Goldman Sachs (GS) analyst Jim Covello says that investors may want to rethink how they are playing the AI boom. Instead of focusing on chipmakers, he suggests favoring large tech companies, often called “hyperscalers,” like Amazon (AMZN), Microsoft (MSFT), and Alphabet (GOOGL). His view is based on the idea that the market is already skeptical about how much money these companies will make from their heavy AI spending. As a result, their stock valuations have been pushed down, which could create an opportunity if those concerns prove to be overdone.
Claim 55% Off TipRanks
Trade AMZN with leverageThis view goes against recent market trends. Over the past year, semiconductor stocks have been the clear winners, with the Philadelphia Semiconductor Index (SOXX) rising nearly 150% as investors bet on strong demand for AI chips. Meanwhile, hyperscalers have lagged, mainly because investors are worried about their large spending on data centers and infrastructure. However, Covello believes that this gap could close. In one scenario, hyperscalers could start showing strong returns on their AI investments, which would likely boost their stock prices while limiting further upside for chipmakers.
In another scenario, hyperscalers could slow down their spending if returns remain weak. While that might sound negative, Covello argues it could actually help their stocks by improving cash flow, while hurting semiconductor companies that depend on that spending. At the same time, chip stocks are now trading at higher valuations compared to their historical averages, while hyperscalers are still below theirs. The main risk, however, is if nothing changes and hyperscalers keep spending heavily without clear returns, which could continue to pressure their stocks while supporting chip demand.
Which Hyperscaler Is the Better Buy?
Turning to Wall Street, out of the three hyperscaler stocks mentioned above, analysts think that MSFT stock has the most room to run. In fact, MSFT’s price target of $553.08 per share implies 35.2% upside potential.


