Semtech Stock Plunges 34% in 3 Months: Should You Buy the Dip?

Semtech shares fell 34% in three months, pressured by weakness in AI and semiconductor stocks, a 5% year-over-year decline in high-end consumer revenue to $39 million, valuation concerns and possible manufacturing-capacity limits in the second half of fiscal 2028. Offsetting those risks, fiscal Q2 2027 data center revenue rose 91% year over year to a record $100 million, while LoRa revenue climbed 58% to $58 million; management projected Q3 data center revenue growth of 160% year over year. Fiscal 2027 and 2028 EPS estimates were revised up 4.9% and 15.8% over the past 30 days, and the article recommends buying the stock.

Analysis

The key underwriting issue is conversion, not demand: backlog and rising 1.6T mix have value only if manufacturing partners deliver qualified capacity on schedule and Semtech turns orders into revenue without margin dilution or working-capital strain. A backlog that is capacity-constrained is not equivalent to secured earnings. LoRa offers a useful non-AI growth stream, but the claimed Amazon Sidewalk opportunity remains distribution potential until device deployments translate into repeatable Semtech revenue.

Near term (days), the drawdown may attract dip-buyers, but SMTC still carries exposure to the same hyperscaler-capex and AI-multiple reset driving sector weakness; estimate revisions are not a valuation floor. Over 1–3 months, watch actual data-center shipment growth versus guidance, backlog conversion, gross-margin trend, and capacity commitments. Over 6–18 months, 3.2T/CPO content gains could widen the opportunity, while delayed capacity or a pause in AI infrastructure spending would defer that upside and leave consumer weakness more visible.

Contrarian point: the bullish case may overstate backlog as a moat. If customers can dual-source or reallocate purchases, unmet demand could benefit competing connectivity suppliers rather than become Semtech’s deferred sales. Conversely, the selloff may be excessive if shipments and margins confirm the ramp; the reported valuation premium alone does not settle that question.

Key Decisions for Investors

• No immediate dip-buy solely on price weakness. Consider a staged SMTC long only after the next report verifies data-center revenue conversion and capacity additions, with position sizing restrained until margin and cash-conversion evidence is available.
• Falsification / risk control: reassess or exit the long thesis if data-center growth materially misses management’s trajectory, backlog coverage converts into delays or cancellations, or margins deteriorate as supply expands. Also monitor hyperscaler capex commentary for a demand reversal.
• Keep MCHP, ON, and NXPI as peer-relative monitors, not automatic hedges: their end-market mixes differ, so a short against SMTC may add unrelated exposure. A broad semiconductor hedge is preferable if the objective is to isolate SMTC execution risk.
• Treat LoRa and Amazon Sidewalk as upside optionality rather than base-case valuation support until Semtech discloses sustained adoption and revenue contribution.