Hangzhou Lion Microelectronics Co., Ltd.

It earns by selling physical wafers and chips outright to other businesses under one-off orders or standing framework agreements, not through subscriptions or recurring fees. Income is only booked once goods are accepted by the buyer, consumed in the buyer's own production, or shipped abroad, tying what it earns directly to volume shipped rather than to an installed base of ongoing customers.

It grows by adding physical manufacturing capacity through large, multi-year construction projects rather than through incremental, low-cost expansion, and it has said it slows those projects when existing capacity is running under-used. Its recent earnings history includes at least one loss year rather than a straight line of growing profit, and it shares this capacity-driven way of operating with many other companies in the same line of work.

It depends on outside suppliers for the industrial gases and other raw materials that feed its wafer and chip lines, on keeping its specialized technical staff, and on keeping pace with fast-moving compound-semiconductor and wafer technology. It names only one supplier directly, a gas provider that makes up a small share of its total purchasing, which points toward a spread-out supplier base rather than reliance on one dominant source, at least among what it discloses. It also sits below many more supplying industries in the wider economy than the small number it in turn supplies.

Its own filings show foundries and chip manufacturers, automotive suppliers and carmakers, and lidar and sensor makers among its buyers, naming companies such as SMIC, Bosch, BYD and RoboSense. It feeds into far fewer industries downstream than the number of industries that supply it, so a comparatively narrow set of buyer types rely on what it makes.

The company points to running the full process from raw crystal to finished chip itself, sharing research and customer relationships across its wafer, power-device and radio-frequency product lines, and holding backup manufacturing capacity across several of its own sites as what sets it apart. Whether other companies could copy this combination is not something the available evidence can measure. The broader capacity-based way of converting materials into finished product is shared with many other companies, so that wider shape on its own is not unusual.

Its own filings describe its newest wafer-expansion project as limited mainly by demand: construction and equipment buying there were slowed because existing capacity was already sitting under-used during an industry downturn, rather than because of a physical capacity ceiling or an approval delay. Separately, it names the availability of skilled staff, the pace of new technology and the cost of raw materials as limits on its operations and earnings.

By its own account, what could most affect it is a downturn in semiconductor demand, falling behind on fast-moving chip and wafer technology, losing ground to competitors, being unable to keep enough skilled technical staff, and swings in raw-material costs. It also points to inventory that can lose value, investments whose value can move against it, and goodwill it could have to write down. It does not itself point to reliance on any single customer or single region as a risk.

By its own account, the pressures it puts first are swings in industry-wide demand, the pace of new chip and wafer technology, competition, the availability of skilled staff and the cost of raw materials, in that order of emphasis, ahead of its financial-risk disclosures. It names international competitors including Shin-Etsu, TSMC and Infineon in its own discussion of the field it operates in. It answers to Chinese securities and stock-exchange rules, and although most of its business is settled in its home currency, the foreign-currency assets and liabilities it does hold sit mostly in US dollars, exposing it to currency movements outside its control.