On July 16, 2026, Taiwan Semiconductor (TSMC) reported Q2 revenue of $40.2 billion (+36% YoY) and net income up 77.4% year-over-year — yet its stock fell over 2% after the company raised its full-year capital expenditure guidance to $60–$64 billion, up sharply from a prior range of $52–$56 billion.
Stellar earnings, stock drops. That's confusing until you sit with how a CFO reads a capex announcement.
TSMC just posted a 67.7% gross margin and 55.6% net profit margin — numbers most companies would frame on a wall. But the market flinched at one line: capex guidance raised to $60–$64 billion for the year. Capex (capital expenditure) is money spent building fabs, buying equipment, scaling 2nm chip production — cash out the door today for revenue that arrives years later.
Here is how a CFO reads this: every dollar of that $64B has to clear a hurdle. The tool is NPV and IRR. NPV (Net Present Value) asks: if I discount all the future cash flows this fab will generate back to today, is the total worth more than what I'm spending? IRR (Internal Rate of Return) asks: what's the effective annual return on this investment? If TSMC's cost of capital is, say, 10%, every project inside that $64B must earn above 10% to create shareholder value.
The market's nervousness isn't about TSMC's past — it's about whether $64B in future commitments clears that hurdle, especially with AI demand that could cool, geopolitical risk in Taiwan, and 2nm ramp costs already pressuring next-quarter margin guidance to 65–67%.
The CFO question isn't 'did we earn well last quarter?' It's 'does this investment earn its keep over the next decade?' That's NPV/IRR in one sentence.
📚 Learn the concept: NPV & IRR
Source: https://www.sec.gov/Archives/edgar/data/0001046179/000104617926000451/a2q26e_withguidancexfinal.htm
▶ Play the 90-second CFO game All daily posts