Broadcom: Incrementally More Bullish Heading Into Earnings

Key Morningstar Metrics for Broadcom

• Fair Value Estimate
  : $550.00
• Morningstar Rating
  : ★★★★
• Morningstar Economic Moat Rating
  : Wide
• Morningstar Uncertainty Rating
  : High

Broadcom AVGO reports fiscal second-quarter results after the market closes on June 3. Shares have risen 33% year to date as of June 1, including rising 57% since a recent nadir on March 30.

Why it matters: We’re focused on management’s outlook into the second half of the year, which we expect to include the first shipments of Google TPUs to Anthropic. We see this as a meaningful catalyst for shares in the next six months.

• Broadcom guided to shipments for 10 gigawatts of artificial intelligence capacity in fiscal 2027, which would likely be well above $100 billion in revenue. We model close to $120 billion in AI revenue in fiscal 2027, led by $60 billion from Anthropic, ongoing growth from Google, and the start of shipments to OpenAI.
• We’re bullish on Broadcom’s XPU opportunity. We expect XPUs to carve out 30% share of total AI compute as customers seek diversification and higher efficiency for large-scale inference applications. We model 75% annualized growth for Broadcom’s XPU sales through 2030.

The bottom line: We raise our fair value estimate for wide-moat Broadcom to $550 per share from $500 as we incrementally raise our bullish medium-term growth estimates. Shares look undervalued heading into the June 3 print.

• We’re 10% above consensus for fiscal third-quarter sales and continue to see the market underpricing the imminent shipments to Anthropic over the next 18 months. We also believe initial shipments to OpenAI, potentially up to 1 gigawatt next year, are being undervalued in the stock.
• We’re above consensus on long-term revenue but more conservative on margins. We see modest operating margin and earnings leverage from XPUs, but gross margin dilution keeps this minimal in our forecast. If Broadcom can use pricing to earn more leverage, there’s upside to our model.