NEXT PICK · Market Insights
Plunged 17%: 40x price-to-sales ratio hits 3.5x order ratio
Thursday, July 2, 2026
Revenue halved, price-to-sales ratio about 40 times, but order-to-shipment ratio 3.5 times, with record-breaking backlog; Long and short bets on the July 14 financial report.
At the close on July 2, Aehr Test Systems ($AEHR) plunged 17.13% to $69.96, sending this $2.2 billion semiconductor testing equipment company into the spotlight in a single day.
Strangely, the company did not issue any financial reports or cancel orders that day—the trigger was far away: Bloomberg reported that Meta plans to sell surplus AI computing power, instantly sensing a "computing power surplus" in the market. The Philadelphia semiconductor index plunged 6% in a single day, marking the worst two-day losing streak in nearly a month.
The AEHR Beta is as high as 3.18, and if you sneeze, it will catch a severe cold. The sector fell 6%, while it dropped 17%. This is not hard to calculate.
What is truly worth asking is
Why is a company with quarterly revenue of only $10.3 million, a 44% year-on-year slash, still worth $2.2 billion after a sharp drop? And with its just-delivered $37 million quarterly orders and a 3.5x order-to-shipment ratio, does this mean the worst is over?
|
Latest price
$69.96
▼ -17.13%
|
Orders in hand
$50.9 million
▲ Setting a historic record
|
Sales-to-market ratio
About 40 times more
▼ 2-5 times that of industry peers
|
The "devil's test hall" before chip shipment has been done for nearly fifty years, doing only one thing
Before delivery, semiconductors undergo extreme "baking" under high temperature and high pressure, screening out chips that may die prematurely. This process is called burn-in. Aehr Test Systems was founded in 1977 and is headquartered in Fremont, California, having devoted nearly fifty years to just one thing.
The core products FOX-XP and FOX-NP systems can directly test and burn on entire wafers, paired with WaferPak contactors supporting 300mm wafers, covering silicon carbide (SiC) power chips, photonics devices, memory, processors, and other categories; FOX-CP offers low-cost single-wafer solutions. FY2024 full-year revenue is $59 million.
SiC power chips in electric vehicles and silicon photonic modules in AI data centers all have extremely high failure costs and must be fully screened at the wafer stage—this is Aehr's unique expertise. Imagine it as the only "extreme test track" in the chip industry: automakers can change tire suppliers, but it's very difficult to change that test track.
A price-to-sales ratio of 40 times, with a free cash flow yield of 0.52%.
Opening the valuation snapshot, the price-to-earnings ratio (P/E) has been directly invalidated due to ongoing losses—a loss of $0.38 per share over the past twelve months. Comparing FY2024 revenue of $59 million to a market value of $2.2 billion, the price-to-sales ratio (P/S) is around 37 times; If you calculate the annualized revenue of the lowest quarters in recent quarters, it exceeds 40 times, while industry peers generally only see 2 to 5 times.
A price-to-book ratio (P/B) of 15.85, return on equity (ROE) of -6.60%, and free cash flow yield of only 0.52%—these three figures together mean shareholders' capital generates neither profit nor cash. Based on the normal range of 10-15 times price-to-sales ratio in the semiconductor equipment industry, the reasonable market capitalization is only $590 million to $885 million, corresponding to 60%-75% downside potential.
On the cash flow side, operating cash flow has been about $4 million to $4.6 million per quarter in recent quarters, but fortunately, existing reserves can support operations for another two to three years.
So why does the market offer a price-to-sales ratio of 40 times? The bulls' answer is just one word: order. The record-breaking $50.9 million backlog is the only collateral worth offering on this highly valued set.
Between the commercialization pilot and the pre-breakthrough lies a financial report from July[14]
Products have long been mass-produced and shipped, with customer lists moving from Customer A to Customer F, and some customers moving from the validation stage to production orders—a typical transition from commercialization pilots to pre-breakthroughs. However, quarterly revenue fell from a high of $18.3 million to $10.3 million, indicating that the volume growth path is far from stable.
The technical side also says "transition."
After the plunge, $69.96 has fallen below the 50-day moving average of $97.74, but remains well above the 200-day moving average of $48.08. The mid-term narrative is cooling down, the long-term trend has not yet been broken, and the market itself has not yet decided which stage to place it in.
The king of niches surrounded by giants
Semiconductor testing equipment is the territory of the giants
Advantest has annual revenue exceeding $3 billion, Teradyne over $6 billion, Cohu about $500-600 million, and FormFactor about $600-700 million. AEHR's $59 million revenue is insignificant compared to them, with its R&D budget and global footprint lagging behind.
But in the wafer-level aging testing niche, the situation has reversed. WaferPak, which supports 300mm full-wafer testing, is an industry-scarce capability. Once customers implement FOX systems, the 6 to 18 month validation cycle, deeply bonded testing procedures, and standardized consumables systems all result in expensive conversion costs.
How much score would this moat get? Its unique technology and customer loyalty earn it 70%, while scale disadvantage mins the remaining 3%. The real question is: when Advantest and Teradyne seriously enter wafer-level aging, can this niche still be held?
A serviceable market of $1.5-2.5 billion, and an awkward denominator
The global semiconductor equipment market is about $100 billion, with test equipment accounting for 8%-10%, about $8-10 billion, and is expected to grow to $12-15 billion by 2027. AEHR truly benefits from wafer-level aging and related segments: SiC power chips, photonics devices, sensor testing combined, serving a market worth about $1.5 to $2.5 billion.
Based on FY2024 revenue, the company holds a share of approximately 2%-4% in the wafer-level aging market. The report's three-year simulation suggests three paths: the baseline scenario returns revenue to $85 million to $95 million by 2027; The optimistic scenario surges to $120 million to $150 million; The pessimistic scenario continues to shrink and approach a cash flow crisis. The ceiling isn't low; the problem is the speed of climbing the ladder.
The bulls' trump card: a 3.5x order-to-shipment ratio
The income report looks grim, but the FY2026 Q3 financial report released in April hid all the reasons why bulls dared to increase positions against the trend:
| ▲ Bull Case |
| ① | Single-quarter orders exceeded $37 million, book-to-bill ratio exceeded 3.5 times, and backlog reached a record $50.9 million |
| ② | An AI accelerator client added $14 million in orders, and a leading hyperscale cloud client awarded package-grade aging mass production orders |
| ③ | A new silicon photonics client was signed, and another $12.2 million in new orders came in in the first five weeks of the fourth fiscal quarter |
| ④ | All end markets are high-growth tracks: electric vehicle SiC, AI data centers, silicon photonic modules, sensors |
| ⑤ | The testing barriers for 300mm whole wafers combined with nearly fifty years of technical accumulation mean that the first-mover advantage in the niche market remains |
| ⑥ | With an income base as low as tens of millions of US dollars, any large order can bring explosive growth potential |
| ⑦ | The 17% plunge that day was caused by rumors of Meta's excess computing power causing a sector crush, not bad news from the company itself |
The bears' ledger: income halved, valuations still hanging in the sky
Bears don't need to tell stories; they just need to lay out the financial reports:
| ▼ Bear Case |
| ① | The most recent quarterly revenue was $10.3 million, a sharp year-over-year drop of 44%, and has declined for several consecutive quarters |
| ② | Quarterly GAAP net loss was about $3.2 million, with net margin deteriorating to -31%, clearly showing economic scale |
| ③ | Operating cash flow has a net outflow of $4 million to $4.6 million per quarter, with a free cash flow yield of only 0.52%. |
| ④ | The price-to-sales ratio is about 40 times, while peers are only 2-5 times, which means there is a 60%-75% downside at normal multiples |
| ⑤ | Customer A and Customer B contribute the majority of revenue; any company cutting orders would be a devastating blow |
| ⑥ | Its Beta reached as high as 3.18, and it was just removed from the Russell index at the end of June, then downgraded by analysts on July 1 |
The earnings report in 12 days will decide on this long-short bet
AEHR's next phase is almost entirely event-driven, with the schedule already set:
| • | FY2026 Q4 Financial Report on July 14: Market expects loss per share to narrow to $0.02, with $50.9 million in backlog orders converting revenue as the sole focus, with both positive and negative factors possible |
| • | Q3 and Q4 2026: Silicon Photonics customers' production orders converted into batch revenue, fulfilling orders restarting the narrative, with a positive impact |
| • | Second half of 2026: If AI infrastructure capital expenditure recovers, semiconductor equipment demand will rebound, which is a sector-specific positive trend |
| • | 2026-2027: Penetration of robotics and electric vehicles drives growth in SiC testing orders, with a positive mid-term outlook |
| • | Continue: New customers will diversify customer concentration, slow variables, but will determine whether valuation discounts can narrow |
From $116 to $70: The cost of expecting to outperform reality
From 2023 to the first half of 2024, AEHR hovered at a low level for a long time, with the market treating it as an ordinary niche equipment vendor. The turning point came after the rise of the AI narrative: in June 2025, a follow-up production order from a silicon photonics customer pushed the stock price from around $30 to $116, nearly quadrupling in two months.
Several subsequent earnings reports ruthlessly disproved it—quarterly revenue slipped from $18.3 million all the way down to $10.3 million, and the stock price plummeted from its peak. In Q2 this year, the Philadelphia semiconductor index posted its best single-quarter gain of 88% in history, and AEHR was pushed back up to the 50-day moving average near 100 yuan, until July 2, when this 17% stubborn stick pushed it back to $70. Expectations preceding reality has always been the only constant script for this stock.
|
⚠️ Risk Notice
|
|
🟡 Neutral Order inflection points and valuation bubbles are real, so it's not advisable to bet on one side before the financial report materializes. |
|
💬 Discussion This is a gaming position, not a bottom position—don't go all out to buy the bottom; Real-time buy and sell positions, posted via AEHR to the official account. |
Data source
| • | Source: NextPick real-time market snapshots, Aehr Test Systems earnings and SEC announcements, Bloomberg and mainstream financial media reports. |