Key Takeaways
- Goldman Sachs equity research co-head Jim Covello advocates pivoting AI investments from semiconductor stocks to hyperscaler platforms
- Cloud infrastructure providers trading below historical valuations while investors scrutinize AI capital expenditure returns
- Semiconductor benchmarks surged approximately 150% over 12 months, creating valuation concerns
- Investment thesis works under two scenarios: hyperscalers demonstrate AI profitability or reduce capital spending to enhance cash generation
- Primary downside involves continued heavy spending by cloud giants without measurable returns
A senior Goldman Sachs analyst is challenging conventional wisdom about AI investment strategies. Jim Covello believes the market has been concentrating on the wrong segment, suggesting cloud infrastructure giants may deliver superior returns compared to chip manufacturers.
Covello, who serves as co-head of equity research at Goldman Sachs while maintaining semiconductor coverage, presented his perspective in a Thursday client communication.
His thesis revolves around valuation dynamics. Major hyperscalers including Amazon, Microsoft, Alphabet, Meta, and Oracle have experienced multiple compression. Market participants remain unconvinced these technology leaders will generate adequate returns from their substantial AI infrastructure investments.
Semiconductor manufacturers, conversely, have dominated AI investment sentiment. The Philadelphia Semiconductor Index has climbed roughly 150% during the trailing twelve months.
This impressive rally has elevated chip stock valuations beyond their long-term average multiples. Hyperscaler equities, however, continue trading at discounts to their historical norms.
Dual Pathways to Investment Success
Covello identified two distinct scenarios supporting a hyperscaler investment strategy over semiconductors.
The first scenario envisions hyperscalers beginning to demonstrate positive AI investment returns. This development would alleviate market skepticism and drive share price appreciation. Semiconductor stocks would face limited upside potential since current valuations already reflect optimistic expectations.
The alternative scenario involves hyperscalers reducing capital expenditures if returns remain disappointing. While this appears negative superficially, Covello contends it could benefit their equities through improved free cash flow generation. This outcome would simultaneously damage semiconductor firms dependent on hyperscaler spending.
Potential Headwinds for the Strategy
Covello highlighted a middle-path outcome as the principal threat to this investment approach. Should hyperscalers maintain aggressive spending levels while failing to demonstrate tangible returns, their stocks could face sustained pressure.
This identical scenario would perpetuate robust chip demand, continuing to underpin semiconductor sector performance.
Covello’s analysis emerges as major technology corporations face heightened investor scrutiny regarding data center and AI infrastructure capital allocation.
Amazon, Microsoft, and Alphabet have each announced significant capital spending commitments extending through 2025 and subsequent years.
Meta has similarly accelerated AI-related investments, prompting questions about revenue conversion from these expenditures.
Oracle has established itself as a significant AI cloud infrastructure provider and reports robust service demand.
The semiconductor industry has capitalized on this spending surge, with AI data center chip suppliers maintaining strong order pipelines.
Covello’s recommendation runs counter to recent market momentum but rests on current valuation metrics relative to historical benchmarks.
