Orders Jump Nearly 70% While Revenue Stalls, INTT Bets on Margins

On October 5, inTest released preliminary third-quarter operating results ahead of its full earnings report, which is scheduled for November 6. Revenue simply hit the top of guidance. The real news was in the orders line.

The headline figures:

The preview contained no profit figures. Gross margin and expenses arrive with the full November report. To see what this jump in orders is built on, it helps to lay out the last four quarters first:

Those numbers hide an easy-to-miss mismatch. Revenue rose for four straight quarters, but orders slid steadily from $37.5 million in Q4 last year to $28.9 million in Q2, and backlog was worked down from $53.9 million to $45.4 million. In the first half, the company was essentially living off past orders. The Q2 revenue mix makes the point even more bluntly: Auto/EV (automotive and electric vehicles) contributed $13.4 million of revenue, or 38.1%, while new orders from that market in the same quarter were only $3.5 million (Q2 earnings release, 2026-08-10). Car-industry projects carried the first half, and almost nothing came in to replace them.

That is why the Q3 order jump matters. The company has changed engines, from fading Auto/EV projects to back-end semiconductor testing. Semiconductor orders in Q2 were already up about 56% from the prior quarter and about 64% from a year earlier, the strongest in six quarters (Q2 earnings release, 2026-08-10). Q3 then set a record for back-end semiconductor orders.

The full-year math is not demanding. First-half revenue was $69.2 million. Add Q3 at $35 million and the nine-month total is about $104.2 million:

Revenue isn't the problem. Margins are. Management guides to about 43% gross margin for the full year. The first two quarters came in at 43.3% and 40.5%, so the second half needs roughly 44% to get there, the same as the roughly 44% Q3 guidance. The thing to watch in the November 6 report is whether Q3 gross margin gets back to 44%. Gross margin fell 280 basis points in Q2, which the company blamed on a bigger share of lower-margin Auto/EV revenue. If margin doesn't recover now that semiconductor orders are back, the case that profits will grow faster than sales gets weaker.

On demand, management offered two direct quotes in the preview: "With orders for semiconductor and other end-markets booked in prior quarters now shipping, we expect revenue for both the third quarter of 2026 and the full year to be at the high end of our guidance ranges." And: "Demand for our advanced semiconductor test solutions is being driven by customers capacity expansions that are fueled in part by AI-driven data center demand." (8-K, 2026-10-05). The company also singled out two new product lines as significant contributors: high-power chillers (cooling systems that control temperature during power-chip testing) and next-generation manipulators (robotic arms that precisely dock a test head to a wafer prober or chip handler). These serve customers' growing need to test power-conversion chips and mixed-signal and analog chips. Note the words "in part" in that quote: AI is one driver, and the company itself does not claim it is the whole story.

inTest is an asset-light business. Its full-year capital spending guidance is just 1% to 2% of revenue (Q2 earnings release, 2026-08-10). So the focus here is earnings quality: how much of the reported profit is real, and how much of it actually turns into cash.

Start with the gap between GAAP and adjusted figures. Q2 GAAP net income was $474,000 and adjusted net income was $1.091 million. The difference is mostly $699,000 of amortization of intangible assets (Q2 earnings release, 2026-08-10). That amortization stems from past acquisitions, such as the Alfamation business line, which brought in $11.7 million of revenue in Q2 alone (10-Q, 2026-08-10). It is money the company really did spend, now being written down gradually on the books. Stripping it out is fine for judging the operating trend, but you shouldn't pretend it doesn't exist.

Adjusted EBITDA deserves a closer look. Of the $2.2 million reported for Q2, $843,000, or about 38%, was stock-based compensation (SBC) added back (Q2 earnings release, 2026-08-10). SBC is about 2.4% of revenue, which doesn't sound like much on its own. But against an EBITDA margin of just over 6%, it is the biggest piece. Put another way, if you treat the stock handed to employees as part of their pay, the company's Q2 profit on a cash basis is much thinner than the adjusted number suggests.

Existing shareholders have already been diluted. Shares outstanding rose from 12.49 million at the end of 2025 (Q4 earnings release, 2026-02-27) to 12.826 million at June 30, 2026 (10-Q, 2026-08-10), about 2.7% dilution in six months. Most of it came from option exercises: 330,000 shares were exercised in the first half, bringing in about $3.41 million (10-Q, 2026-08-10). After a big run in the share price, it makes sense that employees would cash in options, and the company took in cash as a result. But for public shareholders, the share count that earnings get divided by really has grown. The 10-Q showed no at-the-market share sales program and no new shelf registration.

Cash conversion is better than some accounts suggest. One version circulating in the market puts first-half free cash flow in negative territory. The actual disclosure: first-half operating cash flow of $3.004 million, capital spending of $1.049 million, and free cash flow of about $1.96 million (10-Q, 2026-08-10). By quarter, Q1 operating cash flow was about negative $3.3 million and Q2 swung back to positive $6.3 million (Q2 earnings release, 2026-08-10). The swing came mainly from working capital. Inventory fell from $31.58 million to $27.92 million over the half, freeing up about $3.66 million (10-Q, 2026-08-10).

The trap is that cash released by shrinking inventory only happens once. With Q3 orders jumping to the $50 million range and backlog up to the $60 million range, the company will most likely need to restock in the second half, and working capital will eat cash instead of freeing it. Fortunately, the balance sheet can absorb that. At the end of June the company had $22.1 million in cash and about $6.2 million of total debt, for net cash of about $15.9 million. It also has an undrawn $10 million revolving credit line and a $30 million delayed-draw term loan facility (Q2 earnings release, 2026-08-10). Customer deposits and deferred revenue together come to about $5.88 million (10-Q, 2026-08-10), meaning some orders arrive with money up front. It's a small amount, but a good sign.

Bottom line: the company is under no funding pressure and won't need to come to shareholders to pay for materials or expansion. Its problem lies on the income statement. GAAP profit is so thin that a swing in any one of amortization, SBC or working capital can decide whether a quarter looks good or bad.

On the supply side, inTest makes manipulators, docking hardware, test interfaces, temperature-control equipment and induction heating systems. Its inputs are standard industrial materials: precision-machined parts, compressors and refrigeration components, and electronic components. We could not find a list of key suppliers or any single-source dependencies in public filings. Given the type of products, the real bottleneck is more likely engineering and assembly staff. Custom manipulators and temperature-control setups need on-site tuning and application support. With orders jumping nearly 70% in one quarter, whether delivery can keep pace matters more than whether parts are available.

On the customer side, the mix has changed in six months. In full-year 2025, nearly 80% of revenue came from non-semiconductor markets, with semiconductors at just 31.6% (Q4 earnings release, 2026-02-27). In Q2, semiconductor revenue was $9.1 million, or 25.7%, Auto/EV was 38.1%, and defense/aerospace, industrial and life sciences together made up about 30% (Q2 earnings release, 2026-08-10). By Q3, back-end semiconductor had taken the lead in orders. The buyers are IDMs (companies that both design and manufacture their own chips), contract chip foundries, packaging-and-test houses and test service providers. Geographically, $23.9 million of Q2 revenue, about 68%, came from outside the US (10-Q, 2026-08-10).

Customer concentration is the weak spot in what we could verify. The 10-Q contains concentration information by customer, but we could not confirm from public text which customers they are or how large their shares are. Because the Q3 orders were driven by "customer capacity expansions," the expansion timing of a handful of large customers shapes the order curve. Readers should keep that in mind.

Be clear about what the contracts really lock in. Backlog consists of purchase orders customers have placed, and equipment orders like these can be delayed or even cancelled when the industry turns down. In Q2, Auto/EV orders were only about a quarter of Auto/EV revenue, which is what project-based orders look like when they simply stop once delivered. The $58 million to $60 million backlog is closer to a production schedule than a locked-in obligation.

The most fragile link is the Environmental Technologies segment. In Q2 it had $5.8 million of revenue and a $0.6 million operating loss, the only one of the three segments losing money. The Electronic Test segment, with $21.4 million of revenue and $3.8 million of operating profit, carried all the profit (Q2 earnings release, 2026-08-10). One of the new growth drivers management highlighted, the high-power chiller, belongs to this temperature-control product line. Whether new products can push this segment back to profit in the second half will largely determine whether the 44% gross margin guidance holds up.

Within the peer group of semiconductor equipment and materials companies, the comparable figures are mainly revenue growth and order intake:

inTest's growth sits in the middle of the group, and it is moving in the same direction as everyone else: up. Its two closest comparisons are AEHR and CAMT. AEHR also works in testing and has orders of a similar size; CAMT's back-end inspection business overlaps with inTest's customer base. What separates the three is not growth but margins and pricing power. For every $100 of equipment AMAT sells, $33 is left as operating profit. For inTest, it's less than $1.

An analogy helps. Picture two restaurants. For every $100 the big chain takes in, more than $30 is left after food and labor. In Q2, for every $100 inTest took in, about $60 went to ingredients and nearly $40 to rent and wages, leaving just a few dimes in its pocket. That's why order growth means more to inTest than to the giants. As long as gross margin holds, most of each extra dollar of revenue drops straight to profit. That is the reason the market is willing to pay a high multiple, and also the reason profit could vanish the moment gross margin slips.

As for how the market has priced similar results, comparable peers have not yet reported full quarterly results this season, so there is no next-day earnings reaction to compare against. We won't force a comparison.

What does a P/E of 213 mean? On trailing-twelve-month GAAP earnings per share, it would take more than two hundred years to earn back the market value. The number is no longer useful for comparison, because the denominator is so small that a few hundred thousand dollars of extra profit would cut the multiple in half. For a company that has only just climbed from losses to breakeven, the right measuring stick is EV/Sales (enterprise value, or market value plus debt minus cash, divided by revenue), cross-checked against an annualized multiple of adjusted EBITDA.

An EV/Sales of 1.26x doesn't look expensive; semiconductor equipment companies usually trade at more than 2x sales. But this yardstick needs a discount for inTest, because its margins are a sliver of its peers'. The same dollar of revenue turns into dozens of times less profit.

So what has the market already paid for in advance? The stock is up about 91% over the past twelve months. Comparing the multiples on the top of full-year guidance and on 2025 actual revenue, the market has largely priced in this year's 19% to 23% revenue growth. What it hasn't fully priced is two things: the boost to profits once gross margin gets back above 44%, and whether the $60 million backlog can keep producing new growth into 2027. Eight of the nine covering analysts are bullish and one is neutral (Wall Street ratings summary), so this isn't an overlooked stock. Almost all of the remaining upside rides on margins actually showing up.

Over the past two years, AI has turned the big cycle in semiconductor test equipment into a steep climb. SEMI's mid-year forecast has test equipment sales up 55.3% in 2025, rising another 31% in 2026 to $15.3 billion, and reaching $20.8 billion by 2028. Packaging and assembly equipment is forecast to grow 9.6% in 2026 to $6.7 billion (SEMI mid-year equipment forecast, 2026).

Set against inTest, those numbers show a glaring contrast. Industry test equipment sales jumped more than 50% in 2025, yet inTest's revenue fell 12.9% the same year. The reason is that the first wave of AI testing money went to the big makers of testers for HBM memory and high-end system-on-chip processors. inTest works in back-end peripheral equipment (temperature control, manipulators, docking) and in analog, mixed-signal and power-chip testing, a layer that benefits one step later. The makeup of Q3 orders shows the wave is now spreading beyond the GPUs themselves to data-center power conversion and analog chips. The high-power chillers management mentioned are built for power-chip testing.

Three readings help place where we are:

Our read: the overall test-equipment cycle is in the middle of its upswing, while inTest's layer (back-end peripherals plus power and analog testing) is early in its upswing. The former has been confirmed repeatedly by the big players' results. The latter has only just produced its first record order quarter and needs one or two more quarters of follow-through to be on solid ground.

Two early warning signs of a peak to track:

This section is about weekly signals that govern entry and exit timing over the next few weeks. Readers holding for a year or more can skip it.

The sector money-flow data we track doesn't cover stocks as small as inTest, and it isn't part of any current sector-rotation theme. So there's no standalone money-flow reading for its sector. We can only infer from where money is going in the broad market and in neighboring segments.

Readings as of 2026-10-06:

In plain trader terms: money is flowing into tech and along the semiconductor equipment chain, with metrology/inspection and vacuum control the two most crowded destinations, right next door to inTest's niche. That's a tailwind. At the same time, small caps are being sold across the board, high-yield bonds are seeing outflows, and the 10-year yield sits above 5%. For a stock worth less than $200 million with limited daily trading volume, those are real headwinds.

On price, the stock is about 23% above its 50-day moving average and about 13% above its 200-day, after rising about 90% over twelve months. On the first trading day after the preview, it actually slipped 0.7%. When good news fails to push a stock higher, it suggests the rally has run ahead of the money, and that the record orders were partly traded before the announcement.

Long term (quarters to a year): worth following. inTest's fundamentals are at an unusual turning point. Revenue has grown for four straight quarters, orders and backlog just set records, and the growth engine has switched from fading Auto/EV projects to back-end semiconductor testing, which is early in its cycle. The balance sheet has net cash plus $40 million of undrawn credit, so no need to raise money from shareholders. At 1.26x EV/Sales, the valuation hasn't overpaid for revenue growth, but it already treats the coming profit leverage as half done. The costs you'd be accepting are also clear: margins so thin that a swing of a few hundred thousand dollars can rewrite a quarter's GAAP result; SBC eating nearly 40% of adjusted EBITDA; option exercises diluting shareholders by several percent a year; opaque customer concentration; and back-end equipment orders that have always been boom-and-bust. Following this stock should be driven by checkpoints. If Q3 gross margin gets back to 44%, Q4 revenue guidance is no lower than about $35.8 million, and semiconductor orders don't fall noticeably in Q4, all three together would suggest this is the start of a margin upcycle. If any one of them misses, the long-term case should be downgraded a notch and reassessed.

Short term (days to weeks): don't chase, wait. The stock has moved well above its medium-term moving averages, no new buying pushed it higher the day after the good news, small-cap style funds are flowing out across the board, and rates are tight. Money is entering the neighboring semiconductor equipment space but hasn't yet spilled over to stocks this small. Buying now would be a bet on sentiment continuing, without money flow to back it. What to wait for is one of two things: the stock pulling back to near its medium-term average and stabilizing on lighter volume, or small-cap outflows easing noticeably along with gross margin confirmation on November 6.

The second half of this note turns these two calls into conditions you can check daily and three reference price levels.