Credo's Growth Explosion The Market Is Ignoring

Credo reported Q3 FY2026 revenue of $407.0M, up 201.5% YoY and 52% sequentially, driven by accelerating hyperscale AI infrastructure demand. The company guides FY2026 revenue of $1.33B (≈204% growth) backed by hyperscaler adoption of its Active Electrical Cable connectivity, while posting unusually strong profitability for a semiconductor supplier: 68.6% gross margin, 49.6% operating margin and $139.7M in quarterly free cash flow.

Analysis

Credo’s disclosure should be read as a structural signal from hyperscalers, not a one-off beat: hyperscale buyers are de-risking architectures that favor Active Electrical Cable (AEC) silicon, which changes the value chain economics for short-reach interconnects. That tilt favors firms that supply cable assemblies, connector mechanics, and the OSATs that package high-speed PHYs, and creates a pathway for higher ASPs per link versus legacy passive copper while compressing the addressable volume for short-reach optics over the next 12–36 months.

A likely second-order supply-chain effect is a reallocation of capex and factory capacity toward cable-assembly automation and high-density PCB/substrate runs; expect lead times and pricing power to move from commodity cable houses to a smaller set of specialized assemblers and substrate vendors. Concurrently, large silicon system vendors can internalize PHY functionality or pursue tuck-in M&A to capture margin, which would be structurally negative for pure-play connectivity ASIC specialists if it occurs within 12–24 months.

Near-term catalysts to watch are hyperscaler design-win cadence and OSAT capacity fill rates — both will drive whether ASPs and gross margins for connectivity stay elevated or revert as competitors scale. The main fragility is customer concentration: a pause or inventory digestion by a single hyperscaler would transmit quickly through orderbooks and could halve growth expectations within a quarter, while a successful standardization by multiple hyperscalers would extend the runway for years.

From a positioning perspective, the market is early in repricing the supplier stack beyond the obvious GPU beneficiaries; we should shift weight toward companies with direct exposure to cable assembly economics and substrate scarcity while hedging against rapid vertical integration by large silicon vendors.

Key Decisions for Investors

• Long NVDA (6–12 month calls) — directional exposure to AI infrastructure demand with limited downside to option premium; target entry on a pullback of 10–15% or on confirmation of further hyperscaler spend signals. Risk/reward: pay premium vs potential asymmetric upside if GPU-led buildout continues (approx 3:1 if rally >30%).
• Overweight AVGO (6–12 months) — buy the equity or out-of-the-money calls on a 3–6% pullback; Broadcom-style diversified silicon + infrastructure exposure acts as insurance if pure-play connectivity margins compress. Risk: diversification cushions downside; set 10% trailing stop.
• Pair trade: Long APH or TEL (6–12 months) / Short LITE (3–6 months) — long cable/connector/assembly exposure and short short-reach optics supplier to capture a migration to AEC. Risk/reward: aim for net 1.5–2x upside if AEC adoption accelerates; cap downside on short leg via bought calls.
• Buy LITE puts or equivalent downside protection (3–6 months) — tactical hedge against near-term volume loss for optics vendors while inventory digestion risk is high. Risk: optics rebound if technical constraints favor optical links; limit position size to 1–2% of portfolio.