VICR · Vicor Corporation

Current Thesis

The AI-datacenter power-delivery story behind Vicor re-accelerated in late May after looking broken for two weeks. On 2026-05-26 the company lifted Q2 revenue guidance to ~$142M from the prior ~$126M (set on the 2026-04-21 Q1 call), crediting higher product revenue plus royalties from a newly signed all-inclusive licensee of its power-system IP portfolio (Factorized Power + Vertical Power Delivery). Management flagged a book-to-bill above 2.0 and a one-year backlog near $300.6M (~70% sequential growth). Shares spiked as much as ~24% intraday to an all-time high on the news. After a sharp early-June pullback to $271.04 (2026-06-05) the stock recovered to ~$306 (2026-06-12), reclaiming the $300 shelf it lost during the May blowoff. The fundamental ramp is accelerating; the open question is whether a name up >600% off its 2025 low, trading above sell-side fair value with relentless insider selling, has fuel for another leg before a fresh price-target-revision cycle prints.

Bull Case

• Second guide raise in five weeks: Q2 lifted to ~$142M (2026-05-26) from ~$126M (2026-04-21) a sequential step-up, the cycle's third upward revision against the April FY26 ~$570M frame.
• Book-to-bill above 2.0 (2026-05-26): orders running at >2x shipments, the cleanest forward-demand signal in the package.
• New IP licensee (2026-05-26): an all-inclusive license to the patent portfolio (power-converter topologies, control systems, Vertical Power Delivery) adds recurring royalty income at near-zero incremental cost structurally margin-accretive.
• Backlog ~$300.6M, ~70% sequential growth (reported 2026-05-26) order book, not pipeline hope.
• Margin mix inflection: Q1 gross margin ~55% ($62.4M) on 2026-04-21 vs ~47% a year prior, as VPD plus royalty mix climbs.
• Vertical Power Delivery moat: VPD enables >1000A delivery to next-gen accelerators; the second-gen ramp for a lead hyperscale customer in 2026 deepens design-in lock-in.
• Structure repaired: price reclaimed $300 (2026-06-12, ~$306) after defending the $271.04 low (2026-06-05), undoing the May break.

Bear Case

• Insider distribution has not stopped: 192 open-market sales / 0 buys across six months; CEO Patrizio Vinciarelli alone ~76 sales / 380,821 sh / ~$61.71M, with a further ~20,000-share CEO sale under a 10b5-1 plan in recent Form 4 filings. No insider is buying these levels.
• Trades above consensus fair value: sell-side targets cluster near ~$282.50 (bull ~$325, bear ~$217.50) versus ~$306 spot the stock has priced in the raise ahead of the analysts, so the next leg depends on revisions that have not yet printed.
• A blowoff top that recovered rather than based: from a $361.89 all-time high the stock fell to $271 — then bounced ~13% in a week; MACD turned negative on 2026-06-05. Volatile chop, not a clean consolidation.
• Customer / royalty concentration: heavy reliance on a single lead AI customer and a concentrated licensing base; one order cut or design loss is a 20-30% air-pocket.
• Crowded field: SemiAnalysis (2026) frames AI power delivery as a contested fight among MPS, Infineon, ADI, Renesas and Delta share gains are not guaranteed as the dollars grow.
• Thin liquidity: spreads have widened on volatility in prior episodes, raising slippage risk on any fast move.

Setup & Price Structure

• 52-week range $41.76 $361.89; last ~$306 (2026-06-12, intraday high $310.11 / low $292.77), market cap ~$13.96B.
• Up >600% off the 2025 low; currently ~15% below the $361.89 all-time high.
• The $300 level is the line that matters: lost in the May blowoff, reclaimed 2026-06-12. Holding it on a weekly-close basis keeps the recovery intact; losing it reopens the $271.04 low (2026-06-05).
• Primary trend remains above the 200-day. RSI cooled from ~95 at the May peak through the pullback and is rebuilding with the bounce mid-range now, neither stretched nor signalling a fresh breakout.
• A clean continuation looks like a higher low above $300 and a push through the $310 intraday shelf toward the prior high; failure to hold $300 turns the move back into a relief bounce inside a topping pattern.

Catalyst Calendar (next 30 days)

• No binary catalyst inside the 30-day window (through ~2026-07-13). The guide-raise catalyst already fired on 2026-05-26.
• Q2 FY2026 earnings: ~2026-07-21 to 2026-07-23 (est., just outside the window) the next binary. The print must validate the ~$142M guide and a sustained book-to-bill above 1.0.
• Analyst price-target revisions (no fixed date): with spot above the ~$282.50 consensus, an upward revision wave following the 2026-05-26 raise is the most likely near-term fuel.
• Ongoing Form 4 insider filings (rolling): any pause in CEO/VP selling or a first open-market purchase would be a tape-changing signal.

What Would Change Our Mind

• A weekly close back below the reclaimed ~$300 pivot forfeits the 2026-05-26 breakout and reopens the $271 low, flipping the recovery into a failed bounce.
• A Q2 print (~late July) with revenue below the raised ~$142M guide, or book-to-bill falling back under 1.0, breaks the acceleration thesis.
• A lead-customer order cut, design loss, or backlog reversal (the ~$300.6M figure rolling over) is a structural break worth a 20-30% repricing.
• On the upside: a higher low above $300 plus a clean break of $310 alongside upward target revisions would re-confirm accelerating-leg behavior and justify pressing the move.

Correlation Notes

• Trades with the AI-datacenter power-delivery cluster MPS, Infineon, ADI, Renesas, Delta (SemiAnalysis, 2026). Peer breakouts confirm the theme; a peer warning would hit VICR first given its premium multiple.
• Second-order linkage to hyperscale AI capex and the GPU/accelerator cycle; a hyperscaler capex-cut headline is a direct demand risk to the backlog.
• High beta to the broad semiconductor tape (SOX) and to risk-on/risk-off rotation, amplified by the rich valuation and thin float.
• Royalty income partly decouples revenue from unit shipments, dampening though not removing sensitivity to any single quarter's hardware demand.