MCHP · Microchip Technology Incorporated

Current Thesis

A beaten-down broad-line analog/MCU name being re-rated on a freshly disclosed, accelerating AI-data-center revenue leg layered on a cyclical MCU recovery. On 2026-06-01 management quantified the Data Center Solutions (DCS) unit for the first time: $302.7M in CY2025, guided ~$500M in CY2026 (+65%), with the Mar-2026 quarter already +62.9% YoY reported acceleration, not a promise. The tape's reaction tells the harder story: an +8% after-hours pop on 06-01 and +5.9% on 06-02 to ~$97 has fully round-tripped to ~$87.83 by 06-06, parking price on the 50-day EMA (~$86.47). So the read now is a make-or-break breakout retest, not a runaway. The DCS unit is only a high-single-digit slice of revenue; the broader data-center+compute end market is ~18%; the remaining ~80% rides the industrial/auto MCU upcycle off a trough.

Bull Case

• DCS unit +62.9% YoY (Mar-2026 qtr), $302.7M CY2025 → ~$500M CY2026 (+65%), disclosed 2026-06-01. The product stack storage controllers/expanders, PCIe + CXL memory controllers, and Switchtec PCIe switches/retimers is direct AI-rack content that wins independent of which GPU vendor takes share.
• Pricing power surfacing: on ~2026-06-02 management announced selective price increases across the portfolio citing input-cost pressure, explicitly stating it does not change guidance for the quarter ending 2026-06-30 a margin/demand tell into a recovering book.
• Second AI-data-center attach point: 2026-05-26 launch of 3.3kV HV-D3 mSiC power modules for solid-state transformers in AI hyperscale facilities power content stacked on connectivity content.
• Cyclical MCU recovery underneath: DCS growth lands on a base coming off the 2024–25 inventory-correction trough, so operating leverage builds as utilization normalizes rather than fighting a declining core.
• Headroom to consensus: avg 12-mo PT $113.24 (high $135, low $75) vs ~$88 spot → ~29% to the average before the long is crowded.
• Tail overhang removed: 2026-06-04 U.S. Commerce export-license authorization clears a regulatory question on advanced-tech shipments.

Bear Case

• The catalyst pop is already gone. The +8% guidance move (06-01) round-tripped to ~$87.83 by 06-06; buyers who chased ~$97 are underwater and the breakout is unconfirmed. Momentum that fails to hold its own news is a warning, not a dip to buy blindly.
• DCS is a slice, not the company. The high-growth unit is ~$303M against a multi-billion revenue base (high-single-digit%); a 65% rip there moves consolidated numbers modestly. The ~80% MCU/industrial/auto core can swamp the DCS win on any macro wobble.
• Late-attention marker: 2026-06-03 leveraged single-stock/sector ETFs launched on MCHP/NXP/ON ("look beyond Nvidia"). Leverage products appear when retail wants the theme an early-saturation flag.
• Valuation rich on trough earnings: GuruFocus GF-Value ~$57.57 vs ~$88 multiple is carrying the move; air below if the MCU recovery stalls.
• No near-term catalyst. Next print is ~2026-08-06, two months out. Nothing scheduled in the next 30 days re-feeds the move; the guidance is banked.

Setup & Price Structure

• Spot ~$87.83 (2026-06-06 range $87.38–$95.10), market cap ~$47.75B. The 06-01/06-02 guidance pop to ~$97 has been given back nearly in full.
• Sitting on the 50-day EMA (~$86.47); 200-day SMA (~$73.97) well below primary uptrend intact, but price is testing the lower edge of the ~$86–90 breakout shelf rather than extending.
• 52-wk range $48.52–$105.91: ~$88 is ~17% off the high, ~+81% off the low most of the cyclical re-rating is already in the tape.
• RSI has cooled to neutral after the fade; no overbought blow-off, but also no momentum thrust a digestion/retest tape. Reclaim of ~$95 re-arms the breakout; a weekly close under ~$86 fails it.

Catalyst Calendar (next 30 days)

• No earnings in window FY27 Q1 print is ~2026-08-06 (outside 30d). Any entry inside the 3 trading days prior is an avoid.
• ~2026-06-09 → ~2026-07-07: watch for sell-side PT revisions/upgrades clustering off the 06-01 DCS disclosure analyst confirmation of the narrative would re-arm momentum; silence/cuts would confirm the fade.
• Price-increase rollout (announced ~2026-06-02): monitor for distributor pre-buy or order commentary; selective increases take effect across the portfolio without changing the Jun-30 quarter guide.
• DCS design-win / customer press releases: any named hyperscaler/CXL/Switchtec traction inside the window is a narrative-acceleration tell.
• No ex-dividend or PDUFA-type binary in the 30-day window.

What Would Change Our Mind

• Bullish confirmation: a weekly close back above ~$95 reclaiming the breakout shelf, ideally with peer (NXP/ON) cluster strength and clustered analyst PT raises.
• Thesis break (exit the read): a weekly close below ~$86 (50-day EMA) fails the post-guidance breakout; or at the ~2026-08-06 print, DCS CY2026 guide cut below ~$500M or YoY growth slipping under 50% kills the re-rating leg.
• Core-cycle break: MCU/industrial book-to-bill back under 1.0 or renewed distributor inventory builds the ~80% base rolling over swamps DCS.
• Saturation flip: more leveraged-ETF/retail crowding paired with NXP/ON rolling over → theme to SATURATED, no fresh thesis.

Correlation Notes

• Cluster basket: NXP + ON Semi trade together with MCHP on the "beyond-Nvidia" AI-infra narrative the 2026-06-03 leveraged ETFs explicitly bundle the three; stacking the basket is one correlated bet, not three.
• Broad semis beta: SOXX/SMH; analog/MCU peers TXN, ADI, STM read across on the cyclical-recovery leg.
• Data-center connectivity competitive read-through: Astera Labs (ALAB), Marvell (MRVL), Broadcom (AVGO) in the PCIe/CXL/retimer space their order/commentary moves the Switchtec + CXL part of the DCS thesis directly.
• Leverage products amplify single-name moves both ways; the retest of $86 is the level the basket trades around.

Notes

DCS unit (~$303M CY2025) is a high-single-digit slice of revenue; the broader data-center+compute end market is ~18%. The AI-data-center story is a re-rating kicker on part of the company, not the whole size with that asymmetry. Earnings blackout: next print ~2026-08-06 (FY27 Q1); avoid any entry inside the 3 trading days prior. The guidance pop has round-tripped to the 50-day EMA treat ~$86 as the make-or-break level, ~$95 as the reclaim trigger.