AI to Value Sector Rotation: The Tip of the Iceberg

July saw a clear shift in market psychology. Sector rotation began, with capital flowing away from previous momentum and into value and small caps. For the first time in years, the Nasdaq (QQQ) is lagging, while the Russell 2000 small cap ETF (IWM) is outperforming.

Is this just a temporary blip or the tip of the iceberg?

I believe this is the start of a much larger reversal; one where small cap and value will be much better. In this article, we will discuss the reasoning behind this view.

1. Long road to price normalization
2. Macroeconomic changes favoring small capitalization and value
3. Capex on AI vs ROIC on that capex
4. Pockets that look especially casual

Relative prices are still tight

The Nasdaq-smallcap price disparity began in the pandemic with what was a genuine fundamental tech advantage. Lockdowns forced the world to go digital, and tech companies were the overwhelming beneficiaries. QQQ’s initial surge was well supported by earnings advantages.

Over time, price performance attracted momentum investors, which caused the technology market price to surge. The market price delta exceeded the earnings growth delta, so that relative multiples widened. Barron’s noted a Nasdaq 100 P/E ratio of 31.89X.

You would think that the 14% advantage of small caps over tech over the past month would have brought prices back to more normal levels, but there is still a long way to go. The scale of the tech advantage has been so large that the small correction seen so far is just scratching the surface. Below is the 5-year price performance of the Russell 2000 (IWM) vs. QQQ.

From a pricing perspective, it appears that small caps and value stocks still have plenty of room to run.

Some of the outperformance for technology stocks has been fundamental in nature, but there is reason to believe that fundamentals are tilting in favor of small-cap and value stocks.

Macroeconomic changes favoring small capitalization and value

One of the most distinguishing features of trillion-dollar or near-trillion-dollar mega-caps is that they have little to no debt and, in many cases, huge sums of cash on their balance sheets.

A decade ago, that cash was almost a drag on earnings because it wasn’t generating much profit while highly leveraged companies like those in the Russell 2000 were growing their profits.

Skyrocketing interest rates upended this paradigm. Suddenly, mega-caps’ cash hoards were earning as much as 5% interest, and small caps were suffering from skyrocketing interest costs.

That was a big part of the mega-cap fundamental advantage, and it looks like it’s at least partially reversing. The market overwhelmingly believes the Fed will cut rates in September, with some insiders like former Fed official William Dudley calling for cuts as early as July.

The drop in interest costs would boost earnings for small caps and other sectors that use debt, such as utilities and REITs. At the same time, cash-rich megacaps would lose some earnings as their cash yields rise from 5% to perhaps 3% or 4%.

As such, earnings momentum could shift in favor of small caps and value. Sentiment also appears to be shifting as the market begins to question the ROIC of AI-related capex.

Capex on AI vs ROIC on that capex

So far, providing AI solutions has proven to be very profitable, but AI itself has not yet delivered significant profits.

I’m not saying it won’t be profitable or can’t be profitable, but rather that it hasn’t been profitable yet. I think the profitability of AI has been clouded by the lack of distinction between revenue generated by delivering AI and revenue generated by AI itself.

Revenue from providing artificial intelligence

There are many companies that have already made big profits from delivering AI. The most obvious example is NVIDIA (NVDA), which has enjoyed astronomical profit growth. A few others that have gained from delivering AI include:

• Data centers have benefited from providing physical structures to store AI processors.
• Utilities are ramping up production to sell more electricity for use in artificial intelligence

They all have something in common. They generate revenues from the capital expenditures that other companies invest in AI. It is not AI itself that generates these revenues.

In other words, the profits that used to be attributed to AI are largely covering the capital expenditures of other companies.

The market is starting to question the ROIC in these expense lines. In most cases, AI functionalities have not yet been monetized.

Personally, I have no idea whether ROIC will end up being high or low, but the fact that the market is starting to question it suggests that multiples could be coming down, especially for some of the more speculative AI plays.

There is some potential for AI services to become a necessary expense rather than a new profit center. Consider the following:

When hotels first introduced WiFi, it was an added convenience that brought guests to their hotels. It increased occupancy, and hotels could even charge customers to connect to WiFi. Increasingly, WiFi was available in all hotels and was no longer a special attraction. It became a penalty for occupancy for hotels that didn’t have it, rather than an incentive for those that did. Today, WiFi has become a necessary expense. Hotels have to have it to attract guests, but it doesn’t generate revenue because it is almost always free.

Similar things could happen in some areas of AI. Companies will have to invest in AI to stay competitive, but since competitors are also offering such services, it may not generate additional demand or revenue.

I’m sure there are other areas where AI will deliver significant benefits.

The point is that the market will have to become more selective and it will become increasingly difficult to apply a general high profit multiple to everything in the industry.

Continuous rotation

Given the uncertain return on invested capital in AI and the very high multiples, I believe capital will continue to flow into small cap and value stocks.

The rotation into small caps has been rapid, up 11% in just one month. While I think it will last for a while, the Russell 2000’s run of good fortune may be a bit short-lived due to valuation.

While cheaper than the Nasdaq 100, the Russell 2000 is also relatively expensive – according to the same Barron’s report, its P/E ratio is 28.75X.

As they approach full value, I suspect rotation will continue to flow into those that still have some room to run, the value sectors.

For example, utilities have all but halted their gains, with the Utilities ETF (XLU) up just 3.36%.

Contrary to the market price, energy companies have strong fundamental momentum. They currently have a better than usual mix of growth and cheap valuation.

REITs were a slightly larger part of the July rotation, with the Vanguard REIT ETF (VNQ) up 7.68% in the past month.

However, the sector was so badly hit by the decline prior to this rally that valuations remain quite low by historical standards.

As of 7/25/24, the median REIT was 13.7X 2024 FFO. Historically, REIT FFO multiples have been fairly close to the earnings multiple of the broader market. The significant discount at which REITs are currently trading provides a good runway for further rotation in the sector.

Application

July marked a significant rotation out of momentum names into small caps and value. This inflection point is supported by fundamentals and given the significant valuation spreads that exist today, I think this is just the tip of the iceberg.

Small-cap stocks still have some room to grow, while larger pockets of the market are poised for a longer period of growth.