Should You Buy the Dip in These 2 Beaten-Down Artificial Intelligence (AI) Stocks?

CoreWeave reported Q1 revenue up 111.6% year over year to $2.1 billion, but net loss widened to $740 million from $315 million and guidance missed expectations. Broadcom posted Q2 revenue of $22.2 billion, with AI semiconductor revenue up 143% to $10.8 billion and adjusted EPS up 54% to $2.44, but next-quarter AI guidance disappointed. The article views CoreWeave as riskier due to an 8.2x sales multiple and 67% revenue concentration with Microsoft, while Broadcom appears more attractive despite a 33x forward P/E.

Analysis

The market is no longer paying for AI exposure indiscriminately; it is starting to distinguish between revenue quality, customer concentration, and the path to self-funded growth. That is a meaningful regime shift: infrastructure names with visible demand still work, but anything with one or two anchor customers or a large cash burn is now being marked as a financing and concentration risk rather than a growth premium. In that setup, the better risk-adjusted winners are not necessarily the fastest growers, but the names that can convert AI capex into durable free cash flow without relying on continual multiple expansion.

CoreWeave’s setup is attractive only if AI buildout remains a multi-year straight line. The second-order issue is that hyperscalers can stretch, delay, or re-source spend faster than the market expects, and that would hit a model with heavy operating leverage and a narrow customer base disproportionately hard. The stock can still work as a long-duration call option on AI capex, but the current valuation leaves little margin for guidance misses; the move lower should be treated as a reminder that growth without diversification is fragile once the market starts demanding proof of monetization.

Broadcom is in the stronger tactical position because it has a profitable core that can absorb AI volatility and a custom-chip angle that benefits from customers seeking supply-chain control and better economics versus merchant GPUs. The hidden upside is that custom silicon adoption can expand faster than headline AI spend if large buyers optimize for cost per inference, not just performance per watt. Even so, the customer concentration and elevated multiple mean the stock will likely trade on next-quarter AI bookings and design-win visibility; the catalyst is not broad AI demand, but whether the company can keep converting those design wins into a multi-quarter revenue ramp.

Key Decisions for Investors

• Long AVGO on weakness over a 3-6 month horizon; risk/reward favors a 15-20% recovery if AI ordering visibility stabilizes, while downside is better contained by recurring cash flow and dividend support.
• Avoid initiating a full-size long in CRWV; if exposure is desired, use a small starter position only after a pullback and treat it as a 12-24 month call option on AI capex with high volatility and potential 30%+ drawdowns.
• Pair trade: long AVGO / short CRWV for 1-3 months to express preference for profitable AI exposure over high-burn infrastructure; thesis is multiple compression on CRWV if guidance remains uneven while AVGO re-rates on cash flow resilience.
• Buy AVGO call spreads instead of stock if seeking event-driven upside into the next earnings cycle; this captures rerating potential while limiting damage if the market stays skeptical of AI guidance.
• Watch MSFT capex commentary closely; if hyperscaler spending inflects lower, cut CRWV exposure first, as that is the fastest-transmission risk channel and the most likely source of a 20-40% downside move.