At midday on September 10, LONGSYS (09976.HK) traded at HK$218.40, trimming its gain to 4% after touching an intraday high of HK$225.20 before retreating, with turnover reaching HK$205 million. On the same day, Shenzhen Longsys Electronics Co.,Ltd. (301308.SZ) stood at RMB 348.89, up a marginal 0.25%, with a TTM price-to-earnings ratio of 13.38 times. The HK-listed shares carried a TTM multiple of just 7.28 times, underscoring a persistent valuation gap between the two listing venues. Two days earlier, this Shenzhen-based storage leader had completed its Hong Kong IPO. On September 8, LONGSYS debuted on the Main Board of the Stock Exchange of Hong Kong, becoming the first memory enterprise to achieve dual "A+H" listings. The H-share offer price of HK$236 represented a discount of more than 40% to the A-share closing price of RMB 358.22 on September 7, and stood nearly 58% below the RMB 560 private placement price from a month earlier. The company reported first-half net profit attributable to shareholders of RMB 10.577 billion, a staggering year-on-year surge of over 71,500 times, backed by cornerstone investors including Transsion and Lenovo. Yet the shares broke below the issue price on day one, closing at HK$233.60, down approximately 1.02%. Peak earnings meeting a cold market reception, deeply underwater placement investors alongside strategic industry capital — together these weave a tapestry of paradoxes surrounding this capital extravaganza. Through interviews with an international veteran investment banker, Huang Lichong, president of Huisheng International Capital, industry experts, and LONGSYS' investor relations team, this investigation reconstructs the pricing logic and cyclical dynamics behind the drama.

RMB 358 on A-Shares, HK$236 on H-Shares

The HK$236 issue price for LONGSYS' H-shares, roughly equivalent to RMB 202, sat about 43.6% below the A-share closing price of RMB 358.22 on September 7. Meanwhile, that A-share price itself had already fallen approximately 36.0% from the RMB 560 placement price. Simply summing those two discount rates yields a figure of nearly 64% below the placement price — a number that has sparked widespread debate in the market. "The so-called 64% discount to the placement price combines share price movements across two different time points and the price differential between two markets for the same company. It cannot be entirely attributed to H-share discounting, nor should the two discount rates be directly added together," Huang Lichong stated bluntly. He argued that the A+H structure has not failed; rather, it is untenable to treat the A-share peak as the valuation floor that Hong Kong must accept. He further explained that the investor bases in the two markets are fundamentally different: A-share investors can only access Hong Kong stocks through special channels and cannot buy stocks in other markets, whereas H-share investors have global options. "Differences in required returns, available investment targets, float supply, trading depth, and currency risk between the two markets can all lead to price divergence." Institutional frameworks also do not support forced convergence — the Securities and Futures Commission of Hong Kong has clarified that A-shares and H-shares cannot be directly transferred between the two exchanges, ruling out arbitrage that would eliminate the gap by buying H-shares and converting them to A-shares. "However, these factors explain why prices may differ; they do not automatically prove that a 40% discount is precisely justified," Huang emphasized. He stressed that the same set of normalized cyclical earnings and cash flow assumptions must be tested against both price levels. "A higher price demands stronger earnings delivery; a lower price may still price in optimistic growth. Even if the gap narrows in the future, it could happen through H-share gains, A-share declines, or a combination of both — convergence should not be interpreted as a promise of H-share appreciation." Market performance on September 10 validated this view. Hong Kong-listed shares surged in morning trade before retreating amid rising selling pressure, while A-shares moved more steadily. The two markets reacted to the same half-year report with distinctly different pace and intensity. LONGSYS' 2026 interim report showed first-half revenue of RMB 24.088 billion, up 136.26% year-on-year; net profit attributable to shareholders of RMB 10.577 billion, up 71,528.66%; and non-GAAP net profit attributable to shareholders of RMB 10.047 billion, up 31,096.33%. This growth rate is rare in the A-share semiconductor sector. Yet market pricing of results has never been solely about growth numbers. On debut day, LONGSYS' H-shares closed at HK$233.60, down about 1.02% from the issue price. That same day, Reuters reported the Hang Seng Tech Index fell roughly 1.6%, providing some external headwind from a weak broader market. "Strong short-term results are undeniable; the market is not disputing the profit itself. What the market demands is sustained long-term performance — it is reluctant to price the best half-year as the new normal for every future half-year," Huang observed. He noted that a first-day dip below the issue price is insufficient evidence that the entire company has been rejected by the market. He specifically flagged that the prior-year comparable net profit base was only about RMB 14.77 million, meaning the "71,500 times growth" headline carries far more shock value than explanatory power for future growth rates. The non-GAAP adjustment strips out non-recurring items, not the cyclical nature of the storage industry. "I would not frame A-shares and H-shares as two entirely different schools of finance. Both markets have investors focused on fundamentals and long-term value; the difference lies more in who is the marginal buyer at any given time and what assumptions they accept." For strongly cyclical companies, Huang believes investors should first examine gross profit and earnings after excluding abnormal cyclical peaks, then assess how much cash is tied up in inventory and receivables, and finally evaluate whether R&D and capital expenditure can generate cross-cycle returns. "Mechanically doubling half-year profits and declaring a low P/E ratio risks mistaking an earnings peak for a valuation trough. For instance, if sustainable profit were only half that annualized rate, the normalized P/E at the same price would double — this is an arithmetic illustration, not a forecast of LONGSYS' earnings halving." "If the share of high-end products, customer repurchase rates, and cash returns continue to improve going forward, the market may reasonably upgrade its assessment of earnings durability; conversely, if profits stem primarily from favorable price spreads, investors should leave room for a cyclical reversal," Huang concluded. "High growth captures attention, but cross-cycle cash generation capability is what supports valuation."

Who Better Understands This Investment: Industry Capital or Financial Capital?

In this H-share offering, 14 cornerstone investors subscribed for approximately USD 151 million, a roster that included industry players such as Transsion and Lenovo alongside asset management and financial institutions. Meanwhile, among the 21 subscribers to the A-share private placement were industry participants like SigmaStar and a subsidiary of Sungrow Power. Given the different pricing timing, markets, lock-up conditions, and investment objectives between the two transactions, market views on the two groups have diverged. "Cornerstone investors also suffer losses frequently, which is why professional financial investors do not rely on cornerstones' judgments to buy stocks — they conduct their own analysis," Huang said. He argued that the framing of "industry capital versus financial capital" is itself imprecise; one cannot simply compare two price points and declare one group savvy and the other stranded. He clarified that cornerstones subscribe at the IPO price, typically with a six-month lock-up, and do not receive any special discount beyond the HK$236 price. The so-called discounted entry refers to the discount relative to the A-share price, not cheaper terms compared to ordinary H-share subscribers in the same offering. "A first-day decline is equally detrimental to the market value of cornerstone holdings, and the lock-up does not constitute capital protection." Industry investors may simultaneously value supply stability, joint R&D, product validation, and partnership opportunities, all of which can influence their willingness to invest. "But customer status is not a commitment of future orders, and potential synergies are not realized shareholder returns. Absent evidence that they paid a higher per-share subscription price, one cannot label such motivations as an observed 'strategic premium,' nor infer undisclosed purchasing preferences or benefit arrangements." From a banking perspective, cornerstone commitments help reduce offering uncertainty and support bookbuilding, but they do not convince independent-minded investors. "The market must still independently scrutinize the issue price and monitor supply changes after lock-up expiry. Cornerstones commit to subscription and lock-up, not to guaranteeing returns for other investors." For placement institutions, the gap between the RMB 560 placement price and the current A-share price of RMB 348.89 represents an unrealized loss exceeding 30%. The HK$236 H-share price sits nearly 58% below the placement price. This price chasm has made the placement institutions' paper losses a focal point of market attention.

Beyond the placement investors' losses, another market focus is whether LONGSYS' decision to complete the H-share financing during a period of heightened storage industry prosperity was well-timed — specifically, whether it picked the top of the storage cycle. "Raising capital during strong industry conditions can be a rational business decision, but the real test is whether the proceeds can reduce dependence on the next cyclical upturn," Huang said. "However, as of now, I would not characterize this point as the storage cycle peak solely based on slowing price increases." He cited TrendForce's forecast of a 13%-18% quarter-on-quarter increase in traditional DRAM contract prices in Q3, noting this still constitutes "continued increases," and that as of September 8, the third quarter had not concluded, so the projection cannot yet be treated as actualized figures. "A slowdown in price acceleration warrants vigilance, but slower increases don't mean declines have begun; server and consumer end-markets may also be at different points in the cycle." Nevertheless, Huang also cautioned that this does not mean valuations have not priced in the best-case scenario. "The key issue LONGSYS needs to explain is cash absorption," Huang told reporters. The interim report showed net operating cash outflow of RMB 3.151 billion, with period-end inventory of RMB 25.777 billion, up approximately RMB 14.1 billion from RMB 11.678 billion at the end of last year, which the company attributed to strategic stockpiling. Cash flow supplementary notes also indicated that inventory buildup was a significant factor behind the gap between net profit and operating cash flow. "One cannot directly accuse the profit of being distorted, but neither can the label 'strategic stockpiling' exempt inventory from scrutiny regarding sales, collection, and write-down risks," Huang analyzed. For module and storage product companies during price upswings, shipping lower-cost inventory can inflate current profits, yet replenishing new inventory already requires higher procurement costs. If product selling prices fall faster than cost digestion, both gross margin and cash flow could come under simultaneous pressure. "So I would examine inventory composition, turnover, customer order coverage, and the pass-through of selling prices versus procurement costs, rather than focusing solely on record net profit. Stockpiling advantages during an uptrend do not automatically translate into competitive advantages in the next phase." According to CFM flash memory market data, the global storage market reached USD 137.14 billion in Q1 2026, up 81.6% quarter-over-quarter. However, multiple industry research firms have simultaneously flagged that storage price increases may slow in Q3, making module makers' inventory management capabilities a key variable affecting second-half profitability. LONGSYS' inventory increase of approximately RMB 14.1 billion in the first half is at a relatively high level within the storage module industry. Strategic stockpiling is a rational choice during price upcycles, but investors need to monitor whether inventory turnover days keep elongating and when operating cash flow turns positive. According to the prospectus, approximately 78.3% of net proceeds are planned for chip design and high-end storage R&D. Huang believes this at least suggests the financing cannot be simplistically characterized as pure "peak cashing out," though stated use of proceeds does not equal realized returns. "If the cycle turns down while R&D, inventory, and capital expenditure remain elevated, the post-listing period may face dual pressure on earnings and valuation multiples; if investments genuinely improve products and cash returns, they could conversely strengthen counter-cyclical resilience. Financing can be timed for good weather, but valuations must factor in bad weather."

Hard-Tech A+H: Cyclical Stock or Growth Stock?

Beyond financing timing, LONGSYS' H-share pricing is not an isolated event. In recent years, cases of H-shares trading at premiums to A-shares — such as CATL and GigaDevice — sufficiently refute the simplistic assertion that "Hong Kong is always cheaper than the mainland." "The market is disaggregating the unified 'hard-tech' label, not simplistically assigning low multiples to cyclicals and high multiples to growth stocks, but rather reassessing: how much sustainable profit and cash can technological advantages actually retain?" Huang said. He separates the variables determining valuation gaps into two layers: at the company level — technology and customer moats, revenue recurrence, cross-cycle margins, and cash conversion; at the trading level — respective float supply in A/H, trading depth, index and capital accessibility, investors' alternative options, as well as currency and holding costs. "The former determines how investors assess the company; the latter explains why the same company can trade at different prices in two markets. H-share premiums can sometimes reflect short-term supply scarcity — one should not attribute the entire premium to 'international capital certifying higher intrinsic value.'" Growth and cyclicality are not mutually exclusive identities. AI demand can drive long-term growth, but it does not guarantee stability in procurement prices, inventory, or end-demand; traditional module makers that develop more controller, software, and packaging/testing capabilities and establish stable inroads with high-end customers may also reduce reliance on pure price spreads. LONGSYS' interim report disclosed capability-building in chip design, firmware, and packaging/testing, so the company cannot be simplistically equated with a trading house; however, whether technology investments translate into more stable profits and cash requires operational validation going forward. "Thus, LONGSYS' 40% discount is not a permanent label, nor are other companies' H-share premiums a lifetime badge of honor," Huang told reporters. "If the company can demonstrate that future incremental profits derive more from product and customer capabilities than solely from storage price increases, there is a basis for repricing; otherwise, even the 'tech growth' narrative cannot shield it from cyclical discounting. The market ultimately must distinguish between companies that can convert technology into sustained cash flows and those that can only convert price increases into temporary profits."