Ketu Desai
By Ketu Desai
In recent weeks there have been some major developments in AI that have significant implications for index investors. As background, the S&P 500 is a highly concentrated index, where the top ten companies make up just under 40% of the index. All but one of these are tech and AI related companies with major hyperscalers such as Alphabet, Microsoft, Meta, and Amazon carrying large weights.
The investment proposition in hyperscalers is changing. Their AI CapEx spend is expected to reach north of $3 trillion by 2029. They are going from cash flow machines to negative free-cash flow, high return on equity to lower, shifting from asset-light to asset-heavy, issuing equity, and reducing buybacks. Google even announced an $85bn equity issuance.
There are reports that Meta might also issue equity with Amazon and Microsoft likely not too far behind. This is at a time when they are a source of funds for AI-winners and major new IPOs such as SpaceX, OpenAI, and Anthropic. It’s a lose/lose situation if they don’t spend and will fall behind on essential AI. If they do continue to spend, they will have to issue equity, level up, change their business model and have negative free cash flow.
Government regulation and intervention risk has also increased for hyperscalers. The US government issued an export control directive to suspend Anthropic’s Fable and Mythos 5 by any foreign national. As these models get more powerful, federal scrutiny will only increase. Further, it could shift power from a handful of labs toward model orchestration and open-source ecosystems. Ultimately, this feels like models will be commoditized and pricing will go significantly lower, resulting in a low return on investment. Deutsche Bank says, “For the bulk of everyday tasks (perhaps 90% of them) [China’s DeepSeek’s V4-Pro] does much the same job at roughly 1.5% of the cost of Fable 5.”
OpenAI is reportedly taking its first steps to lower pricing and gain share. There are better opportunities for capital than in companies that are in negative free cashflow, issuing equity, have potential for government intervention, and potentially facing a price war. This is reflected in their stock prices, as they are down for the year and underperforming in the broader market by double-digits.
With the headwinds hyperscalers face, equal-weight strategies and small caps are set up to take leadership. With nominal growth running in the mid-single digits, many more companies can provide high growth. The market will look to the companies that are the cheapest for that high growth. This can result in the revenge of value stocks, small caps, old-economy stocks, cyclicals, banks and financials, industrials, certain healthcare and biotech companies.
Basically, most commodities have done very little over the last 15 years, while the market was held up by a select few stocks. Goldman estimates 36% earnings growth for small caps this year and 38% the next. Not only will many of these companies benefit from higher top-line growth from higher nominal GDP, but they will also likely see margin expansion over time from AI & technological advances. Every 1% labor cost savings translates to not only about a 2% boost in EPS for the S&P, but over a 6% boost for small caps. They will also see margin expansion from lower rates which will favor many of these businesses, as they are often more levered. The combination of moving past peak oil prices and hawkishness will allow rates to stabilize and eventually decrease. The 5YR breakeven inflation rate is down to 2.24%, the lowest of the year, allowing the Fed to tone down its bias.
Many of these market areas have low positioning and a smaller cap, so it won’t take much reallocation for big returns. The cap of the top ten stocks is approximately $28 trillion, while the one for the entirety of Russell 2000 is just $3.5 trillion. A 10% allocation away from the top ten S&P names would mean nearly doubling Russell. We are just at the beginning of such a reallocation.
Positioning is only in the 8th percentile, and Russell has more than doubled the S&P’s performance this year. It is just breaking out of a five-year base with long-term momentum and trend indicators pointing higher. The spread between large and small is at historically wide levels. When we had similar levels of valuation for small caps relative to large, these went on to outperform for the next twelve years by 6.5% per year.
Looking ahead, the market will focus on second quarter earnings, AI, and the latest from Washington and the Fed.
