What Should Investors Do With the New Wave of IPOs?

Monday.com reported Q1 2026 revenue of $351 million, up 24% year over year, and adjusted EPS of $1.15 versus $0.93 expected, while raising full-year 2026 revenue guidance to about $1.5 billion. Cerebras also drew attention after increasing its IPO range to $150-$160 per share, implying a roughly $50 billion valuation and nearly $5 billion in proceeds, though the discussion emphasized customer concentration and execution risk. The episode frames Monday.com as showing better-than-feared AI resilience and Cerebras as a highly speculative, heavily hyped IPO.

Analysis

The key read-through is that the market is no longer treating AI software as one homogeneous bucket. monday.com’s quarter suggests investors are willing to re-rate names where AI is an add-on to sticky enterprise workflows, but only if the company can show retention and large-account expansion are offsetting slower SMB demand. That matters for adjacent workflow and collaboration software: the “AI will commoditize you” bear case is losing force in the near term, but only for vendors with enough installed base to monetize usage-based pricing and enough operational discipline to expand margins without hiring ahead of demand.

The more important second-order effect is on AI infrastructure and hyperscaler spending. Cerebras’ reception implies capital is still available for differentiated compute architectures, but the customer concentration profile means the market is paying up for narrative more than diversification. If Cerebras wins even modest share, the pressure falls on incumbent GPU economics at the margin; if it doesn’t, the post-IPO unwind could quickly spill into smaller AI hardware and adjacent software names as investors reprice “AI alternatives” from scarcity premium to execution risk.

The contrarian takeaway is that the current enthusiasm is probably better read as a sentiment reset than a full thesis inflection. For monday.com, the stock can keep working for months if management sustains RPO growth and prevents margin slippage, but the slower guide means the easy part of the rerating may already be done. For Cerebras, the setup looks structurally similar to other heavily oversubscribed AI listings: first-day action can stay strong, but the better entry is likely after the market forces a valuation reset against actual disclosed customer breadth and backlog conversion.

The broader IPO-flows angle is that a large, attention-grabbing listing can temporarily drain capital from second-tier AI beneficiaries. That creates a tactical opportunity to own better-quality software or infrastructure names on weakness if they are sold mechanically to fund exposure to the new issue. The key is to separate true fundamental degradation from simple portfolio rebalancing pressure, which can persist for several weeks after deal launch.

Key Decisions for Investors

• Long MNDY into the next 1-2 quarters on weakness, but only if the market gives back the post-earnings gap; target a tactical 10-15% upside from here with a tight stop if next-quarter growth decelerates below the guided band.
• Avoid chasing Cerebras on IPO day; wait 30-90 days for lockup overhang, trading volatility, and disclosure on customer diversification before initiating. If the stock de-rates 25-35% from IPO pricing while revenue growth remains >50%, that becomes a more attractive speculative entry.
• Pair trade: long quality workflow software vs short AI hype basket. Use a basket long in names with sticky enterprise retention and visible margin expansion, against a short in highly promoted AI infrastructure/adjacent names that trade at extreme sales multiples with customer concentration risk.
• If owning NVDA, treat Cerebras as a monitoring catalyst rather than a thesis breaker. The right posture is to use any post-IPO weakness in NVDA as a buy-the-dip event only if competitive commentary shows share loss risk is contained; otherwise hedge with short-dated calls on smaller AI hardware peers.
• Watch AMZN as the stealth beneficiary: every new compute alternative increases hyperscaler bargaining power and broadens the pool of rented AI capacity. A modest long bias makes sense on any AI-capex enthusiasm that spills into cloud demand, with a 3-6 month horizon.