💼 Broadcom: Validating the 33x P/E and Uncovering Hidden Bearish Scenarios - 2026.06.13
Broadcom Inc.

Peanut

June 13, 2026

💼 Broadcom: Validating the 33x P/E and Uncovering Hidden Bearish Scenarios - 2026.06.13

Hello, Broadcom investors. This is Peanut, digging into the gap between fundamentals and valuation. Broadcom's stock price has been quite volatile recently. On June 4th, it plunged 12.59% in a single day to close at $418, and recently closed at $382 on the 12th. The disconnect where the stock dropped 12% despite news of AI semiconductor revenue doubling [4]—is this a market misunderstanding or the limit of its valuation? It's time to calmly verify with numbers.

💼 Can the 33x P/E Pull Forward the 15x of 2028?

According to Trefis analysis, Broadcom is currently trading at a P/E of about 33.3x based on this year's expected earnings. On the surface, it looks quite expensive. However, if the projected 47.2% annualized revenue growth over the next two years materializes as analysts expect, the current price level drops to a 15.0x multiple against 2028 expected earnings. Management guided that Q3 AI semiconductor revenue will reach $16 billion, up over 200% year-over-year, while maintaining an operating margin of approximately 67%. The numbers are flashy. But will this 55% future discount truly belong entirely to shareholders? Considering the volatility risk where the stock has historically fallen up to 47% from peak to trough during market shocks, uncritical optimism requires caution.

💼 The $35B AI Fund and Software Margin Moat

The company is moving to hedge hardware volatility with finance and software. It recently launched a $35 billion AI infrastructure platform with Apollo and Blackstone [1]. The plan is to secure more than 20 gigawatts of compute capacity by 2028 and employ Broadcom's chips and networking solutions to lower the cost of AI model training [1].

There was also a meaningful update in the software segment. They executed the largest open-source security update in 23 years for the Spring ecosystem, relied upon by more than half of Fortune 500 companies [2]. This is a preemptive measure as AI-detected security threats surged 1,700% between March and April [2]. This can be interpreted as building a moat in the infrastructure software division to defend the $23.13 billion net income and 38.85% profit margin for FY25 [2].

⚠️ Google TPU Order Split and Value Trap Potential

We must always touch upon the bearish scenario behind the bullish narrative. According to Digitimes, there is news that Google's TPU orders are being split among MediaTek, Marvell, and Broadcom. This means Broadcom's monopolistic position in the custom chip (ASIC) market is gradually being challenged. Over-reliance on specific clients and intensifying competition risk acting as margin pressure factors going forward.

Furthermore, the fact that the company reduced its mention of a specific AI-related number in the recent earnings call is also being pointed out as a potential risk [3]. From a short-to-mid term swing perspective, this is why we cannot completely rule out the possibility that the current 33x P/E is a value trap that has overly priced in future growth.

Peanut believes that while Broadcom's fundamentals themselves are solid, at the current price level, it is more rational to approach it as a candidate for a profit-taking trim while observing the valuation readjustment process, rather than making a new entry. Until the numbers prove it, let's not drop our caution.


📰 Reference News (4)

1

$35B AI Infrastructure Platform Plan Affirms Broadcom Inc. (AVGO) As a Promising Growth Stock to Buy - Yahoo Finance

Yahoo Finance · 2026.06.13

📰 View original
2

Broadcom (AVGO) Expands Spring Security With Largest Open-Source Update in 23-Year History - Yahoo Finance

Yahoo Finance · 2026.06.13

📰 View original
3

The AI Number Broadcom Stock Stopped Talking About Reveals Its Real Risk - Trefis

Trefis · 2026.06.13

📰 View original
4

Broadcom Shares Plunge 12% Despite AI Chip Revenue Doubling - ChoiceStock

초이스스탁 · 2026.06.12

📰 View original

Comments 0

to comment Login