601012.SHG

The addressable pie is enormous but the company is not creating it. IEA PVPS put 2024 global PV additions at 554.1GW to 601.9GW with cumulative installed capacity around 2.25TW, and China alone had likely crossed 1TW. LONGi is fighting for share and margin inside an already-huge, oversupplied commodity market, not opening a new end-market. BC modules are a differentiated product within the existing module market, not a new category, and the PotisEdge storage acquisition is an entry into an adjacent market that already exists, not one LONGi is creating. The real ceiling is not volume, which will keep growing with global electrification, but margin capture: IEA PVPS says annual demand would need to move above 1TW just to absorb current overcapacity, so LONGi's growth ceiling is set by how much of a already-large pie it can keep at a profit, not by how big the pie gets.

No, not on the report's own numbers. The three-year scenarios (to 2028) run from CNY 70bn-75bn conservative, CNY 78bn-82bn base, to CNY 90bn-95bn optimistic, all measured against 2025 revenue of CNY 70.35bn. Doubling would require roughly CNY 140bn, which none of the three cases approaches; even the optimistic case is only about a third above the 2025 base. To the extent any recovery happens, it is a mix story, not a volume or price story: BC's share of module shipments is the variable the report tracks (already about two-thirds of shipments in Q1 2026, targeted above 70% in the base case), while total wafer and module volume targets for 2026 (around 100GW wafers, 80GW modules) are only modestly above current levels. Price is not a tailwind because the industry is still in an oversupplied price war, and storage is explicitly called immaterial in the conservative case and only small and visible in the base case, so it is not a near-term revenue driver.

The report treats BC modules as the current growth engine, already running at roughly two-thirds of module shipments in the first quarter of 2026, so the more useful question is what comes after BC. The named candidate is energy storage, added through the January 2026 acquisition of PotisEdge, but the report is explicit that this is still a strategic option, not an earnings pillar, with no primary disclosure showing it as revenue-relevant yet. Hydrogen equipment is mentioned as a third leg but is called financially peripheral. So a second curve exists inside the company already, which is better than having no candidate at all, but it is unproven, and the report's own tracking dashboard treats no visible storage revenue relevance by the end of 2027 as a failure trigger, putting management on a clock to show it is real.

The report calls the moat mixed and names three real components: process and product know-how spanning monocrystalline history into BC industrialization, bankability and customer trust built through utility-scale delivery in more than 150 countries, and a relatively resilient balance sheet that lets the company keep investing through the downturn. It is explicit about what no longer counts: wafers alone and scale alone, both described as fair-weather advantages in the current oversupplied market. On direction, the old moat is narrowing (wafer gross margin was negative 5.30% in 2025), while the new one, BC differentiation, could widen if the pricing premium proves durable, but the report's own margin-of-safety section calls premium retention the most fragile assumption in its base case. The honest read is unresolved: the moat has to be re-earned on the new axis rather than being assumed to be widening.

There is real evidence of a self-reinvention gene. LONGi's original identity, a monocrystalline wafer champion, was structurally devalued by the 2022-2025 industry reset: revenue fell from CNY 129.50bn in 2023 to CNY 70.35bn in 2025, and profit swung from a CNY 10.78bn gain to a CNY 6.42bn loss over the same stretch. Rather than defending that identity, management redirected capital and attention toward BC modules, a pivot disclosed in financing documents as a plan to reach 100GW of BC capacity within three years on more than CNY 25bn of investment, and it has since layered a second pivot into storage on top of that. That is a company willing to cut loose a legacy identity rather than protect it. What the report does not show is how management talks about mistakes or bad news directly; there is no discussion of an admitted misstep or a public reckoning, so that half of the question is an evidence gap rather than a confirmed strength.

Founder Li Zhenguo, trained in semiconductor materials, has led the company's technical and strategic direction since its founding in 2000, and today's leadership, with Zhong Baoshen as chairman, still traces back to that founding generation, which points to continuity rather than a hired-manager rotation. The willingness to sacrifice near-term profit is visible in the numbers: management kept funding the BC transition and then added the storage acquisition while the company was posting real losses, CNY 8.59bn in 2024 and CNY 6.42bn in 2025, rather than retreating to protect short-term earnings. What the report does not disclose is management's personal shareholding or how compensation is tied to long-run outcomes, so the depth of personal financial alignment cannot be verified from this report alone.

Existing utility customers would likely miss LONGi operationally: warranty obligations, service networks, and multi-year project relationships across more than 150 countries create real switching friction mid-project, and that bankability is one of the three things the report credits as a genuine moat. But at an industry level, LONGi is not irreplaceable. The peer table shows JA Solar, Jinko, Trina, and TCL Zhonghuan all supplying comparable or larger volumes, so its disappearance would be a supply disruption rather than the loss of an indispensable partner for the sector as a whole. On sustainability, nothing in the report ties LONGi's growth to regulatory arbitrage or social harm; if anything, the policy backdrop, China's campaign against disorderly low-price competition, is aimed at curbing the destructive pricing that hurts the whole industry's economics, which would favor a well-capitalized survivor like LONGi rather than penalize it.

The segment margins answer this directly and the answer is that scale has not been helping lately. Wafers ran a negative 5.30% gross margin in 2025, and modules and cells, despite 82.19GW of volume, were barely positive at 0.19%; only the small power-station segment held a real margin, at 22.08%. In other words, getting bigger in the core business has not been producing better unit economics through this cycle, which is the central bear argument. On where the money goes, contracted fixed-asset commitments were only about CNY 1.55bn at the end of 2025, far below the levels seen during the earlier expansion era, so capex has shifted from aggressive growth spending to something closer to selective maintenance. The company is holding cash, CNY 53.35bn against CNY 35.20bn of debt, rather than redeploying it into new capacity, which looks like balance-sheet preservation through a trough rather than compounding on incremental investment.

The report does not model a ten-year path, but its own three-year scenarios are a useful proxy, and they fall well short of what a five-times outcome would need. Even the optimistic case implies a fair value of about CNY 17.5 against a current price of CNY 13.12, roughly 33% upside over three years, not the near five-times move a ten-year quintuple would require from today's price. Getting there would need the BC premium to prove durable and scale for a full decade, storage to turn into an actual second profit engine rather than a strategic option, the industry-wide price war to resolve into real pricing discipline, and the market to eventually value LONGi as a quality platform instead of a cyclical manufacturer, none of which the report treats as its central case. At about 1.8 times book, the current price already reflects some credit for the transition without pricing in that kind of outcome.

The report's own view is not that the market is blind, but that it is genuinely split, and reasonably so. It says the market is probably underrating how decisively the revenue mix has already shifted away from wafers, which is a look-too-far-ahead failure since that shift is already visible in the 2025 annual report rather than a future promise, while some bulls are probably overrating how quickly BC's product edge converts into stable margin recovery in a chain that is still badly oversupplied. That is closer to a live, unresolved disagreement than a mispricing waiting to be corrected. The concrete trigger the report points to for a narrative shift is two or three consecutive quarters where BC mix above 65% shows up as real module-margin improvement, narrowing losses, and disciplined inventory; the mirror-image trigger, further margin erosion or a fading BC premium, would confirm the bear case instead.