Teradyne looks misclassified after a 13.6% one-day drop. The market is treating TER like another semi name caught in a sector downdraft, while the company itself says about 70% of revenue is now tied to AI-related demand. That is not a small narrative tweak; it is a revenue-mix shift that changes how the stock should be framed. We think the selloff is giving investors a cyclical-stock price reaction to what is increasingly an AI infrastructure test business.
The cleanest proof is in the latest quarter. Q1 revenue reached $1.282 billion, up 87% year over year, and Semiconductor Test alone contributed $1.111 billion, or roughly 87% of total revenue. Management did not leave much room for interpretation either, explicitly tying about 70% of company revenue to AI-related demand and framing the business around a wafer-to-AI-data-center strategy. If most of the revenue base is being pulled by AI compute and memory test intensity, TER should not be lumped in with slower, more consumer-exposed semiconductor names.
The quality of that growth is what makes the setup compelling. TER carries a 58.8% gross margin, a 26.9% operating margin, and a 22.6% net margin, with ROIC at 25.7%. That is why the TickerSpark Score lands at 70 overall despite a merely middling 40 on Valuation: Profitability is a perfect 100 and Financial Health is 88. Put differently, this is not a speculative AI story living on hope. It is a highly profitable test franchise with real earnings power, and the market just marked it down as if the fundamentals had cracked.
The relative setup also argues for staying constructive. TER is still up 77.8% year to date, beating the Technology sector by 52.7 percentage points, and it has beaten earnings in seven straight quarters, including a 20.8% surprise in April and a 30.4% surprise in February. Consensus still leans bullish with 20 buys against 11 holds and no sells, while recent news sentiment has stayed strongly positive. Against KEYS, the contrast is notable: Keysight grows more slowly at 8.0% revenue growth and posts a lower 17.2% net margin, yet TER is the name getting sold as if its growth engine were less durable.
The obvious pushback is valuation. At 67.97 times trailing earnings and 15.26 times sales, TER is not cheap, and the TickerSpark Score's 40 Valuation sub-score says exactly that. Q2 guidance of $1.15 billion to $1.25 billion also implies some moderation from the record Q1 level, so anyone buying here is paying up for an AI-linked demand cycle that the market already knows about.
There are also some near-term tells that keep this from being a no-risk momentum layup. The stock is below its 20-day and 50-day moving averages, insiders logged $2.28 million of recent selling, and today's drop shows how violently sentiment can reset when semis get de-risked. Even so, those are trading and valuation objections, not a fundamental rebuttal to the core point that TER's revenue mix is now far more AI-driven than the market's cyclical bucket suggests.
That leaves TER looking like a buy-the-misclassification name, not a stock to chase blindly but one we would rather own than KEYS if choosing between test exposure. The business mix, margin profile, and earnings consistency all say TER deserves to trade as an AI infrastructure enabler first and a generic cyclical semi second.
What we'd watch now is simple: the next quarter needs to keep validating the AI revenue concentration and defend margins near current levels. If that happens, today's 13.6% washout will look more like a reset in positioning than a break in the story. If AI-linked demand fades enough to pull Semiconductor Test off its current trajectory, that would change the setup fast. Until then, the market's classification error still looks bigger than the business risk.
