FEATURED

CBRS Is Down ~50% From Its Post-IPO High

The stock began trading on May 14 after pricing its IPO at $185. It opened at $350, hit an all-time intraday high of $386.34, then closed its first day at $311.07. It is now trading in the $182 to $191 range. That is not a dip. That is a near-halving in six weeks from one of the largest U.S. tech IPOs since Uber went public in 2019.

And yet, every major analyst who covers it still has a Buy rating. All ten of them.

What is interesting is that both of those facts can be true at the same time. The tension between them is exactly what makes Cerebras Systems (CBRS) worth understanding right now, not just as a story, but as an options structure.

What the Earnings Actually Said

Cerebras reported Q1 FY2026 results on June 23 for its first earnings release as a public company. Core revenue came in at $191.3 million, up 92% year over year. GAAP revenue was $193.4 million. Cloud and other services revenue grew 178% year over year to $82.8 million. Hardware revenue rose 60% to $111.6 million. Core gross margin for the quarter was 46.5%, up from 42.1% a year ago. Core operating loss narrowed to $3.5 million from $19.3 million, essentially breakeven on a core operating basis.

The company ended the quarter with $3.3 billion in cash. Full-year FY2026 core revenue guidance of $855 to $865 million came in ahead of the Wall Street consensus of roughly $828 million going into the report. Q2 core revenue guidance of approximately $194 million also beat consensus of about $178 million.

On paper, that is a strong quarter. The stock still sold off sharply. The part people keep skipping over is why.

The Number That Broke the Session

Q2 core gross margin guidance: 36% to 38%. Full-year core gross margin: 38% to 41%. That compares to 46.5% in Q1. The compression from Q1 to Q2 is roughly 1,000 basis points in a single quarter. For comparison, Nvidia operates in the low-70% gross margin range. The gap is not new, but the guided step-down made it more visible than before.

The explanation management offered: a temporary mix and capacity issue tied to meeting near-term demand while the company expands its data center footprint. CEO Andrew Feldman said on CNBC that investors “misunderstood” the margin guidance and that the dynamic is temporary. Mizuho analyst Vijay Rakesh noted in his post-earnings commentary that gross margins could temporarily dip into the low-30% range through FY2026 as a leaseback dynamic plays out, before recovering as Cerebras builds out its own facilities. His long-term target for the Cloud and Services segment is 60% gross margin in FY2027.

That may be correct. The problem is that at a high multiple, there is no cushion for a “temporary” anything.

The Structural Picture Underneath the Margin Story

This is where it gets complicated. Customer concentration is real and quantifiable. Cerebras has described a $24.6 billion contracted backlog, with the OpenAI deal accounting for approximately 80% of that figure. The OpenAI agreement is valued at more than $20 billion for 750 megawatts of compute, announced in January and set to expire in 2028. AWS announced a partnership in March to deploy Cerebras chips in its data centers, but management confirmed on the earnings call that meaningful revenue contribution from the AWS deal is not expected until 2027.

Slight tangent, but it matters: Cerebras’ lock-up structure is staggered rather than standard. About 28 million Class A shares held by directors, officers, and nonemployee shareholders became available to trade on the second trading day after the June 23 earnings announcement, per the company’s prospectus. That placed the first unlock window around June 25. The full 180-day insider lock-up does not expire until mid-November 2026. That is a meaningful overhang that does not go away quickly.

And the valuation math. At roughly $185 to $191 and a market cap in the $46 to $48 billion range, CBRS is still priced at approximately 54x current-year revenue guidance at the midpoint of $860 million. That is an extreme multiple for a company that just showed its first major public-market margin compression in guidance.

Options Market Structure After the Selloff

Going into the June 23 earnings, options on CBRS were implying a 12.8% move in share price post-earnings. The actual move landed well beyond that. Front-month implied volatility was elevated for a stock that had only been public for six weeks, and the realized move through the following session exceeded what was priced.

Post-earnings, notable put activity further out the curve reflected traders pricing in a more extended downside path. With IV still elevated in the near term, premium-selling structures can look attractive on paper, but the lock-up overhang and binary customer dynamics make that a higher-risk posture than it appears at first glance.

Analysts largely held or raised price targets after the quarter. Wedbush raised its target to $280 from $270. Morgan Stanley raised its target to $273 from $250. The consensus average across ten analysts now sits around $294 to $299, with a high estimate of $340 and a low of $250. All ten analysts maintain buy-equivalent ratings. If the margin compression story deepens into Q2, those targets come down.

Structured Trade Framework

Bull Case: The margin headwind is genuinely temporary. OpenAI ramps through H2 2026, the AWS relationship begins contributing in 2027 as guided, and customer diversification accelerates beyond the current OpenAI concentration. For traders expecting a recovery toward analyst targets, a defined-risk call spread in the October or November expiry, buying a call at current levels and selling one 30 to 40% higher, captures that upside with limited premium at risk in an elevated-IV environment.

Bear Case: The demand and capacity mix issues persist longer than advertised. Gross margin stays in the low-30s through most of FY2026. Customer diversification fails to materialize fast enough to offset OpenAI concentration. Put positioning further out the curve reflects this scenario. A bear put spread in the August expiry frames the downside with defined risk.

Neutral Case: The stock consolidates in the $180 to $220 range as the market waits for Q2 results, scheduled for September 2, 2026. With IV still elevated post-earnings, premium-selling structures can work if realized volatility collapses. But this carries meaningful risk given the lock-up overhang, the 180-day full expiry in mid-November, and the binary nature of a backlog that is 80% concentrated in one customer.

What to Watch Between Now and the September Report

• OpenAI revenue recognition timing: the $20B+ deal is central to the bull case and the backlog math
• Lock-up share supply dynamics through late June and into mid-November, when the full 180-day window expires
• Data center build-out progress: management flagged capacity as the main constraint, not demand
• Q2 gross margin actuals vs. the 36 to 38% guide: any upside to guidance would matter to market sentiment
• AWS revenue commentary: meaningful contribution not expected until 2027, but any pull-forward would be significant
• Customer diversification signals: any new customer announcements beyond the OpenAI and UAE-affiliated accounts

The technology is real. The wafer-scale architecture does what management says it does. At 15x the inference speed of leading GPU-based solutions on benchmark open-source models, Cerebras has a genuine performance edge in the market it is targeting. The question has never been about the chip. It is about whether a company with heavy customer concentration can diversify fast enough to justify a 54x revenue multiple before the market loses patience. And whether margin compression is a three-quarter story or a three-year story.

The first real answer comes September 2.