History Says a $1,000 Investment in Broadcom Will Be Worth This Much by 2028

Broadcom reported $16.7 billion in AI semiconductor revenue in fiscal Q3 2026, up 221% year over year, driven by custom AI-chip demand from customers including Alphabet, Meta, OpenAI and Anthropic. The article projects AI semiconductor sales could exceed $65 billion in FY2026, rise to $115 billion in FY2027 and reach $230 billion in FY2028. Using Wall Street's FY2028 EPS estimate of $28.08 and a 30x multiple, the analysis values Broadcom at $842 per share versus about $360 currently, implying 134% upside and a potential $2,340 value for a $1,000 investment.

Analysis

The relevant debate is no longer whether hyperscalers will fund custom silicon, but whether Broadcom can retain sufficient economics as customers gain design sophistication and use multiple foundry/packaging routes. ASIC revenue carries unusually high customer concentration, long qualification cycles, and potentially lumpy shipment recognition; a large design win can support estimates, but it also gives GOOG, META and other buyers substantial leverage on pricing and inventory commitments. The market should not capitalize projected AI revenue at GPU-like scarcity margins without evidence that gross-margin mix and working-capital conversion are holding.

Near term, AVGO remains levered to any upward revision in hyperscaler capex, particularly if inference demand shifts from flexible training hardware toward stable, high-volume workloads. The second-order beneficiary is TSMC (TSM), whose advanced-node and CoWoS capacity are likely the actual bottleneck; however, AVGO's upside can be delayed rather than lost if packaging allocation constrains deliveries. NVDA is not necessarily the direct loser: custom ASIC adoption expands total AI spend, while NVDA retains the more valuable frontier-training and software-ecosystem workloads. The investable question is incremental wallet share, not binary displacement.

Consensus appears to accept a steep multi-year revenue ramp while also assuming earnings multiple expansion. At AVGO's scale, even a modest miss in custom-chip timing, yield, or customer deployment schedules can produce a material estimate reset and expose the stock to multiple compression. Validate the thesis through the next two earnings cycles: AI semiconductor backlog, customer prepayments, segment gross margin, and management's commentary on supply availability matter more than aspirational end-market forecasts. A deceleration in META/GOOG capex guidance or evidence that a major program is dual-sourced would falsify the near-term bullish case.

Key Decisions for Investors

• Maintain a measured long AVGO versus a short SOXX hedge over the next 1-3 months rather than add outright after bullish retail coverage. The pair isolates custom-silicon execution; exit if AVGO underperforms SOXX by 10% following earnings or if AI semiconductor guidance fails to rise.
• Add TSM as the cleaner supply-chain expression only on confirmation that advanced packaging utilization remains tight into the next calendar year. Target a 6-12 month horizon; key risk is a hyperscaler capex pause that turns packaging from a constraint into excess capacity.
• Do not short NVDA solely on ASIC substitution. Instead, monitor NVDA data-center gross margin and hyperscaler commentary for evidence of inference share loss; a sustained margin/guidance deterioration would justify revisiting an AVGO/NVDA relative-value long AVGO trade.
• For existing AVGO longs, buy downside protection around the next earnings print via a 3-6 month put spread financed with out-of-the-money calls only if implied volatility remains below the stock's post-earnings realized move. The main event risk is delayed program ramp or lower-than-expected gross-margin conversion.