Dormant
POWI · Power Integrations, Inc.
Last analysed ·
Resolved Graded and closed 2026-06-15 at medium conviction — the published kill line did not fire. Coverage continued after the close; the read below is dated 2026-09-19 and is not part of the scored record. Research has since re-rated the name low; the record keeps the graded tier.
Current thesis
The 800VDC AI-power re-rate has cracked. POWI lost the $74–78 May breakout shelf in a -20% week to $70.45 (2026-07-17) and now sits below the $76.20 average target for the first time since spring. The GaN cohort broke with it — NVTS is -41% in a month on a Wolfspeed patent suit and ON Semi's GaNEXUS entry. The 2026-08-05 Q2 print is where narrative has to convert into disclosed data-center dollars.
Kill line
A weekly close below $60 — the July swing low and the pre-breakout base — completes the round-trip of the entire 800VDC re-rate and ends the 2026 uptrend. Secondary: the 2026-08-05 Q2 print passing without a named AI data-center design win or quantified bookings, or the GaN cohort (NVTS, ON, MPWR) printing fresh lows together.
Pick status
Played out resolved published kill line did not fire graded at medium · since re-rated low How this is scored →Latest analysis and events for POWI —
As of 19 September 2026, the latest FrontierPicks analysis for Power Integrations, Inc. (POWI): 5 June 2026: broke -9.71% to ~$77 (AH ~$75.98) in a SECTOR-WIDE semi selloff (Broadcom soft AI/hyperscaler outlook, Samsung strike risk, TSMC stake-sale chatter, rare-earth supply, hot jobs print) — Microchip fell with it. NOT company-specific news.
Kill line: A weekly close below $60 — the July swing low and the pre-breakout base — completes the round-trip of the entire 800VDC re-rate and ends the 2026 uptrend. Secondary: the 2026-08-05 Q2 print passing without a named AI data-center design win or quantified bookings, or the GaN cohort (NVTS, ON, MPWR) printing fresh lows together.
Next dated event on file: — catalyst in 10d.
Current Thesis
Power Integrations' recovery case rests on industrial demand while artificial-intelligence (AI) power revenue remains distant; delivery against Q3 guidance and a weekly reclaim of the August reference close would establish repair, while a weekly close below $50 would invalidate it. This is a low-conviction recovery hypothesis following the earlier thesis failure.
The September 18, 2026 adjusted close was $50.39, compared with $50.45 on September 4. The 14-period relative strength index (RSI) increased from 18.4 to 46.6 between those dated observations, but the sample is too small to establish a durable base. The latest supplied series reports a three-month decline of 42.1%.
The inference remains that the narrative is dead — the August 5, 2026 earnings call placed potential auxiliary data-center revenue in 2028, and the August 21 weekly close of $55.18 breached the previously published $60 thesis boundary. That failure remains part of the record. A recovery in industrial demand would establish a different earnings foundation; it would not itself validate the earlier AI valuation expansion.
Bullish and bearish views on Power Integrations, Inc.
The model's bull view on Power Integrations, Inc. (POWI), in brief: Industrial demand supports reported sales. The bear view: Commercial AI evidence remains distant. Management's August 5, 2026 call described potential auxiliary data-center revenue in 2028, with the graphics-processing-unit main-power opportunity further out. That call disclosed neither a named data-center customer nor quantified… Both cases follow in full.
Bull Case
- Industrial demand supports reported sales. On the August 5, 2026 earnings call, management reported industrial revenue up 16% year over year and channel inventory of 7.3 weeks. Those observations support the recovery hypothesis, which fails fundamentally if Q3 revenue falls below management's $122 million guidance floor.
- Reported revenue grew sequentially. The August 5, 2026 release reported Q2 revenue of $118.9 million, up 10% sequentially and 3% year over year. Company earnings release.
- Analyst expectations remain above price. Stock Analysis displayed a $77.50 average analyst target across five analysts when checked September 20, 2026. This measures published expectations, not capital committed or evidence that the market will reach those targets. Analyst forecast table.
Bear Case
- Commercial AI evidence remains distant. Management's August 5, 2026 call described potential auxiliary data-center revenue in 2028, with the graphics-processing-unit main-power opportunity further out. That call disclosed neither a named data-center customer nor quantified bookings; it does not establish a near-term revenue catalyst.
- The price structure remains damaged. The September 18, 2026 adjusted close of $50.39 was 42.1% below the supplied $86.99 annual high. The improved RSI reading does not demonstrate recovery of the August 21 reference close of $55.18.
- Earnings measures differ materially. The August 5, 2026 release reported Q2 diluted earnings per share of $0.17 under generally accepted accounting principles (GAAP), versus $0.37 excluding specified expenses. Company earnings release.
Setup & Price Structure
The September 18, 2026 market reference is the split/dividend-adjusted close of $50.39. The $50 round-number boundary is an analytical invalidation threshold, not demonstrated support. A weekly close above the August 21 reference close of $55.18, accompanied by Q3 guidance delivery, defines the proposed recovery case; a weekly close below $50 beforehand ends it. The earlier $60 boundary has already failed and is not being presented as an intact thesis.
Positioning evidence is limited. The August 6, 2026 Form S-8 cited in prior coverage registered 2,000,000 common shares for employee plans; No current moving-average distance, short-interest series, or dated retail-attention sample is available here. Neither crowding nor expanding participation can be established from the September 18 RSI observation.
Catalyst Calendar (next 30 days)
- 2026-09-30 — Scheduled dividend payment. The August 5 release specifies $0.215 per share. The payment does not resolve data-center commercialization. Company announcement.
- ~2026-11-04, est. Q3 results. This remains an unconfirmed estimate outside the next month. The decisive comparison is with the August 5 guidance of $122–130 million revenue and 54–55% non-GAAP gross margin. Revenue below $122 million or that margin below 54% would contradict the operating-recovery case. Published guidance.
What Would Change Our Mind
Failure to hold the $50 round-number boundary would end the proposed stabilization: a weekly close below $50 is the observable price condition. The September 18, 2026 close of $50.39 leaves little separation from that boundary, and the available observations do not establish support there.
Positive resolution requires both a weekly close above the August 21, 2026 reference of $55.18 and Q3 results meeting the August 5 revenue and non-GAAP gross-margin ranges before invalidation. Separately, a named data-center customer with quantified orders and a disclosed revenue schedule would challenge the assessment of the AI narrative; the August 5 call supplied none of those measures.
Correlation Notes
This is a single-name recovery hypothesis. The August 5, 2026 call's reported industrial growth supplies company-specific evidence, but no current matched return series establishes correlation with Navitas, ON Semiconductor, or Monolithic Power Systems. The September 18 price and RSI observations cannot establish a shared gallium-nitride semiconductor recovery, and no group advance is assumed in the case.
Notes
- 2026-06-05: broke -9.71% to ~$77 (AH ~$75.98) in a SECTOR-WIDE semi selloff (Broadcom soft AI/hyperscaler outlook, Samsung strike risk, TSMC stake-sale chatter, rare-earth supply, hot jobs print) — Microchip fell with it. NOT company-specific news.
- GAAP and non-GAAP diverge widely here — Q2 2026 was $0.37 non-GAAP against $0.17 GAAP; check which figure a headline uses.
- Moving-average and RSI feeds on this name have returned inconsistent values; the swing levels and the 52-week high are the more reliable structure.
- Levels quoted in coverage before August 2026 came from an unadjusted price series; levels here reference split/dividend-adjusted closes and differ by a few percent.
- Target aggregators disagree: $77.50 across 5 analysts (stockanalysis.com) versus a $72.67 consensus and a 4 Buy / 3 Hold / 1 Sell mix (MarketBeat), August 2026.
- Insider sales clustered in late May 2026, into the parabolic leg — before the July break, not after it.
- The adjusted 52-week high itself moves with the series: quoted at $87.35 on 2026-08-21 and $86.99 on 2026-09-04.
Related · shared themes
HPE
Hewlett Packard Enterprise Company
Hewlett Packard Enterprise's AI and networking earnings transition supports a repricing thesis that requires intact growth guidance and the first weekly close above $54.44 after the 2026-09-30 Investor Day. A weekly close below $50 invalidates the price thesis.
NTAP
NetApp, Inc.
NetApp's AI storage recovery requires delivery of September guidance before a weekly close below $164 breaks the thesis. Fiscal-Q2 revenue of at least $2.025 billion and non-GAAP gross margin of at least 67.0%, with the annual revenue outlook issued on 2026-09-02 maintained, would confirm the case.
XMTR
Xometry, Inc.
Xometry's AI manufacturing thesis requires enterprise adoption to deliver its August 4 guidance before a weekly close below $85. Third-quarter revenue of at least $234 million and adjusted EBITDA of at least $16 million, with the 33–34% full-year revenue-growth outlook maintained, would confirm the case.
ATEX
Anterix Inc.
Anterix’s utility-spectrum conversion thesis failed its published price test when the 2026-09-18 weekly close of $76.99 breached $85.31. A new utility agreement with disclosed payments and a weekly recovery above $85.31 would support a replacement case.