Aehr Test Systems (AEHR) FY2026 10-K Analysis: Semiconductor Test Backlog Up 430%
The Hook
Two years ago Aehr Test Systems $AEHR reported the best year in its history: $33.2 million of net income on $66.2 million of revenue, a 50 percent net margin that would be remarkable for a software company and is close to unheard of for a maker of industrial test equipment. Operating income that same year was $10.1 million.
The gap is a $20.7 million income tax benefit, and the filing is specific about where it came from: a valuation allowance release of $21.9 million, recognized in the fourth quarter of fiscal 2024, when management concluded that deferred tax assets it had previously written off were more likely than not to be realized. Two thirds of the record year came from a bookkeeping conclusion about future profits rather than from selling anything. Those future profits did not arrive. Revenue has fallen every year since, gross margin has given up nearly 14 points, and fiscal 2026 closed with an operating loss of $14.1 million, the worst in at least five years.
And the stock is up roughly six times in twelve months.
Company Snapshot
Aehr Test Systems, incorporated in California in 1977, designs and manufactures test and burn-in equipment for the semiconductor industry. Its systems are the FOX-XP, FOX-NP and FOX-CP wafer contact parallel test and burn-in platforms, plus the Sonoma, Tahoe and Echo package-level burn-in products. The consumables and fixtures around them are the WaferPak full wafer contactor and the DiePak carrier, and those matter more than they sound: contactors are the recurring revenue attached to an installed base.
The demand story has been silicon carbide and gallium nitride power semiconductors, which are the preferred technology for electric vehicle power conversion, and more recently artificial intelligence and inference processors.
This analysis covers the 10-K for fiscal year 2026, the 52 weeks ended May 29, 2026, filed July 27, 2026. One structural note for anyone modelling forward: on April 2, 2026 the board approved moving the fiscal year end from the Friday nearest May 31 to the Friday nearest June 30, effective with fiscal 2027. Fiscal 2027 runs June 27, 2026 to June 25, 2027, so year-over-year comparisons will need care.
The Financial Story
The income statement has deteriorated for three consecutive years, and the deterioration is broad rather than isolated to one line.
| Fiscal year | Revenue | Gross profit | Gross margin | Operating income | Income tax expense / (benefit) | Net income | Operating cash flow |
|---|---|---|---|---|---|---|---|
| FY2023 | $64.961M | $32.746M | 50.4% | $13.375M | $0.060M | $14.557M | $10.011M |
| FY2024 | $66.218M | $32.543M | 49.1% | $10.078M | ($20.698M) | $33.156M | $1.756M |
| FY2025 | $58.968M | $23.933M | 40.6% | ($5.677M) | ($0.381M) | ($3.910M) | ($7.400M) |
| FY2026 | $50.001M | $17.651M | 35.3% | ($14.149M) | ($4.610M) | ($7.126M) | ($3.310M) |
Revenue is down 15.2 percent this year and 24.5 percent from the fiscal 2024 peak. Gross margin has fallen from 49.1 percent to 35.3 percent in two years, which for a capital equipment business usually signals either pricing pressure, adverse product mix, or fixed manufacturing cost spread across fewer units. Operating income has swung by $24.2 million across the same two years, from a $10.1 million profit to a $14.1 million loss, and the loss grew about two and a half times this year alone.
Management is unambiguous about the cause. The MD&A attributes the fiscal 2026 decline "primarily due to a $19.9 million decrease in wafer-level contactor revenue driven by significantly lower shipments, reflecting continued softness in demand" related to electric vehicle power semiconductors, with revenue down in Asia for the same reason. Notice the arithmetic: total revenue fell $9.0 million while that one product line fell $19.9 million. Everything else grew by roughly $11 million, offsetting more than half the contactor collapse. This is not a company where demand broadly evaporated. It is a company that lost one product line into one end market while the rest of the business expanded.
The customer table makes the same point more starkly, and it is the disclosure most likely to be skipped.
| Share of total revenue | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Customer A | under 10% | under 10% | 26.3% |
| Customer B | under 10% | under 10% | 14.2% |
| Customer C | under 10% | 15.1% | 10.9% |
| Customer D | 67.3% | 38.6% | under 10% |
| Customer E | 16.7% | under 10% | under 10% |
In fiscal 2024, the record year, a single customer was 67.3 percent of revenue. By fiscal 2026 that customer is below the 10 percent disclosure threshold entirely, and three different customers account for 26.3, 14.2 and 10.9 percent. So the revenue decline and a genuine reduction in concentration risk are the same event: Aehr lost a whale and replaced roughly half of it with a broader base. Whether that is a disaster or a de-risking depends entirely on which number you were worried about.
Then there is the balance sheet, which moved in the opposite direction from the income statement and moved hard.
| May 30, 2025 | May 29, 2026 | |
|---|---|---|
| Cash and equivalents | $24.529M | $116.358M |
| Current assets | $88.778M | $184.448M |
| Current liabilities | $15.638M | $17.855M |
| Inventory | $41.997M | $41.354M |
| Total assets | $148.508M | $246.694M |
| Shareholders' equity | $122.871M | $219.511M |
Cash rose $91.8 million in a year in which operations consumed $3.3 million, and equity rose $96.6 million in a year the company lost $7.1 million. That combination is only consistent with a large equity issuance. The result is a company with $116.4 million of cash against $17.9 million of total current liabilities and no liquidity question worth asking. Inventory held flat, which is mildly reassuring given falling shipments.
And the forward indicator, which for a capital equipment maker matters more than any trailing line on the income statement:
| Backlog at fiscal year end | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|
| $24.5M | $7.3M | $15.2M | $80.6M |
Backlog went from $15.2 million to $80.6 million, and now stands at roughly 1.6 times the revenue Aehr just reported for the entire year. Aehr also reported Q4 revenue of $18.84 million and record quarterly bookings of $60.7 million in its July 14, 2026 earnings release, which is consistent with the backlog figure in the 10-K. Whatever the trailing income statement says, the order book inflected sharply.
One last item worth flagging because it follows directly from the fiscal 2024 story. Aehr carries $23.8 million of net deferred tax assets with no valuation allowance against them, up from $19.1 million, and the net operating loss component grew from $7.9 million to $13.3 million. The 2024 release was justified on three years of cumulative pre-tax income plus estimates of future profits. Two subsequent loss years have eroded that evidence. If losses continue, the same judgement gets revisited, and re-establishing the allowance would be a large non-cash charge with no cash consequence at all. That is not a prediction, but it is the visible tail of the entry that created the record year.
Valuation
Where it trades. As of July 22, 2026, AEHR traded around $93.52 with a market capitalization near $3.21 billion, against a 52-week range of $15.94 to $126.62 (Investing.com, stockanalysis.com). The stock is up roughly six times off the low.
On fiscal 2026 results that is approximately 64 times revenue and about 14.6 times book value, on a business with a $14.1 million operating loss and negative operating cash flow. There is no meaningful P/E, because there are no earnings. Conventional multiples do not describe this security; the market is not valuing fiscal 2026 at all.
What the price implies. Working backwards is more honest than a DCF here, because projecting free cash flow from a company with one loss-making year and a 430 percent backlog increase produces a range so wide it carries no information. Instead, consider what has to be true.
Under a bull case, the $80.6 million backlog converts substantially within eighteen months, revenue roughly doubles from $50 million toward the $90 to $100 million range, gross margin recovers toward the high-40s as fixed manufacturing cost is spread across more units, and Aehr re-establishes double-digit operating margins on a broader customer base. On $95 million of revenue at, say, a 15 percent operating margin, that is roughly $14 million of operating profit. At a generous 40 times, the equity would be worth on the order of $560 million. That is well under a fifth of today's market capitalization. To justify $3.21 billion on those margins, revenue has to reach several hundred million dollars, which requires the AI burn-in market to become a far larger business for Aehr than power semiconductors ever were.
Under a base case, backlog converts more slowly and unevenly, revenue recovers to the $65 to $75 million range, and margins land in the low-40s with operating income near breakeven to modestly positive. That is a materially better company than fiscal 2026 and still nowhere near the current price on any earnings-based measure.
Under a bear case, the backlog proves to be lumpy system orders that slip or cancel, the AI opportunity attracts better-capitalized competition, and revenue stalls near $50 million with margins in the mid-30s. Cash of $116.4 million funds that for years, so the outcome is not insolvency, it is a long period of a $3 billion valuation attached to a $50 million revenue base.
The output. On any assumption grounded in the reported financials, implied value sits far below the current price, and the gap is not small. What the price is actually expressing is a claim about the size of the AI test and burn-in market and Aehr's position in it, on a time horizon longer than the fiscal 2026 statements cover. That claim may prove right. It is not a claim these statements can verify. Implied value under stated assumptions, not a price target, and not investment advice.
Risks
Concentration cuts both ways, and the improvement is recent. Three customers were 51.4 percent of fiscal 2026 revenue, and one customer was 35.0 percent of gross accounts receivable at year end. Aehr's own risk factors say it expects sales to a limited number of customers to keep accounting for a high percentage of net sales. A single large customer deferring orders is what produced the last two years, and the new base is not yet broad enough to absorb that.
Backlog is an order book, not revenue. For lumpy capital equipment, timing slips, and $80.6 million concentrated in a few system orders is a different asset from $80.6 million spread across many. The 10-K discloses the total, not the composition. The gross margin question is unresolved too: a 35.3 percent margin recovering toward the high-40s is an assumption, and if the new AI-driven mix carries structurally lower margins than the old contactor business, the operating leverage the bull case depends on does not appear. Layer on the deferred tax asset exposure above, the fiscal year change complicating comparisons, and a valuation with essentially no support from trailing financials, and the distribution of outcomes is unusually wide in both directions.
The Bottom Line
Aehr's income statement is the worst it has looked in five years and its balance sheet and order book are the best they have ever looked. Both are true simultaneously, and which one deserves the weight is the entire question. The trailing numbers describe a company that lost a dominant customer in a soft end market; the forward numbers describe one that has replaced that customer several times over in bookings and holds $116 million of cash to deliver against it. What the record fiscal 2024 headline should teach is narrower and more durable: read the whole income statement, because a 50 percent net margin at an equipment maker was never operations. What to watch next is the conversion rate, specifically how much of the $80.6 million backlog turns into revenue in the first two quarters of fiscal 2027, and whether gross margin recovers as it does. That single ratio will settle most of the disagreement in the valuation.
Every figure in this analysis was pulled from Aehr Test Systems' SEC filings through the RoboSystems SEC Shared Repository. No analyst wrote it, and the same pipeline reads the XBRL of any of the roughly ten thousand companies that file. Run your own queries at robosystems.ai/pricing.
Sources: Investing.com · stockanalysis.com