CEO committed to sequential growth every quarter in 2026, then told analysts the actual AI step function lands in 2027.
Thesis: Navitas has a differentiated position in high-voltage GaN and ultra-high-voltage SiC where the competitor pool genuinely narrows to Infineon. But the revenue that pays for this position won't arrive until 2027 when 800V HVDC and grid architectures transition. Between now and then, 2026 sequential growth rests on low-ASP chargers and AC-DC PSU SiC in a market with real competition. $237M cash funds roughly five years of burn at $46M/year, but the CFO just left with no successor named, and the growth composition is unverifiable.
Verdict: AVOID — Conviction: MEDIUM
Catalyst: Q1 print at or above $8.5M with management reiterating full-year sequential growth and Q2 margin expansion. Two consecutive beats would confirm the pattern management committed to.
Key Risk: Sequential growth stalls in Q2 or Q3 as the mobile drain outpaces high-power additions, forcing another narrative reset while the real 2027 inflection remains unmonetized.
The Tell: When pressed on 800V timeline, CEO volunteered: 'the real step function is really coming from the 800-volt DC, which is really kind of linked to the Kyber rack, which is the higher integration of GPUs in '27.' Unprompted admission that the AI thesis he's selling doesn't generate revenue for another year, while committing to sequential growth all 2026 from businesses he just called low-priority when mobile.
Friction Level: MODERATE_FRICTION — Both sides agree revenue is $30M, inflection is 2027, cash is real. Disagreement: whether the 2026 micro-ramp from low-ASP performance computing and AC-DC PSU SiC justifies owning a stock priced on a $3.5B SAM narrative.
2025Q4 · Feb 25, 2026
3 of 5 readings agreed3 AVOID · 2 SHORTRead the full call
Guiding Sequential Growth All 2026, Real Revenue Inflection Pushed to 2027
Navitas has a differentiated position in high-voltage GaN and ultra-high-voltage SiC where the competitor pool genuinely narrows to Infineon. But the revenue that pays for this position won't arrive until 2027 when 800V HVDC and grid architectures transition. Between now and then, 2026 sequential growth rests on low-ASP chargers and AC-DC PSU SiC in a market with real competition. $237M cash funds roughly five years of burn at $46M/year, but the CFO just left with no successor named, and the growth composition is unverifiable.
When pressed on 800V timeline, CEO volunteered: 'the real step function is really coming from the 800-volt DC, which is really kind of linked to the Kyber rack, which is the higher integration of GPUs in '27.' Unprompted admission that the AI thesis he's selling doesn't generate revenue for another year, while committing to sequential growth all 2026 from businesses he just called low-priority when mobile.
Every segment carries an AI prefix but design wins are samples and evaluations: 12+ customers evaluating GaN samples, 15+ OEMs in grid evaluation, co-development on 10kW brick with no module sales path. CEO confirmed no path to selling modules. None of this is booked revenue.
Management says Q4 was bottom and sequential growth runs through 2026, yet CEO pushed the real step function to 2027: 'the real step function is really coming from the 800-volt DC, which is really kind of linked to the Kyber rack, in '27.' Grid revenue starts 2027 too. 2026 rests on low-ASP businesses.
Language shifted from 'stabilization and transition' to 'revolution,' 'energizing,' 'bright future' from a CEO five months in. Revenue fell 45% YoY to $45.9M. Confidence words increased while the numbers did not.
CEO gave specific content per megawatt ($25-35K), clearly filtered the competitive list (650V GaN in right package + 2kV+ SiC), and answered the SiC-vs-GaN 800V question directly: 'we've not seen any significant use case of board implementation using SiC on the primary side.' No deflection.
$237M cash, zero debt, $96M private placement. Enough runway for the pivot without near-term dilution. Headcount cut 19%, distributors consolidated from ~40 to <10 in one quarter. Actions match words on cost structure.
+55
CONFIDENT"CEO made unprompted forward commitments: sequential growth throughout all of 2026, margin expansion from Q2, mobile immaterial by year end. Direct answers with specific numbers on content per megawatt ($25-35K) and competitive filtering. No hedging language in prepared remarks."
Specific Q1 guide, margin path, and opex commitment are testable. But CFO refused to segment high-power revenue, so composition of the growth cannot be verified. No backlog or book-to-bill disclosed. Balance sheet and competitive claims are real but unverifiable from outside.
Q1 print at or above $8.5M with management reiterating full-year sequential growth and Q2 margin expansion. Two consecutive beats would confirm the pattern management committed to.
Sequential growth stalls in Q2 or Q3 as the mobile drain outpaces high-power additions, forcing another narrative reset while the real 2027 inflection remains unmonetized.
“Revenue in the fourth quarter of 2025 exceeded the high end of guidance at $7.3 million compared to $10.1 million in the third quarter of 2025.”
“We expect revenue to increase sequentially to between $8 million and $8.5 million.”
Figures are extracted from the transcripts and checked in code: every quote appears word for word. Descriptive only: none of this feeds the verdict.
Gross margin held at 38.7% flat despite 28% sequential revenue decline. At $7.3M revenue with zero operating leverage, that margin stability says the cost structure is working. Opex dropped $0.5M with 19% headcount cut and 40 distributors collapsed to under 10. Real execution. The competitive position in high-voltage GaN is differentiated: CEO identified Infineon as the only serious competitor, filtered the field to vendors with both 650V GaN in the right package AND 2kV+ SiC. Content per megawatt at $25-35K with $10-12K outside data center gives a real sizing framework. SiC AC-DC for current-gen AI PSUs grows through 2026 before 800V arrives. Q1 guide of $8-8.5M is the first sequential growth print. GlobalFoundries partnership gives domestic supply with production this year.
Revenue is $7.3M per quarter, down 45% YoY, burning $46M annually. The CEO removed mobile from the denominator and called the remainder a transformation. AR collapsed from $9.8M to $3.6M and inventory declined from $14.7M to $13.3M: no build ahead of a claimed ramp. CFO refused to segment high-power revenue, so growth composition is unverifiable. CFO departing after 10 years with no named successor. $96M private placement at a moment revenue is collapsing. Shares went from 222M to 230M guided. The 800V HVDC revenue and grid infrastructure revenue both start in 2027 per CEO. 2026 growth rests on performance computing chargers and AC-DC PSU SiC: low-ASP, high-competition. $16.6M restructuring charge, $12.8M cash, tells you what the old business was worth exiting.
Say/Do Gap
Management commits to sequential growth all 2026 but CEO explicitly pushed both real growth drivers (800V HVDC and grid) to 2027. The 2026 ramp depends on businesses management itself deprioritized. Composition undisclosed.
AI Label Without Substance
12+ customers evaluating samples, 15+ OEMs in grid evaluation, co-development on 10kW brick with no path to selling modules. CEO: 'At this stage, we don't see a path where we're going to sell the modules.' None of this is booked revenue.
Unsustainable Trend Confidence
Language shifted to 'revolution' and 'bright future' while revenue fell 45% YoY. CEO five months in making full-year sequential growth commitments on a $30M run-rate business.
AVOID. The trade is binary: either 2026 sequential growth holds or it doesn't. If it holds, you're up on a micro-cap with a real technology position. If it stalls, the whole narrative resets while the real 2027 catalyst is still 18 months away. I can't size what I can't verify because CFO won't segment high-power revenue. Inventory declining and AR collapsing before a claimed ramp is the wrong pattern. Wait for two more prints with disclosed segment composition. If Q1 comes in above $8.5M and Q2 shows margin expansion with high-power breakout disclosed, re-evaluate as a LONG. Right now the asymmetry is unknown, and unknown risk is unpriceable.