Credo is worth buying on its post-earnings dip. x.com 

Credo's 45% drop from its high was about customer concentration and valuation, not demand, making it a great long-term buy. The latest quarter beat on revenue and guidance: revenue grew 115% to $479M, a seventh straight quarter of triple-digit growth, and management guided to more than 85% growth for the fiscal year. Optical revenue is targeted to pass $600M this year, on top of the AEC business five hyperscalers already buy. x.com 

Credo is buying its re-acceleration at a flat gross margin, and at 36 times forward earnings, growth without margin is worth less. Q2 revenue guidance of $525-535M topped the $514.7M consensus, but non-GAAP gross margin went 68.6%, 68.3%, 68.0% and is guided at 67-69% again, while the beat over estimates shrank from 41% to 0.8%. Three customers are 77% of revenue. Credo sells copper, and Marvell expects scale-up optics from 2027. A print above 69% on $530M would prove this too cautious. x.com  … Show more

Credo's growth rides surging AI interconnect demand, and management expects revenue to grow more than 85% in FY27. F1Q27 revenue rose 115% to $479 million, EPS of $1.20 slightly beat, and F2Q27 guidance is $525-535 million. Active electrical cables remain the near-term engine as adoption broadens across hyperscalers and neoclouds; FY27 optical revenue is still expected above $600 million, though a higher optical mix dilutes gross margin. The $227 target rests on 20x FY28 EPS of $11.40. x.com 

Credo's near-total share of AI data-center cables and unchanged estimates make its lower multiple look like a fear-driven opportunity. Its stock went from 27x to 20x forward earnings while the supply chain building those systems guided 2027 revenue to double. Buyers this week will probably be very happy a year from now. x.com 

Credo is a trade rather than a long-term hold because co-packaged optics is the technology that closes the copper-to-optics gap its retimers and active electrical cables fill. If CPO ramps faster than expected in the 2027 architecture cycle, those products get designed out. That is not a 2026 problem, but it is why a decelerating hyper-growth chipmaker can't keep a 15x sales multiple forever. Customer concentration, which hit the stock before, adds risk, and Credo fits no AI endgame theme. x.com  … Show more

Credo's selloff on a raise reflects impatience, since its next growth leg is a fiscal 2028 story that begins in May 2027. It printed $479M, guided $525-535M and lifted full-year growth to more than 85%, yet sold off for a second straight print. CEO Bill Brennan called fiscal 2027 just a stepping stone for optics: OmniConnect, micro-LED cables and near-package optics turn on in fiscal 2028, when AECs step up to 1.6T. x.com  x.com 

Credo's post-earnings selloff is overdone, reflecting normalizing hypergrowth and first-quarter DustPhotonics costs rather than any demand break. Q1 revenue rose 115% to $479M, its seventh straight triple-digit quarter, Q2 guidance of $525-535M beat the Street, and FY27 optical revenue is targeted at $600M+ as optics outgrows AECs. Management sees thousands of dollars of Credo content per GPU. Risks include customer concentration, Broadcom and Marvell in optical DSPs, and new AEC rivals. x.com  … Show more

Credo's results were underwhelming for a company that added $32 billion in market value in a few months, and shorts may punish the stock. Per @WisemanCap, fiscal Q4 EPS of $1.16 beat the $1.02 estimate and revenue of $437 million was just above estimates, while Q1 fiscal 2027 revenue guidance of $465-476 million was roughly in line with the $464.7 million consensus. x.com  x.com 

Credo has the widest gap between fundamentals and share price among optics names, but also the worst chart, so it merits a staged, half-size entry rather than a full one. The optics selloff is macro, not fundamental: markets priced a 92-93% chance of a 25bp Fed hike, the 10-year yield hit 5.04%, the highest since 2007, and chip stocks fell 5% after Dario Amodei called for slowing frontier AI. Rate moves cut optical multiples before the rest of AI hardware, and they are still compressing. x.com