Thesis: Credo Technology (NASDAQ: CRDO) just posted its third consecutive quarter of 200%+ revenue growth, generated $157 million in net income, and sits on $1.29 billion in net cash with zero meaningful debt. The stock is down 59% from its late-2025 peak. Either this is the AI infrastructure trade of the year, or the market knows something Seeking Alpha’s 12 “Strong Buy” analysts don’t.
Let’s figure out which one.
What Credo Actually Does
Credo makes the cables and chips that connect GPUs in AI data centers. Not the GPUs themselves — the stuff that moves data between them.
Specifically: Active Electrical Cables (AECs) and optical DSP chips. Inside a modern AI training cluster, you’ve got 50,000+ GPUs that need to talk to each other at blazing speeds. Traditional copper cables (DAC cables) work fine up to a few meters. Beyond that, they lose signal and burn power. Credo’s AECs solve this with active electronics embedded in the cable itself.
It sounds mundane. It isn’t. When Microsoft and Google are spending $50 billion each building out AI infrastructure, the cables matter as much as the GPUs.
As of March 2026, CRDO’s biggest customers are Microsoft Azure and Google — and the company is already developing 1.6T (terabit-per-second) connectivity for the next generation of AI clusters.
The Numbers That Matter
Q3 FY2026 (quarter ended January 2026):
- Revenue: $407 million — up 201.5% year-over-year, up 52% sequentially
- Gross margin: 68.52%
- Operating margin: 36.76% (was -24% one year ago)
- Net income: $157.14 million
- Diluted EPS: $0.82 (up 412% YoY)
- Free cash flow: $139.73 million for the quarter
The trajectory here is unusual even by AI infrastructure standards. Three quarters ago CRDO was losing money every quarter. Now it’s printing 37% operating margins on $400M+ quarterly revenue. The flip happened fast.
Full Year FY2026 (analyst consensus):
- Revenue: ~$1.36 billion (up ~211% from FY2025’s $436.78 million)
- EPS: ~$3.18
Balance sheet (as of January 2026):
- Cash and short-term investments: $1.30 billion
- Total debt: $12.62 million (essentially zero)
- Net cash position: $1.29 billion
- Shareholders’ equity: $1.85 billion
TTM free cash flow: $283.69 million — up 878% over the prior year
This is a company that went from burning $48 million in free cash flow two years ago to generating $284 million. That’s not a trend, that’s a business model that hit escape velocity.
Why the Stock Got Hammered
CRDO went from $11 in early 2024 to a peak of roughly $140 in late 2025 — a 12x run. The market eventually looked at Q4 guidance and decided growth was slowing.
After Q3 FY2026 earnings (reported early March 2026), management guided for Q4 revenue of approximately $430-440 million — mid-single-digit sequential growth versus the 52% sequential jump in Q3. Analysts had been modeling more aggressive ramps.
Several firms cut their price targets: Susquehanna trimmed from $230 to $170, Rosenblatt went neutral at $125. The stock dropped 14.8% on March 3rd alone on volume nearly 4x average.
The result: from a peak near $140+ in February to $104 today, a 26% drawdown in six weeks.
Here’s the bear case in plain language: CRDO’s explosive growth was driven primarily by Microsoft and Google ramping AI clusters. When those hyperscalers pause spending or shift to in-house silicon, CRDO could see revenue cliff-dive. Customer concentration risk is real — Microsoft reportedly accounts for a disproportionate share of Q3 revenue. And 20x forward revenue is a demanding valuation even with 200% growth, especially when 2027 estimates imply deceleration to ~52%.
The Bull Case
Twelve analysts cover CRDO. Eleven have Buy or Strong Buy ratings. Average price target: $176.36 — 69% above today’s price. Highest target: $240.
The bull case isn’t complicated: AI infrastructure spending is not slowing down. The Iran conflict has complicated energy markets, but hyperscaler capex plans haven’t changed — Microsoft alone guided $80 billion in FY2026 datacenter investment. Every GPU cluster needs connectivity. Credo owns that market at the high end.
A few things bulls point to:
1. FCF is self-funding growth. CRDO raised $551 million in new equity in FY2026 (TTM) but is also generating $284 million in FCF annually. They don’t need capital markets. That’s a significant shift from the pre-2025 era where they burned cash every year.
2. 1.6T roadmap. Current hyperscale deployments run at 800G speeds. The shift to 1.6T is coming in 2026-2027. Credo is building the next-gen product. If they win sockets there, the revenue expansion continues beyond Microsoft and Google.
3. The stock has already been punished. This isn’t priced for perfection. At $104, CRDO trades at roughly 14x the consensus FY2026 revenue of ~$1.36B. That’s still high by traditional metrics but much more reasonable than the 20x+ multiple when the stock was at $140.
4. Net cash gives options. With $1.29 billion in net cash and zero meaningful debt, Credo can buy back shares, acquire complementary technology, or simply ride out a macro downturn without diluting shareholders.
The Risks You Need to Understand
Customer concentration. Microsoft and Google are the engine of this growth story. If either pulls back on AI cluster spending — or develops their own custom connectivity silicon — CRDO’s revenue could decelerate faster than the market expects. This isn’t a theoretical risk; it’s why 10 analysts cut their targets after Q3.
Valuation still isn’t cheap. Even after the selloff, CRDO’s enterprise value is ~$18.5 billion against TTM revenue of ~$1.07 billion. That’s a ~17x EV/Revenue multiple. Astera Labs (ALAB), a direct competitor in AI connectivity, trades at similar multiples. The sector isn’t cheap.
The growth rate will slow. Growing 200% per year is unsustainable. FY2027 consensus revenue estimate is $2.07 billion — that implies 52% growth. Still excellent, but the multiple will compress as growth normalizes. If the stock stays at $104 while revenue catches up, you make money through earnings expansion, not multiple expansion. Slower and lumpier.
Macro headwinds. The Iran conflict has pushed oil prices up and complicated the Fed’s rate trajectory. Higher-for-longer rates make high-multiple growth stocks less attractive. CRDO isn’t immune to that pressure.
Comparable Plays in AI Infrastructure
CRDO doesn’t exist in isolation. The broader AI infrastructure investment theme has driven several similar stories:
Nebius Group (NBIS) is building AI cloud infrastructure in Europe at significant scale — different geography, similar tailwind. Babcock & Wilcox (BW) just signed a $2.4B AI power deal, addressing the electricity bottleneck that Credo’s cables are the data side of. Gorman-Rupp (GRC) has been one of the sleeper picks for data center infrastructure exposure.
The theme is the same: build out the AI data center, and everything from the power to the cooling to the connectivity wins.
Verdict
CRDO is not a simple buy here. It’s a high-quality company with genuinely exceptional fundamentals — 68% gross margins, positive FCF, net cash position, and growth rates that most companies never see in a lifetime. But the valuation still prices in a lot of that excellence.
The best entry was clearly at $97 during the March 3rd flush. At $104, the risk/reward is more balanced. The 12-month analyst consensus target of $176 implies strong upside, but that assumes AI infrastructure spending accelerates again in late 2026 and CRDO wins next-gen sockets.
The scenario where you lose money: Microsoft and Google saturate their current cluster builds, a competitor undercuts on the 1.6T spec, and CRDO’s 2027 revenue comes in at $1.5B instead of $2B. The stock probably revisits $75-80.
The scenario where you make money: AI cluster expansion continues through 2027, Credo’s ZeroFlap and 1.6T products ship on schedule, and earnings hit $4-5 per share by FY2027. At 30x earnings, that’s a $120-150 stock — roughly where it is now. At 40x, you’re at $160-200.
This is a stock for people who have a real view on hyperscaler capex, not a passive hold. If you believe the AI buildout is a multi-year cycle and CRDO maintains competitive position, the selloff from $140 to $104 is an opportunity. If you think we’re in the early innings of hyperscaler digestion, wait for $85.
This is not financial advice. Do your own research before making any investment decisions.