In the morning review we flagged Navitas (NASDAQ: NVTS) as a report requiring particular caution, because of the gap between the story and the numbers. The report published last night confirms the caution - and adds a layer to it: the bottom line shows a loss of $228.2 million, but almost none of it relates to the business itself.
About the Company: The Chips That Convert Electricity
Navitas manufactures power chips from two advanced materials: gallium nitride (GaN) and silicon carbide (SiC). Both replace ordinary silicon in one role: converting electricity.
To explain why that matters. Electricity reaching a server passes through a chain of conversions - from high voltage to low, from alternating current to direct. In each conversion, energy is wasted as heat. GaN and SiC chips do it more efficiently: less loss, less heat, less cooling required. In a data center consuming tens of megawatts, an improvement of a few percentage points in conversion efficiency is an enormous saving.
And that is precisely the company's strategy. Navitas is in a transition it calls "Navitas 2.0" - a shift from the consumer market (phone chargers, cheap consumer products) to infrastructure: AI data centers, the power grid and energy. In the first quarter it launched a 20-kilowatt GaN platform converting 800 volts to 6 volts for data centers, as well as fifth-generation SiC products for AI power supplies.
What Was Reported: Read the Bottom Line Carefully
The loss in the headline. The GAAP loss totaled $228.2 million - an enormous figure relative to a company with $10 million of quarterly revenue. But $203.1 million of it is a non-cash charge on an earnout liability - an accounting adjustment of a future obligation, not money that left and not a deterioration in operations.
Excluding the unusual items: the adjusted loss stood at $9.3 million, or $0.04 per share - exactly in line with the analyst estimate.
This is the same lesson that recurred throughout the season, from Intel to Philip Morris: check what entered the bottom line before drawing a conclusion from it.
Revenue. $10.5 million - growth of 22% versus the prior quarter ($8.6 million), above the estimate of $9.97 million and also above the midpoint of the company's own guidance. And yet, the figure is still below $14.5 million in the comparable quarter a year ago - meaning on a year-over-year basis this is still a decline.
Gross profitability improved: the adjusted gross margin rose to 39.5%.
The engine that is working: the high-power market - AI infrastructure, power grid and energy - grew over 50% year-over-year. Meaning the strategic transition is indeed happening: the new part is growing fast, but it is still too small to offset the contraction in the old part.
The Guidance: The Most Meaningful Point
For the third quarter the company guides to revenue of $13.5 million (within a range of ±$0.5 million) - sequential growth of about 28% - and an adjusted gross margin of about 39.7%. Adjusted operating expenses are expected at $15.5-17.5 million.
And the important detail: per the company, in that quarter it is expected to return to year-over-year growth - the inflection point investors have been waiting for.
The Bull Thesis
Whoever reads it positively will point out that the transition is working: the high-power market is growing over 50%, the gross margin is improving, revenue beat the estimate, and guidance calls for a 28% jump in a single quarter and a return to annual growth. If that trend line holds, the current quarter is the bottom.
Beyond that: the company is positioned in a real link of the electricity thesis we have been tracking. If Baker Hughes reports its energy equipment orders doubled, and GE Vernova a record turbine backlog - someone also has to supply the conversion components inside the building. Navitas is one of the few offering both GaN and SiC across the full power path.
The Bear Thesis
And the other side must be said clearly, because it concerns magnitude.
The company is very small. Quarterly revenue of $10.5 million means an annual run rate of about $40-50 million. Every figure in the report - loss, margin, guidance - must be read relative to that size.
It is still losing money and burning cash. An adjusted loss of $9.3 million in the quarter, against revenue of $10.5 million, means expenses are nearly double revenue. At that pace, the cash balance is a critical variable.
The year-over-year comparison is still negative. $10.5 million versus $14.5 million a year ago. The sequential growth is encouraging, but the annual picture has not yet turned.
And the gap between story and valuation. The market prices Navitas on the potential in AI infrastructure, not on current performance. That is legitimate in a growth company, but it means the stock is highly sensitive to any deviation from the timetable.
The proportion that must be kept
This is a company with about $10 million of quarterly revenue, reporting an adjusted loss of nearly the same magnitude, whose year-over-year comparison is still negative - inside a story about data centers worth billions. The strategic transition is real and the quarterly numbers are improving, but the gap between the company's size and the size of the market it targets is precisely both the source of risk and the source of opportunity. There is no halfway thesis here.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
In the morning I flagged Navitas as a report requiring reading from both sides, and the report itself justified that - but not in the direction I expected.
The gap I had in mind was between the story and the numbers. The actual gap was between the headline and the content: a loss of $228 million whose overwhelming majority is an accounting adjustment of a liability. Whoever read only the first line got an entirely wrong picture.
And what I take from the numbers themselves: the direction is right, the size is not yet. The high-power market is growing over 50%, the margin is improving, and guidance calls for 28% sequential growth and a return to annual growth. That is exactly what should happen at a company shifting itself from consumer to infrastructure. But the revenue scale - about $10 million a quarter - is a reminder that we are at the beginning of the story, not the middle.
The figure I will follow next quarter is one: whether the company hit $13.5 million. At companies at this stage, meeting guidance matters more than the guidance itself - because it is what separates a story that materializes from a story that gets postponed.
Summary
Navitas delivered a quarter that confirms both sides at once. The positive side: revenue beat the estimate, grew 22% sequentially, the gross margin improved, the high-power market leapt over 50%, and guidance calls for a further 28% jump and a return to annual growth. The side requiring caution: the company is still very small, still losing money, and the year-over-year comparison is still negative.
And the $228 million loss in the headline? Almost all of it is accounting. That is an excellent reminder of the rule that has accompanied us all season - the headline is not the report.
Sources: Navitas Semiconductor's official results announcement for the second quarter of 2026 (July 27, 2026), including revenue, GAAP and adjusted loss, the detail of the non-cash charge, gross margin, the high-power market figures and third-quarter guidance; prior management guidance and the first-quarter report; analyst consensus as covered by Zacks and StockTitan. Data accurate as of the time of writing. The chart is shown in real time via TradingView. Nothing herein constitutes a forecast, recommendation or advice - see the full disclaimer at the bottom of the page.