The CEO overrode his CFO's guidance floor on the call while the segment it covers just declined 2% for the full year.
Thesis: MPWR is a well-executed analog power franchise with real design wins expanding content per socket, but the forward setup depends on enterprise data recovering from a 2% decline to 50%+ growth on backlog that management admits they can't yet separate from double ordering. The margin story stalls because mix hasn't shifted enough yet. The asymmetric risk is that H2 2026 backlog unwinds when hyperscaler capacity fears normalize, arriving into the same window as MPWR's own capacity expansion.
Verdict: HOLD — Conviction: MEDIUM
Catalyst: Q1 2026 print confirms backlog converts to revenue without channel inventory building. Gross margin ticks above 55.8% as Balow promised 10-20bps sequential cadence resumes. Packaging innovation shipping this quarter with customer qualification already complete.
Key Risk: Backlog unwinds from double ordering. Bernie explicitly said customers are extending orders "concerned about capacity constraints" industry-wide, not MPWR-specific. When perceived shortage resolves, orders cancel. MPWR's $4B capacity arrives into the same window.
The Tell: Hsing interrupting Balow's guidance mid-sentence: "50%. I thought that we can do a lot more than that, conservatively." He then rambled about design wins across "magnetic fixing seven or eight" customers. The CEO publicly overriding the finance team's number on the record, unprompted, in his final CFO's last call. Either he's supremely confident or he's selling the print.
Friction Level: MODERATE_FRICTION — Both sides agree on facts. Bull reads the raised enterprise data guide and clean channel inventory as structural demand. Bear reads the same guide as double ordering inflating backlog against a shrinking base. Bernie himself flagged the split: "trying to figure out what's real demand and what may be some double ordering."
2025Q4 · Feb 5, 2026
5 of 5 readings agreed5 HOLDRead the full call
Enterprise Data Fell 2% in 2025, 2026 Guide Raised 20 Points in One Quarter
MPWR is a well-executed analog power franchise with real design wins expanding content per socket, but the forward setup depends on enterprise data recovering from a 2% decline to 50%+ growth on backlog that management admits they can't yet separate from double ordering. The margin story stalls because mix hasn't shifted enough yet. The asymmetric risk is that H2 2026 backlog unwinds when hyperscaler capacity fears normalize, arriving into the same window as MPWR's own capacity expansion.
Hsing interrupting Balow's guidance mid-sentence: "50%. I thought that we can do a lot more than that, conservatively." He then rambled about design wins across "magnetic fixing seven or eight" customers. The CEO publicly overriding the finance team's number on the record, unprompted, in his final CFO's last call. Either he's supremely confident or he's selling the print.
Enterprise data, the largest end market and the AI power narrative driving the multiple, declined 2% in 2025 while total revenue grew 26.4%. The CFO raised 2026 enterprise data from 30-40% to 50% floor based on one quarter of ordering patterns. Hsing walked back the 40% non-enterprise figure in real time. Balow declined to put any growth number on automotive.
Book-to-bill "well in excess of one" with backlog into Q2 2026, but Bernie flagged the core uncertainty directly: customers extending orders because of industry-wide capacity concerns produces inflated orders that unwind when shortage perception resolves. He cited adapting to this "as well as anybody... in late 2020 and early 2021". That precedent ended badly for those who held stale backlog.
Balow raised enterprise data floor from 30-40% to 50%. Hsing immediately interjected: "50%. I thought that we can do a lot more than that, conservatively." The founder publicly undercut his own CFO's guide by 10+ points on the record. 14 consecutive years of growth backs the credibility, but one of those numbers is wrong.
Gross margins stuck at 55.5-55.8% for 4-6 consecutive quarters, low end of the 55-60% model, during a period of 26.4% revenue growth. Finance attributes the stall to lacking "longer time horizon" on backlog and promises 10-20bps improvement once visibility improves. Improvement deferred to a future quarter requiring better backlog quality than currently exists.
Book-to-bill well above one, backlog extending into Q2 2026, channel inventory at low end of range. Packaging innovation (doubling current density to 3A/mm²) already customer-qualified and shipping this quarter. $4B in geographically balanced capacity secured. Real assets backing the order book.
-25
CAUTIOUS"Balow explicitly deferred automotive guidance citing tariffs, EV subsidies, memory shortage. Hsing said 800V revenue is "not this year, not even next year." Bernie hedged on H2 visibility: "still a lot of variables." The CEO was confident but the finance team was hedging every forward number."
Specific numbers throughout: $751.2M, 26.4% growth, 55.5-55.8% margins, $4B capacity, 5% optical module mix. But guidance was deliberately ambiguous on automotive and H2. Backlog quality flagged by management themselves as unquantified.
Q1 2026 print confirms backlog converts to revenue without channel inventory building. Gross margin ticks above 55.8% as Balow promised 10-20bps sequential cadence resumes. Packaging innovation shipping this quarter with customer qualification already complete.
Backlog unwinds from double ordering. Bernie explicitly said customers are extending orders "concerned about capacity constraints" industry-wide, not MPWR-specific. When perceived shortage resolves, orders cancel. MPWR's $4B capacity arrives into the same window.
“For Q4 2025, we had a record quarterly revenue of $751.2 million, 1.9% above Q3 2025 and 20.8% higher than 2024.”
Changed more than 78% of calls
Wording compared with the Oct 30, 2025 call, ranked among 61 calls reporting on Q4 2025. A measure of how much the story moved, not a signal about the stock.
Figures are extracted from the transcripts and checked in code: every quote appears word for word. Language change is computed from word frequencies, with no model involved. Descriptive only: none of this feeds the verdict.
Fourteen straight years of growth with record $751.2M in Q4. Book-to-bill well above one, backlog into Q2 2026, channel inventory at low end. Non-enterprise data grew 40%+, automotive 43%, optical modules hit 5% of sales in just over a year. The packaging innovation doubling current density is customer-qualified and shipping this quarter. $4B capacity commitment is real. CEO overriding CFO guidance floor to push enterprise data above 50% suggests buy-side whispers are too low. Content expansion from silicon to modules to systems raises ASP per design win without needing share gains. 14 straight years of executing.
The AI segment shrank 2% in 2025 while the stock priced on AI power narrative. CEO overrode his CFO's guidance floor on the record, one quarter after it was set. Backlog quality is explicitly flagged as uncertain: Bernie said customers are extending orders due to industry-wide capacity fears, which is exactly the pattern that produced double ordering. Hsing wouldn't underwrite the 40% non-enterprise figure for 2026. Balow wouldn't put a number on automotive. Margins flat for 4-6 quarters despite 26% growth. 800V GaN revenue is 2027+ per Hsing. CFO retiring during record quarter replaced by interim controller. $4B capacity arrives into the same window backlog matures.
Backlog Quality
Bernie himself flagged the split between real demand and double ordering. He cited adapting to the 2020-21 dynamic. Customers extending orders over industry-wide capacity fears, not MPWR-specific capacity, is textbook double-order setup.
Say/Do Gap
Enterprise data declined 2% in 2025 and guidance was raised from 30-40% to 50%+ in one quarter on ordering patterns alone. CEO walked back the 40% non-enterprise figure. Automotive guidance withheld. Management underwrites almost none of the 2025 growth drivers for 2026.
Unsustainable Trend Confidence
Gross margins stalled at 55.5-55.8% for 4-6 quarters during 26.4% revenue growth. Margin expansion promise depends on backlog visibility that management admits they can't fully assess. Improvement pushed to a future quarter.
HOLD. Execution is real: 14 straight years, clean channel inventory, qualified products shipping now. But I can't size a position when the forward thesis rests on a segment that shrank 2% last year growing 50%+ this year, on backlog management admits may be inflated, with margin expansion deferred to a future quarter that requires better visibility than they have. The asymmetry favors the bear on H2 2026 if backlog unwinds, but favors the bull through Q1 2026 on printed numbers. Wait for Q1 to see if backlog converts without inventory building. If channel inventory stays low and margins tick above 55.8%, size in. If inventory builds or margins stay flat, stay out.