The numbers tell two different stories, and the tape tells a third. That’s the first thing you notice when you sit down with Credo Technology Group Holding’s latest run. On one screen, you have a stock, CRDO, that closed at $213.15 on July 24th. That’s a 29.5% plunge from its high just five weeks prior, a full re-rating in the span of a summer month. On another screen, you have the forward estimates. Over that same brutal period, as the stock shed another 21.6%, the consensus figures for next-twelve-month revenue, EBIT, earnings, and levered free cash flow all finished slightly higher than where they started.
This isn’t just noise. It’s a fundamental disagreement between the market’s mood and the company’s projected math. The selloff had a clear sector catalyst. Chip stocks led a Nasdaq slide that day, and Credo, with its five-year beta of 3.23, is engineered to amplify sector moves. But the magnitude was its own. You can see the compression in a single line: the forward price-to-earnings multiple collapsed from around 44 times to about 35 times. The forward enterprise value-to-revenue multiple dropped from roughly 20 times to 16 times. All of this happened on a set of financial estimates that did not decline. The forward EBITDA consensus held firm at about $1.3 billion on both dates. Even the sell side held the line; Barclays and Susquehanna actually raised their price targets during the stock’s descent. When the estimates and the targets point one way and the tape points another, you have a puzzle worth solving.
The fear fueling July’s semiconductor selloff is a macro one: that the breakneck pace of AI capital spending might stall. But in a conversation at the Bank of America Global Technology Conference in early June, Credo’s management addressed a more specific, supply-side constraint. CEO Bill Brennan zeroed in on the 3-nanometer silicon node. He noted that every 200-gigabit-per-lane, 1.6-terabit solution he’s aware of is built on this advanced node, as power consumption rules out older technology. This, he suggested, could ultimately cap deployments if buyers aren’t strategic. For Credo, however, Brennan expressed confidence. “I feel comfortable that we’ve underpinned through ’27 based on our growth trajectory,” he said, reasoning that the company’s connectivity chips are physically small next to the GPUs and switches they enable, and that shorting them would stall entire AI cluster deployments. He also outlined a potential workaround, where server sleds using eight 100-gigabit lanes per port could achieve the same 1.6-terabit throughput without immediately jumping to the tighter 200-gigabit lane. These are expectations, not committed capacity orders, but they frame management’s operational mindset.
Where does this leave Credo in the competitive landscape? On a forward earnings basis, its roughly 35x multiple sits in a intriguing pocket between Analog Devices at around 27x and Marvell Technology at around 43x. It’s below Advanced Micro Devices near 59x and, notably, above Nvidia at around 21x. Flip to a revenue multiple, and the ranking inverts: Credo’s roughly 16x tops AMD’s 15x, Marvell’s 14x, and Nvidia’s 11x. The reconciliation between these two valuation views lies in margin. Credo’s fiscal 2026 non-GAAP net income margin landed at 49.5%. Essentially, half of every revenue dollar made it to the bottom line. When CFO Daniel Fleming was asked at that same conference if Credo was investing enough, he pointed to the underlying efficiency: “the growth rate of revenue was 1.5x that of OpEx.”
- Financial estimates have not declined.
- Forward price-to-earnings multiple collapsed from 44 times to 35 times.
- Forward enterprise value-to-revenue multiple dropped from 20 times to 16 times.
- Effective strategic positioning for 3-nanometer supply.
- Confidence in growth trajectory till 2027.
- CFO highlighted efficiency ratios.
| Metric | Value |
|---|---|
| Forward Price-to-Earnings Multiple | 35x |
| Forward Enterprise Value-to-Revenue Multiple | 16x |
| Fiscal 2026 Non-GAAP Net Income Margin | 49.5% |
| Projected Share Price by April 2031 | $670 |
| Annualized Internal Rate of Return | 27% |
| Low-case Revenue Growth | 26% |
A detailed financial model built on institutional data points to the stakes of this divergence. A mid-case scenario suggests a potential path to roughly $670 per share by April 2031, implying a total return of about 213%, or an annualized internal rate of return near 27%. This projection rests on two core drivers over a decade-long window. First, that volumes for Credo’s active electrical cables scale directly with the expansion of AI cluster sizes. Second, that the company’s positioning for 3-nanometer supply holds, preventing a hard cap on shipments from wafer availability. The margin thesis, meanwhile, is underpinned by vertical integration via the DustPhotonics acquisition. Brennan highlighted that this allows Credo to build optical transceivers using its own components instead of purchasing an $80 digital signal processor from an outside supplier. The primary risk is concentration; a small group of customers drives the majority of revenue. A low-case model assumes slightly slower revenue growth near 26% annually and a margin settling toward 47%, with gross margin drifting toward the guided floor of 67%. It’s worth noting that in this model, a shift in product mix toward lower-margin optics could break the margin assumption long before growth itself faltered.
All of this sets the stage for a clarifying moment. Credo is scheduled to report its first-quarter fiscal 2027 results after the close on September 9th. Two lines from that report will settle the current disagreement. Revenue at or above the $475 million top end of guidance, coupled with a non-GAAP gross margin holding near the 68.3% posted last quarter, would signal that the estimates and the analyst targets were correct, and the market’s July selloff was premature. A print at the low end of the range, with margin sliding toward the 67% floor, would indicate the market read the quarter’s challenges before the analysts did. In the end, the only way to truly know where you stand is to examine the numbers yourself, with the same depth and tools that inform the professional analysis. It’s the difference between watching the story unfold and understanding the plot.
