Earnings call transcript: Credo Technology Q4 2026 earnings beat forecasts

Credo reported a strong Q4/FY2026 beat, with revenue of $437 million and EPS of $1.16 both ahead of expectations, while FY2026 revenue topped $1.3 billion, up 206% year over year. Management guided Q1 FY2027 revenue to $465 million-$475 million and reiterated a larger FY2027 ramp driven by optical products, with more than $600 million of optical revenue expected across DSPs, ZeroFlap Optics, and silicon photonics. Shares rose 1.13% after hours as the company highlighted robust AI infrastructure demand, though supply-chain tightness and customer concentration remain key watch points.

Analysis

The market is still treating this as a simple “beat-and-raise,” but the real signal is that CRDO is moving from point-solution vendor to architectural toll collector. If management is right, the mix shift into ZeroFlap and silicon photonics creates a step-function in content per node, which means the company can grow into the same cluster build multiple times as customers migrate from 100G to 200G lanes and from copper to optical. That gives CRDO a rarer profile: demand visibility that is less about unit volume and more about increasing dollars per deployment.

Second-order winners are the supply-chain enablers and adjacent interconnect vendors that can secure capacity in 5nm/3nm and photonics earlier than peers. The flip side is pressure on slower-moving transceiver and module vendors whose products look increasingly commoditized when network reliability and power efficiency become the buying criteria. The commentary also implies that neo-clouds are becoming a meaningful end market, which matters because these customers are faster adopters and less tolerant of multi-quarter design delays — that should compress the gap between design win and revenue ramp, but also increase quarter-to-quarter lumpiness.

The biggest risk is not execution on the current copper ramp; it’s timing slippage on the optical hockey stick and whether the 200G transition arrives fast enough to justify the current multiple before the narrative shifts again. If the industry delays 1.6T adoption into late FY27/FY28, the stock is vulnerable to multiple compression even if fundamentals remain strong, because expectations are now heavily loaded into second-half acceleration. In other words, the near-term upside is more about guidance credibility than outright earnings beats; any disappointment on supply, customer timing, or optical attach rates could re-rate the name quickly.

Consensus appears to be underestimating how much of the value is now driven by portfolio expansion rather than the core AEC franchise. The contrarian setup is that the market may be overpricing the near-term optical ramp while underpricing the durability of copper share gains and the optionality from new product categories like Weaver. That suggests the right way to own it is not as a pure momentum name, but as a basket of staged catalysts across FY27/FY28 with clear timing risk.

Key Decisions for Investors

• Stay long CRDO, but use strength to trim into the post-earnings pop; add only on pullbacks toward the 20-day moving average. Risk/reward is best if the market gives back premium while preserving FY27/FY28 optical optionality.
• Buy CRDO Jan-2027 call spreads to capture the second-half FY27 optical inflection while capping theta burn. Prefer structures that target the next two earnings dates rather than outright stock given elevated expectations.
• Pair trade: long CRDO / short a slower-moving optical component or transceiver peer with less vertical integration and weaker telemetry story. Thesis: CRDO’s mix shift should compress the valuation gap as reliability becomes the buying criterion.
• Watch for a tactical hedge via short-dated CRDO puts if channel checks show 1.6T or 200G-per-lane timing slips by more than one quarter. The stock’s setup is vulnerable to any delay because the multiple is already discounting a clean ramp.
• Over the next 3-6 months, accumulate exposure to names tied to AI connectivity capex and capacity bottlenecks rather than pure compute. CRDO is the clearest expression, but the second-order beneficiaries are the foundry, packaging, and photonics supply chain.