Why Did Broadcom Stock Shed 15% After Bookings Tripled?

Broadcom’s AI semiconductor revenue jumped 143% year over year to $10.8 billion, with management reporting more than $30 billion in new AI bookings and guiding next-quarter AI chip sales to $16.0 billion. However, gross margin fell 230 bps to 77.1% and is guided to around 74% next quarter, reflecting lower-margin ASIC mix and pressure from AI-related expansion. The stock dropped 14.0% premarket as investors focused on margin compression rather than the strong AI demand backdrop.

Analysis

The market is not reacting to growth, it is repricing quality of growth. AVGO’s AI mix shift creates a subtle but important tension: the company is becoming more strategically embedded with hyperscalers just as its economics move closer to a custom-manufacturing model, which typically deserves a lower multiple than a standardized semiconductor platform. That means the first-order revenue surprise can coexist with a second-order de-rating if investors conclude incremental dollars are less profitable and more customer-concentrated.

The key competitive implication is that AVGO is helping enable inference cost compression for the largest AI buyers, which should accelerate deployment and extend the AI capex cycle. That is broadly negative for NVDA’s share of wallet at the margin, not because demand weakens, but because some workloads migrate from expensive general-purpose compute toward customized silicon where economics are optimized for hyperscaler-specific use cases. The real winner is the hyperscaler ecosystem itself: lower per-query economics can improve AI ROI and justify more spend, even if semiconductor suppliers capture less margin per unit.

The risk over the next 1-3 quarters is that investors focus on the gross margin glide path as a proxy for whether AVGO’s AI business is becoming accretive enough to offset mix dilution. If margin stabilization does not show up quickly, multiple compression can outrun earnings growth; if margins trough sooner than feared, the stock can re-rate hard because the current move likely embeds a lot of bad news. The most important catalyst is not another revenue beat, but evidence that bookings are converting into a durable backlog with less incremental margin leakage than feared.

The move looks somewhat overdone in the near term because the market is extrapolating a structural margin reset before it has proof that scale and customer lock-in can more than compensate. That creates a tradable setup: the stock can bounce on any sign that 74% is a floor rather than a stepping stone lower, but downside remains if management keeps signaling mix deterioration. The better question is whether AVGO is transitioning from a premium compounder to a high-growth but lower-quality earnings stream; if so, the discount is rational, but if not, this selloff may be an entry point rather than a warning.

Key Decisions for Investors

• Fade the panic: initiate a short-dated AVGO put spread or sell cash-secured puts into the post-earnings volatility if implied remains elevated, targeting a 2-4 week mean reversion trade with defined downside risk.
• Relative value: long AVGO / short NVDA for 1-3 months if the thesis is that custom ASIC adoption captures more AI inference spend while NVDA remains exposed to broader competitive pricing pressure in accelerators.
• Buy AVGO only on confirmation: add on evidence that consolidated gross margin stabilizes above the guided level for two consecutive quarters; otherwise treat rallies as sellable until margin inflection is visible.
• Pair trade on quality of growth: long hyperscaler beneficiary baskets vs short semiconductor hardware beta, since lower AI compute cost should improve cloud ROI faster than it improves chip supplier economics.
• If owning NVDA, use AVGO weakness as a hedge signal rather than a direct bearish read: reduce NVDA exposure on any sign that custom silicon share gains are accelerating over the next 1-2 quarters.