Teradyne CEO sells EUR 1.17m while AI test demand stays hot

AI test gear is still riding the AI capex wave

AI spending is still doing the heavy lifting for semiconductor equipment names, and Teradyne sits right in that current. The company sells automated test equipment into a market that has been supported by high-volume testing needs for advanced logic, memory, and packaging, the kind of demand that has kept the group resilient even as some back-end segments cool. That matters because this is not a sleepy industrial tape story. It is a capital-spending story with a very specific bottleneck, and the bottleneck is testing. When chip makers and their supply chains keep pushing more complex devices through the line, the test step becomes less of a back-office afterthought and more of a gating function for throughput, yield, and time to revenue.

The broader sector backdrop still favors that logic. AI infrastructure spending remains the dominant macro force, and the equipment names tied to it have continued to benefit from order activity even as parts of the semiconductor cycle soften. That split is important. It means the market is not simply rewarding all chip equipment equally. It is rewarding the names that can show direct exposure to AI-related capital intensity, especially where advanced logic, memory, and packaging all need more sophisticated test coverage. In that sense, Teradyne is not just participating in a sector move. It is positioned inside one of the tighter bottlenecks created by the AI buildout.

Advantest is the obvious peer to watch here. It has been stronger in some memory test segments and has talked up AI demand with a raised operating profit forecast, while Teradyne has still put up a powerful share-price run of its own. Year to date through mid-September, Advantest was up about 56 percent and Teradyne about 76 percent, according to the comparison data in the research. Cohu has lagged. So the sector is not short of momentum, but it is also not moving in lockstep. In a group like this, the market keeps rewarding the names that can show exposure to AI test content without sounding like they are simply borrowing the theme.

That distinction matters because the test market is not a generic semiconductor proxy. It is a place where product mix, insertion point, and customer demand can change the story quickly. A company can be in the right sector and still underperform if it is not capturing the right part of the cycle. Teradyne’s recent strength suggests the market believes it is capturing enough of the AI test wave to matter, but the peer comparison also shows that investors are still discriminating. Advantest’s strength in memory test and its raised outlook are a reminder that the market is rewarding specific execution, not just broad exposure.

The stock has already done the hard part

Teradyne’s July quarter gave the market a reason to keep paying attention. Revenue reached $1.33 billion, up more than 100 percent year over year, and non-GAAP EPS came in at $2.47, according to the company’s earnings transcript. Management guided third-quarter revenue to between $1.2 billion and $1.3 billion and said it expected ATE market expansion and share gains to support growth into 2027. That is a sturdy setup, and the market has treated it that way. The numbers matter not just because they were strong, but because they validated the idea that AI demand is not a vague future opportunity. It is already showing up in the company’s reported results and forward commentary.

The share price tells the same story. A stock that closes at $332.85 and then trades at $341.15 the next day is not being priced like a company under stress. It is being priced like a name with a real AI-linked earnings bridge and enough credibility to keep the multiple elevated. That is the context you need before you look at the filings. Selling into a flat chart is one thing. Selling after a run like this is another. The market has already done a lot of the work for the bulls, which means any insider sale now lands against a backdrop of strength rather than weakness.

That backdrop also helps explain why the market is willing to tolerate a premium. Teradyne is not being valued as a cyclical test vendor waiting for a rebound. It is being treated as a company with a visible path from AI demand to revenue growth, and management has reinforced that path by pointing to market expansion and share gains. When a company can connect current results to a longer runway, investors tend to look through near-term noise. That is exactly why insider selling in a strong tape can be read in more than one way. It can be routine monetization, but it can also be a signal that insiders see the easy part of the rerating as already behind them.

September 15: the CEO and a director both sold

Gregory Stephen Smith, Teradyne’s president and CEO, sold 4,000 shares on September 15 at about $337.83 each, for roughly EUR 1.17m in euro-normalised filing value. He retained about 108,495 shares afterward. Marilyn Matz, a director, sold 800 shares at the same price for about EUR 234,157 and was left with roughly 14,241 shares. Both filings were reported the following day and both were tied to pre-arranged Rule 10b5-1 plans.

That is the part that matters. The sales were not random, and they were not hidden. They were scheduled. But scheduled selling still lands in the tape, and when the chief executive is the seller, the market notices. Teradyne’s insider pattern has also included multiple 4,000-share sales by Smith earlier in 2026 at higher prices, so this is not a one-off print that arrived out of nowhere. It is part of a repeated cadence of disposals while the stock has been strong. The repetition is what gives the filing more weight than a single isolated transaction would have carried.

The timing is also important because it lines up with the stock’s own trajectory. The sales came after a strong quarter, after a powerful share-price move, and while the sector backdrop still favored AI-linked equipment names. That combination does not automatically make the sales bearish. Executives often diversify after large gains, and 10b5-1 plans are designed to remove discretion from the timing. But the market still reads the pattern, especially when the same names keep appearing on the sell side. In a stock that has already rerated, insider selling can be less about immediate alarm and more about the market learning where the insiders themselves are comfortable taking some exposure off the table.

What the cluster shows, and what it does not

Insiders Trades data flags this as a cluster, and the cluster is real. The dossier shows 3 distinct insiders and 12 recent declarations, with Smith and Matz both selling on September 15 and September 16, and both also active on August 18. That is enough to say the selling is not isolated. It is a pattern of repeated disposal by the same names over a short window. The fact that the same two insiders recur across several dates makes the signal more coherent than a one-day burst of unrelated activity.

At the same time, cluster data needs to be handled carefully. A cluster can mean shared conviction, but it can also mean shared timing around a pre-arranged plan or a routine liquidity window. Here, the filings were under Rule 10b5-1 plans, which lowers the chance that the trades were a reaction to a specific near-term event. That does not erase the signal, but it does change how you interpret it. The cluster is telling you that multiple insiders are monetizing into strength, not that they are making a fresh directional call on the business.

Our scoring leans on the chief executive role, the cluster, and the fact that the filing value is tiny relative to Teradyne’s market value, under 0.01 percent. That last point matters because it keeps the filing in scale. EUR 1.17m is real money, but it is not a balance-sheet event for a company with a market cap of about EUR 48.4bn. The trade is about behavior, not corporate finance. You should read it that way. It is a window into how insiders are managing their own exposure, not a sign that the company itself is changing course.

The historical cohort data is less dramatic than the stock chart. For chief-executive buys at mega-cap names, the 90-day sample shows a 46.7 percent win rate and an average return of -1.17 percent, with a 365-day average return of 57.28 percent across 2,045 observations. That is historical cohort data for that role-and-size bucket, not a forecast for Teradyne and not a promise that this filing leads anywhere in particular. It simply tells you that the bucket is noisy over shorter horizons and can work very differently over longer ones. The sample is large enough to be useful as context, but not so clean that it should be mistaken for a trading rule.

Teradyne’s own numbers still justify attention

The reason this filing gets attention at all is that Teradyne is not a broken story. The company’s second-quarter numbers showed AI demand doing what management said it would do, and the guidance pointed to continued strength into the next quarter. That gives the stock a real fundamental anchor. The market is not buying a rumor here. It is buying a business that has already printed the revenue. In a sector where many names are still trading on the promise of future AI content, that matters. Teradyne has already demonstrated that the demand is flowing through the income statement.

Insiders Trades data gives the company a fundamental score of 63, with a quality score of 84 and a value score of 41. Those are not a thesis by themselves, and they are not meant to be. They are a screen. Still, they fit the picture. Teradyne is not cheap in the way a cyclical trough name is cheap, and it is not low quality in the way a story stock can be. It sits in the awkward middle where growth is real, expectations are high, and the stock can still move a lot on any change in tone. That middle ground is often where insider sales matter most, because the market is already debating how much of the good news is in the price.

The company’s quality score is especially relevant in this context because it suggests the market is not simply rewarding a one-quarter spike. It is rewarding a business with enough operational credibility to support a higher multiple. But a higher multiple also means less room for disappointment. If the AI test cycle remains strong, the stock can justify the premium. If the cycle slows, the same premium can compress quickly. That is why the insider sales are worth noting even though they are small relative to market cap. They arrive at a point where the stock is already priced for a lot of success.

That is also why the sector comparison matters. Advantest has been the cleaner benchmark for some of the AI test enthusiasm, while Teradyne has still delivered a stronger year-to-date share move. Cohu has not kept up. If you are trying to decide whether the insider sales are a warning sign or just a monetization event after a strong run, the answer depends on whether you think the AI test cycle is still early or already well owned. The filings do not settle that argument. They just tell you the CEO and a director chose to take some chips off the table while the market was still paying up.

The filing sits inside a broader 2026 selling pattern

The September 15 sales were not the first time Smith showed up on the sell side this year. The research notes multiple 4,000-share transactions earlier in 2026 at higher prices, and the cluster data shows both Smith and Matz active again in August before the September prints. That is a useful distinction. A single sale can be noise. Repeated sales by the same executive, in the same stock, across several months, are less easy to write off as a one-day housekeeping event. The pattern suggests a deliberate and ongoing approach to trimming exposure rather than a one-off response to a particular price.

Still, you should not overread the size. Smith’s sale was about EUR 1.17m, and his remaining stake after the trade was about 108,495 shares. Matz’s sale was about EUR 234,157, with about 14,241 shares left. Those are meaningful holdings. They are not exits. The executives are still exposed to the stock. That is the part the market often misses when it sees a headline about insider selling and jumps straight to a bearish conclusion. Retained ownership matters because it keeps the insiders aligned with the company’s future, even if they are reducing some exposure after a strong run.

The better question is whether the timing fits the chart and the business. Here, it does. The stock had already run, the company had already posted a very strong quarter, and the sector backdrop was still favorable. Selling into that kind of setup is rational even when the underlying business remains strong. It can also be a sign that insiders see less upside from here than the market does. Both readings can be true at once. That ambiguity is exactly why insider filings are most useful when they are read alongside fundamentals and sector context rather than in isolation.

There is also a subtle point about market psychology. When a stock has already moved sharply on a credible growth story, the market becomes more sensitive to any sign that the people closest to the business are reducing exposure. Even if the trades are pre-planned, they can still influence sentiment because they arrive at a moment when expectations are elevated. That does not mean the stock is about to roll over. It means the burden of proof shifts back to the company to keep delivering numbers that justify the move.

What to watch after the September 16 print

The next thing to watch is not another filing headline by itself. It is whether Teradyne can keep converting AI demand into the kind of revenue and margin profile that justifies the current valuation. The July quarter showed the company can still surprise on the upside, and management’s third-quarter guide left room for continued strength. If that holds, the stock can absorb insider selling. If growth cools, the same filings will look more pointed in hindsight. The market will care less about the fact that the trades were pre-arranged and more about whether the business kept compounding after the insiders trimmed.

Watch the peer tape too. Advantest remains the cleaner comparison for AI test momentum, especially where memory test strength is concerned. If Advantest keeps outperforming in the segments that matter most, Teradyne’s relative multiple can get harder to defend. If Teradyne keeps showing share gains and the AI test cycle broadens, the market will probably treat these sales as routine monetization rather than a warning flare. The relative performance of the peers is important because it helps separate sector enthusiasm from company-specific execution.

The other thing to watch is whether the cluster continues. A single month of repeated sales can be explained by plan-based execution. A longer run of similar filings would make the pattern more meaningful. For now, the filing tells you the CEO and a director sold into strength, under pre-arranged plans, while the stock was already near recent highs and the sector backdrop still favored AI-linked equipment names. That is a concrete fact pattern, and it is enough to keep the name on the list for the next earnings update and the next cluster of Form 4s.

Dig deeper: Teradyne, INC's full insider filing history.

This is not investment advice.