Megmeet (麦格米特), a Shenzhen-based power-electronics and automation supplier, has become a market darling even as its underlying profitability deteriorates. On January 28 the company warned that 2025 net profit attributable to shareholders will fall to ¥120–150 million, a drop of roughly 66–72% year-on-year, while non-recurring-adjusted net profit is set to plunge to just ¥20–30 million, down more than 90%.
The headline numbers mask a sharper story: Megmeet generated ¥212 million of attributable net profit and ¥96 million of non-GAAP profit in the first three quarters of 2025, and the firm expects the fourth quarter to swing into loss, taking the full-year totals to the guided low levels. Management attributes the deterioration to two structural pressures — falling gross margins as it faces intensified cost pressure in legacy lines while acting largely as a component supplier, and sharply higher R&D and administrative spending as it pivots into AI server power supplies.
Investors have nonetheless poured money into the stock on the back of Megmeet’s AI ambitions. Since October 2024 the share price has surged some 360%, lifting market capitalisation to about ¥70.9 billion. Market enthusiasm was fuelled further by Megmeet’s appearance among more than 40 component suppliers listed on Nvidia’s website for a GB200 NVL72 rack project — an endorsement that the market treats as proof of entry into the Nvidia ecosystem and a potential path to rapid AI-server revenue growth.
Brokers and fund allocations reflect that optimism: a December equity private placement priced at ¥85.01 per share raised ¥2.663 billion and attracted major asset managers, including E Fund, UBS and other institutional investors. Research houses paint a binary future for the company: some forecasts assume AI power-supply revenue jumps to ¥40–70 billion by 2026–27 with gross margins near 40%, while sceptical commentators warn current valuations imply market-share assumptions that outpace conservative industry estimates.
The tension — extreme market valuation versus near-term financial pain — encapsulates wider dynamics in the AI supply chain. Server power supplies (PSUs), high-voltage DC transmission units and cabinet-level designs are high-stakes, certification-heavy products where early design wins and ecosystem access (notably to major OEMs and Nvidia’s platform) matter more than niche cost reductions. Megmeet’s heavy R&D and capacity investments are intended to convert positioning into share, but they also make the company vulnerable to certification delays, customer concentration, and margin squeezes if AI demand or market share falls short of the rosy scenarios.
For global investors, the Megmeet episode is a cautionary example of how AI narratives can detach stock prices from current cash flows. If the company executes — shipping Rubin rack PSUs, CBU supercapacitor systems and HVDC modules at scale while defending gross margins — the market’s optimism could be vindicated and the short-term losses prove to be necessary phase costs. If not, the stock could suffer a sharp re-rating once forecasts colliding with reality force investors to reconcile lofty expectations with fragile fundamentals.