Alexander Sikov/iStock via Getty Images
Neither the market nor the economy is fully efficient in real time, but both trend toward efficiency over the long run.
Economic inefficiency occurs when supply and demand get out of line with each other, resulting in a temporary window of excess value capture. Companies that produce the undersupplied good or service can experience higher than normal margins, volume, or both, thereby achieving greater than normal profit.
Over time, this corrects. The abnormally high profit attracts competition.
- If a good is undersupplied, competitors will add capacity.
- If a product is superior, peers will spend on R&D until they can match or exceed it.
Periods of excess opportunity are unequivocally positive for the companies positioned to capture the profits, but they are usually limited in scope. Eventually, economic equilibrium is restored and profits return to normal levels.
A significant inefficiency in the stock market today is that these periods of excess opportunity are being priced as if they are permanent. Multiples suggest that equilibrium will not be restored. The market is assuming capacity will not be added or that competitors will not be able to catch up in product quality.
The stocks with unusually high profits presently are broadly overvalued, while those that profit in the restoration of equilibrium are broadly undervalued. This article will discuss the process of equilibrium restoration regarding current economic distortions and point out areas of excess and areas of cheapness.
Micron serves as an excellent case study on how consistently and repeatably equilibrium does indeed restore
Micron Technology, Inc.’s (MU) growth is explosive right now. It is projected to grow from $8.29 normalized earnings in 2025 to $96.92 in 2027.
S&P Global Market Intelligence
The nature of this growth is that there is a severe undersupply of memory in both NAND and DRAM due to the buildout of AI infrastructure.
Quite simply, not enough units are being produced to satisfy demand, which has 2 financial implications for MU:
- Sales volume is full (every unit produced can be sold)
- Price per unit is up substantially.
With both high volume and abnormally high margins, Micron’s earnings are going parabolic. The stock is pricing the period of extreme growth as if it is permanent.
Seeking Alpha
A cyclical rise in earnings would not result in the stock surging 300%+.
Yet Micron’s history shows that similar cycles are routine. Gross profit margin is a rollercoaster.
S&P Global Market Intelligence
There is a clear pattern:
- Memory gets undersupplied, causing scarcity, which makes prices spike.
- MU’s gross margins soar to about 50%
- The industry increases capacity
- Undersupply becomes oversupply, and prices go back down
- Margins drop off and earnings collapse
Net margin, rather than gross margin above, shows that when the downside of the cycle happens, it often results in negative margins.
S&P Global Market Intelligence
Each period of excess earnings has been followed by a period of weak or negative earnings.
S&P Global Market Intelligence
Bulls suggest that this time is different. Perhaps the demand is more permanent because the extreme buildout of data centers will continually need memory, even just through replacement.
I tend to agree that demand is likely to stay elevated, but that is only one side of the equation. The reason margins and thus earnings are so high right now is because supply is insufficient to keep up with that demand.
High prices make new manufacturing facilities economically viable. Supply will inevitably increase, and when it does, margins drop back down. Even Micron itself is materially increasing supply with new fab facilities and $25B of capex.
Mark Murphy, MU’s CFO, discussed their increased capacity on the earnings call:
We increased our outlook on CapEx to over $25 billion, which was up from the $20 billion we did on the last earnings call. The updated investment reflects this investment in the Tongluo fab, which we communicated in February conference and increases in U.S. expansion. Again, it’s DRAM and HBM driven, including the increase. Today, we also provided more detail on construction, which we’ve said in the past was becoming a more material part of the build-out, obviously, because we need greenfield capacity. On the December call, we said expect that FY ’25, ’26 construction would double. And so we expect FY ’26 construction to be mid- to high single-digit billions, net. When I talk CapEx, we’re talking net CapEx. Now as we look out to ’27, we did say that we project approximately a $10 billion incremental construction cost and also for equipment spend to increase. Now that ’27 spend, NAND will begin to increase, but it will still be a much smaller portion of the spend compared to DRAM.
Extrapolating the current growth deep into the future ignores the feedback loops that are inherent to commodity products.
The high prices mechanistically cure themselves. High prices stimulate more supply, which brings prices back down.
Micron is merely one example. I believe the market is incorrectly extrapolating earnings growth for most of the companies that are getting an immediate revenue burst in the buildout of AI.
Demand is temporarily out of sync with supply, but the mechanisms that restore equilibrium are still present. The period of abnormally high earnings is great for the companies positioned to capture it, but it will not last forever. I believe the following categories of companies are having their temporary earnings surge overly extrapolated by the market:
- Turbine and power plant equipment makers like GE Vernova (GEV)
- Chips, memory, and other commodity hardware
- Independent power producers selling power at surged auction prices.
The factor that each of the above categories of companies has in common is that they sell a product in which revenues are earned immediately at the point of sale.
There is a whole different set of companies that addresses the same supply/demand imbalances but does so through recurring revenue models.
Data center REITs like Equinix (EQIX) build and lease data centers. The revenues start as a trickle but will continue indefinitely. The growth is substantially slower than that of the explosive immediate revenue recognition companies.
S&P Global Market Intelligence
However, there is no cliff on the other side.
Eventually equilibrium will be restored with a sufficient number of data centers to meet demand. When that happens, EQIX’s growth is likely to slow down, but they will still have the recurring revenue streams of the data centers they already built and leased.
So while MU’s earnings look far more impressive in the near term, there is a huge difference between sustainable earnings growth and a cyclical spike.
Once data centers are built, they draw power continuously. This power demand is in some instances overwhelming existing capacity such that grids like PJM are paying through the teeth to procure more capacity in the immediate term:
Utility Dive
IPPs are getting explosive growth by selling into these auctions, but their margins will drop back down once supply catches up to demand. The growth is great, but it is not permanent.
Unlike the IPPs, regulated electric utilities are capturing sustainable growth through permanent expansion to their rate bases. It takes much longer to kick in and is less flashy, so the market is not bidding them up as much, but the growth is essentially permanent.
Long-term sustainable growth should trade at a higher multiple than cyclical peak earnings.
Right now it is not.
The explosive cyclical growth is trading at multiples higher than that of the secular growers.
That is substantial mispricing, and one can capture this market inefficiency by trading out of companies with the immediate revenue bursts and into companies that address the same underlying demand with repeatable revenue streams.
