Aehr Test Systems Is Losing Money and Gaining Customers. The Market Is Betting on the Latter.

The bookings story

The headline number requires context. Aehr's $37.2 million in quarterly bookings is roughly 3.6 times the $10.3 million in revenue it recognized during the same period. That ratio implies a sharp inflection in demand that has not yet flowed through to the income statement.

The demand is coming from two primary vectors. First, Aehr's package-level burn-in (PLBI) business, built around its Sonoma systems, is seeing what CEO Gayn Erickson described as "significant forecasts" from a lead hyperscale customer for production burn-in of custom AI processor ASICs. That customer has now awarded Aehr the production win for its next-generation, higher-power AI accelerators, with a "significant near-term follow-on production order" expected to ship during fiscal year 2027.

Second, Aehr's wafer-level burn-in (WLBI) business is expanding into silicon photonics. The company announced on March 31 that it had won a major new customer developing advanced silicon photonics-based transceivers for hyperscale data center optical interconnects. That initial order was for multiple FOX systems for both qualification and production. A follow-on order from an existing silicon photonics customer followed shortly after.

Erickson framed the silicon photonics opportunity in structural terms: as data center architectures scale to support AI workloads, fiber optic interconnects offer advantages over copper wiring in data rates, power consumption, and thermal performance. The implication is that burn-in demand for silicon photonics devices will grow alongside the broader buildout of AI data center infrastructure.

The income statement gap

The current financials tell a different story. Fiscal third-quarter revenue of $10.3 million was down from $18.3 million a year earlier. The GAAP net loss widened to $3.2 million (or $0.10 per diluted share) from a loss of $0.6 million in the year-ago quarter. Non-GAAP net loss was $1.5 million, or $0.05 per diluted share, which beat the consensus estimate of a $0.08 loss.

For the first nine months of fiscal 2026, revenue totaled $31.2 million (down from $44.9 million), and the company burned $5.1 million in operating cash. Cash on hand stood at $37.1 million as of February 27, up from $31.0 million at the end of the prior quarter.

Management reiterated its second-half guidance: revenue between $25 million and $30 million, and a non-GAAP net loss per diluted share between $0.09 and $0.05. The company now expects full-year fiscal 2026 revenue to land on the "high side" of the previously stated $45 million to $50 million range, and second-half bookings to hit the high end of the $60 million to $80 million range.

The valuation question

Here is where the math gets uncomfortable. At a recent price near $50, Aehr's market capitalization sits around $1.5 billion. Trailing twelve-month revenue is approximately $53 million. That implies a price-to-sales ratio north of 28x on a company that is currently unprofitable.

To be sure, the forward-looking indicators are genuinely strong. A 3.5x book-to-bill ratio, a $50.9 million effective backlog, and production wins with hyperscale AI customers represent a credible path to significant revenue growth in fiscal 2027. Aehr is also scaling manufacturing capacity, including a newly upgraded contract manufacturing facility capable of producing more than 20 additional Sonoma systems per month.

But the stock's 188 percent year-to-date gain has priced in a substantial portion of that growth already. The median analyst price target of $36.33 (per Yahoo Finance) sits well below the current trading price, and only three analysts cover the stock (one Buy, two Hold). Lake Street Capital raised its target to $50 from $29 in early March, but that target has already been reached.

What comes next

The near-term catalyst calendar is straightforward: Aehr's fiscal fourth quarter (ending May 29) will determine whether the company can convert its record backlog into recognized revenue. Management has guided for $25 million to $30 million in second-half revenue, meaning the fourth quarter needs to deliver roughly $15 million to $20 million to hit the range. The expected follow-on production order from the lead hyperscale customer, if announced, would further validate the fiscal 2027 growth trajectory.

The broader question is whether Aehr can sustain its positioning as AI chip complexity increases. The company's thesis is that as semiconductor devices grow in size, power, and packaging complexity (particularly in advanced packages like CoWoS that combine multiple die), the cost of discovering defective devices after packaging far exceeds the cost of burn-in testing beforehand. That argument has resonated with at least one major hyperscale customer and several silicon photonics manufacturers.

The risk is concentration. Aehr's bookings are heavily weighted toward a single hyperscale customer for PLBI and a small number of silicon photonics customers for WLBI. Any delay or reduction in orders from those accounts would disproportionately affect the growth story the market is currently pricing in. And at 28x trailing sales with negative earnings, the stock leaves little room for disappointment.