Prediction: This Unstoppable Stock Will Rejoin the $2 Trillion Club by 2027

Broadcom reported fiscal Q3 2026 revenue of $29.6B, up 86% year over year, driven by AI semiconductor sales of $16.7B that surged 221%; net income more than tripled to $13.1B. Management expects Q4 AI semiconductor revenue to rise 236% to $21.7B and indicated total AI revenue could reach $115B in fiscal 2027 and $230B in fiscal 2028. At a $1.77T market capitalization, Broadcom needs roughly a 13% stock gain to return to $2T, although its trailing P/S multiple of 20.2x remains well above its 10-year average of 10.5x.

Analysis

The investable issue is not whether custom accelerators displace GPUs outright, but whether hyperscalers shift the marginal dollar of AI capex toward internally controlled silicon. That mix shift is favorable to AVGO because ASIC programs create multi-year engineering, networking and packaging attach revenue, while NVDA’s risk is primarily incremental pricing power and ecosystem rent rather than an abrupt unit-demand collapse. TSM is a second-order beneficiary if custom-silicon volume translates into advanced-node and CoWoS demand; however, AVGO’s economics are more concentrated because it captures design ownership and customer-specific integration.

AVGO’s near-term rerating requires management’s AI outlook to convert from design-win narrative into revenue recognized on schedule. A 13% equity move is plausible over 1-3 months around the fiscal-year reset, but the valuation leaves little tolerance for a pushed customer tape-out, constrained advanced packaging, or a single large customer moderating deployment. The critical verification point is not aggregate AI revenue: investors should track custom-accelerator mix, networking growth, backlog conversion, and whether gross margin holds as lower-margin bespoke silicon becomes a larger share of sales.

Consensus likely overstates the binary threat to NVDA. ASICs win where workloads are stable, inference is at scale, and a customer can amortize design cost; frontier-model development and fast-changing workloads still favor GPU flexibility. The more durable relative-value expression is AVGO versus NVDA only if custom silicon is taking incremental budget share without a broader AI-capex slowdown. If hyperscaler capex decelerates, AVGO’s concentrated program exposure and elevated multiple could make it the higher-beta downside vehicle despite its apparent diversification.

Key Decisions for Investors

• Initiate a modest 3-month long AVGO / short NVDA dollar-neutral pair only on confirmation that AVGO reiterates or raises its next-fiscal-year AI outlook; target 10-15% relative outperformance. Exit if AVGO reports custom-silicon shipment deferrals, AI semiconductor growth decelerates materially, or NVDA raises forward data-center guidance enough to negate the mix-shift thesis.
• For outright AVGO exposure, wait for a 5-8% pullback or use a 3-6 month call spread rather than chasing a headline-driven rerating. The upside case is a fiscal-year guide reset; define downside at a guidance cut or gross-margin compression attributable to ASIC mix, which would likely trigger multiple compression faster than revenue estimates adjust.
• Add TSM as a watch-list confirmation trade rather than a direct recommendation: sustained advanced-packaging capacity expansion, wafer-start commentary, or stronger custom-ASIC revenue from AVGO would validate the supply-chain read-through. Lack of CoWoS capacity additions or customer prepayment signals would weaken the thesis.
• Do not short NVDA outright on this development. Maintain exposure hedged through the relative pair, since accelerator specialization can expand total AI workload demand and NVDA retains the strongest position in rapidly evolving training workloads.