Innolight's performance in recent years has been very impressive. Full-year revenue in 2025 reached 38.24 billion, maintaining a compound growth rate of over 60% for three consecutive years. Net profit attributable to shareholders was 10.74 billion, doubling year-on-year.

Regarding expectations, according to a summary of forecasts from 31 institutions by Tong H Shun as of July 17, the median forecast for net profit attributable to shareholders in 2026 is 30 billion, representing a 178% year-on-year increase, with revenue of 100.9 billion, a 164% year-on-year increase.

This forecast is not baseless. According to the Q1 report this year, the company's Q1 revenue increased by 192.1% year-on-year, and net profit increased by 264.6% year-on-year.

In summary, the company's fundamentals are indeed very attractive, making it one of the biggest beneficiaries of the AI wave. Thus, the core question shifts from whether the company is good to whether this IPO is priced expensively? Is there money to be made in subscribing to new shares?

The upper limit of the issuance price for this offering is 1010 HKD, corresponding to a discount rate of 17.8%. Yesterday, the discount rate was 23%, but today it dropped another 6 percentage points.

Some people compare Zhongji Innolight with CATL. CATL currently has a premium of 41.7%. If it truly benchmarks against the "Ning King" (CATL), assuming the A-share price remains unchanged, there is still a 72.4% upside potential in the Hong Kong market.

However, I believe that although Zhongji is indeed the leader in optical modules, it is unlikely that the discount rate will benchmark against CATL.

The key lies in two points. First, CATL's lithium battery sector presents a pattern of one superpower and several strong competitors, with a clear gap between the leader and the second tier. In contrast, in the optical module sector, there are dual leaders at the top of the pyramid: Zhongji Innolight and Coherent. Moreover, the generational gap for the second tier, represented by Eoptolink, is not that large. For instance, Eoptolink explicitly stated in September 2025 that it achieved mass shipment of 1.6T products, only one quarter later than Zhongji's Q2 shipment.

Furthermore, CATL's 39% global market share is crushing (BYD, the second place, has only 13.7%). Zhongji Innolight's 21% share across all categories is merely "leading" rather than "monopolistic." Therefore, first, the leader premium is not as high as CATL's.

Secondly, CATL is the chain master of the entire industry chain. It not only extends upstream to lock in long-term contracts for lithium ore resources but also binds deeply with downstream automakers, even participating in pricing. During periods of industry overcapacity, where second-tier manufacturers suffer widespread losses, CATL maintains profitability through economies of scale, and the Matthew Effect continues to intensify.

In contrast, Zhongji is very strong in the midstream packaging segment, but lacks sufficient depth in the industry chain, facing stronger players both upstream and downstream.

This brings us to Coherent. Coherent has extremely strong autonomous capabilities in core components, producing InP wafers, EMLs, CW lasers, PDs, silicon photonics chips, etc., capable of manufacturing almost everything. It resembles a "full industry chain platform covering optical chips, optical components, optical modules, and even CPO," whereas Zhongji relies more on upstream suppliers.

This is why, if aligning fiscal years with calendar years, Zhongji's forward valuation is 10%-20% lower than Coherent's, precisely because Coherent's platform value is higher.

Therefore, as I wrote in my article yesterday, I expect a reasonable discount rate of 10%, giving a range of 5%-15% with a 5% margin. Today, it is already close to the lower bound. If it drops further, it might not be very interesting anymore.

Luxshare Precision, which switched from electronic contract manufacturing to optical modules two days ago, currently has a discount rate of 17.5%. Zhongji is naturally better than Luxshare, but given the current sentiment, I do not recommend being overly optimistic about how much the discount rate can narrow.

Finally, a complaint: it was said to be issued at an 80% discount on Monday, but after rising yesterday, the issuance price was adjusted again. Retail investors are one thing, but isn't this treating anchored and cornerstone institutional investors like Japanese people (being messed with)?

The A-shares fall every day, and just as they rebounded for one day, you raise the price, giving no room for maneuver. Who will support you after listing? No wonder there are so many pure anchors in the market. Perhaps these cornerstones have long-term confidence in the company, but short-term uncertainty is too high.

Sell the anchors first to make a small profit, then talk?

New Stock Subscription Index: ★★★~★★★★★ (Final strategy and rating will be posted on Xingqiu) $ZJ INNOLIGHT(03308.HK) $Zhongji Innolight(300308.SZ)