TSMC plans to spend $64 billion on plants and equipment, and that figure, according to Invezz, should push investors toward the stock rather than away from it. The number looks alarming at first glance. It is one of the largest capital budgets any company in any industry has ever committed.

The worry is simple. Heavy spending means cash leaves the balance sheet today, while the factories it buys earn revenue only years from now. Investors have punished other chipmakers for exactly this pattern. The argument from Invezz, however, runs in the opposite direction.

Why spend $64 billion at all?

TSMC makes chips for customers who cannot build their own at scale, including the largest designers of smartphone, data-center and AI processors. Those customers keep asking for more wafers than the company can currently supply. Capital spending is the mechanism that closes that gap.

When demand outruns capacity, a foundry has two choices. It can ration wafers and watch customers look elsewhere, or it can build. TSMC has chosen to build, and the size of the budget reflects the size of the order book behind it. Invezz frames the spending as evidence of demand strength rather than as a cost problem.

What does the spending mean for investors?

The core of the Invezz case is that capital spending at TSMC is not money lost. It is money converted into a physical asset that the market currently struggles to replicate. Advanced chip manufacturing has few credible competitors at the leading edge, and the equipment, cleanrooms and process knowledge behind it take years to assemble.

That conversion matters for returns. A dollar spent on a fab becomes a dollar of capacity that can price against scarce supply. As long as customers need leading-edge wafers and have limited alternatives, the spending supports future revenue rather than diluting it.

Invezz also points to the risk of under-investment. A foundry that spends too little can find itself short of capacity when demand accelerates, losing allocations and design wins to rivals that built ahead of the cycle. In that reading, the $64 billion plan is insurance against exactly the outcome investors fear most.

Are there real risks?

The spending plan is not risk-free. Chip demand moves in cycles, and a downturn can leave expensive capacity idle. Construction costs, equipment lead times and geopolitical exposure around Taiwan all press on the economics of each new site. Any of these can stretch the time between spending and returns.

The counterweight is TSMC's position in the supply chain. Its customers depend on it for the most advanced logic available, and switching foundries is slow and expensive for them. That stickiness, Invezz argues, gives the company a path to recover its investment even if the cycle turns.

What happens next?

The plan will unfold over several years, with the money flowing into plants, tools and process capacity. Investors will watch whether customer demand stays strong enough to fill the new wafers as they come online. If Invezz's read holds, the $64 billion outlay will look in hindsight like a floor under TSMC's future earnings rather than a hole in them.