Wistron Raises $1.47 Billion to Fund the AI Server Supply Chain

The Nvidia supplier priced 25 million depositary receipts at a 5.5% discount as raw-material needs rise alongside AI-server demand.

By Kenji Mori • • Markets

Robotic assembly lines build unbranded server racks as a blue capital stream crosses stacks of industrial raw materials.

Taiwanese electronics manufacturer Wistron has raised $1.47 billion through a global depositary-receipt sale, giving the Nvidia supplier fresh working capital as demand for AI servers stretches the hardware supply chain. The company priced 25 million receipts at $58.88 each, near the bottom of the marketed range.

Each receipt represents ten Wistron common shares, equivalent to about T$186.24 per share. That was roughly a 5.5% discount to the company’s closing price on September 7. The new shares dilute existing ownership by approximately 7.29%, a meaningful transfer of future earnings even though Wistron said it did not expect a material weakening of shareholder interests.

Wistron plans to use the proceeds to buy raw materials denominated in foreign currencies. That detail distinguishes the financing from a conventional factory expansion. AI-server manufacturers must secure memory, networking equipment, processors, power components and other inputs before customers pay for completed systems. Rapid sales growth can therefore consume cash even when profitability improves.

The company has already expanded capacity. It opened a $700 million Texas plant in July to manufacture Nvidia’s latest AI systems and approved additional spending in Taiwan. Management said in August that demand from cloud-computing and enterprise customers continued to exceed supply.

Second-quarter figures support that account. Net profit increased 128% from a year earlier and revenue rose 64%. Wistron’s shares have gained about 30% in 2026, giving it a stronger valuation base from which to issue equity. The offer discount reflects the need to attract a large pool of international capital without placing too much immediate pressure on the local share price.

The transaction also illustrates the financing burden moving through the AI ecosystem. Cloud operators may announce the largest capital budgets, but equipment suppliers must finance inventory, production and receivables. Equity reduces balance-sheet pressure compared with debt, though it permanently spreads ownership across more shares.

Foreign-currency funding creates its own risks. Wistron is matching dollar-based receipts with imported inputs and overseas operations, which can reduce currency mismatches. However, movements in the Taiwan dollar, component prices and customer payment terms still affect how much purchasing power the proceeds ultimately provide.

Customer concentration is another issue. Demand tied to Nvidia and a limited number of cloud providers can generate exceptional growth while leaving suppliers exposed to design changes, delayed data-centre projects or shifts in sourcing. A shortage can also become excess inventory if customers revise orders after components have been purchased.

Why it matters

The sale turns AI optimism into a large and measurable capital requirement. Wistron is not raising money for a distant concept; it is financing the materials needed to fulfil current server demand. That makes the transaction a useful indicator of how much working capital the AI build-out is absorbing beyond the best-known technology companies.

For shareholders, the trade-off is immediate dilution in exchange for capacity to capture a fast-growing market. For suppliers, the proceeds can support larger orders. For cloud customers, a better-funded manufacturer may reduce bottlenecks, although the capital raise does not guarantee that scarce components will become available.

The next evidence will come from inventory, receivables, margins and utilisation at the Texas and Taiwanese facilities. Strong revenue alone is not enough: investors need to see that the $1.47 billion converts into profitable deliveries rather than expensive stock held against volatile forecasts.

Sources