The 'Margin Moat' Trade: Nvidia, Broadcom Drive BofA's $1 Trillion AI Thesis Inside Semiconductor ETFs

The artificial intelligence boom isn't slowing down; instead, it is quietly being packaged into ETFs. Bank of America's latest semiconductor outlook suggests that for investors hesitant to chase popular stocks, ETFs might be the best way to keep up as AI spending heads toward historic levels.

In its report "2026 Year Ahead: choppy, still cheerful," cited by Yahoo Finance, BofA analyst Vivek Arya stated that the semiconductor industry is only halfway through a decade-long change. He predicts that global chip sales will rise by 30% year over year. This will finally push the sector past the $1 trillion annual revenue mark by 2026.

The ‘Margin Moat’ Strategy

Arya's investment strategy is straightforward: rank semiconductor companies by gross margins and invest in the leaders. His top six picks for 2026—Nvidia Corp (NASDAQ:NVDA), Broadcom Inc (NASDAQ:AVGO), Lam Research Corp (NASDAQ:LRCX), KLA Corp (NASDAQ:KLAC), Analog Devices Inc (NASDAQ:ADI), and Cadence Design Systems Inc (NASDAQ:CDNS)—dominate their specific markets, holding market shares typically between 70% and 75%.

This approach aligns well with major semiconductor ETFs. Funds like the VanEck Semiconductor ETF (NASDAQ:SMH) and the iShares Semiconductor ETF (NASDAQ:SOXX) hold significant stakes in Nvidia and Broadcom. They also provide exposure to equipment makers and chip designers who profit from AI capital spending.

Why The $1 Trillion Semiconductor Estimate Matters

According to BofA, AI data center systems alone represent a total market potential of over $1.2 trillion by 2030. This market is growing at a compound annual rate of 38%. AI accelerators make up about $900 billion of that potential, making chips crucial to AI expansion.

However, the high cost of AI infrastructure has made some investors hesitant. A single 1-gigawatt data center can require over $60 billion in capital expenditures, with nearly half of that going directly to hardware. The key question is whether the returns will be worth the spending. Arya argues that Big Tech has little choice; the investment is both proactive and protective, aimed at safeguarding existing platforms.

Nvidia And Broadcom Remain Central To The Trade

Nvidia is still the main focus of most semiconductor ETFs. Arya cautioned against valuing it like a traditional chipmaker, pointing out that while the average chip sells for around $2.40, Nvidia's GPUs can go for about $30,000. Despite its large rally, BofA suggests Nvidia's valuation remains attractive when factoring in growth.

Broadcom has also become essential to AI infrastructure due to its custom silicon for hyperscalers aiming to reduce dependence on Nvidia; this shift is increasingly shown in ETF weightings.

ETFs Provide A Smoother Path Through Volatility

Arya warned that the journey to $1 trillion will be uneven, and no stock is safe from risk. However, semiconductor ETFs allow investors to express long-term confidence in AI's growth while spreading risk across several leading players. This strategy may be particularly valuable as the AI cycle evolves.

Read Next:

• A Selective Santa Rally? ETFs May Reveal Where Conviction Still Exists in 2025

Photo: Shutterstock