Transcription
There is a mechanism operating in the equity market right now that almost nobody is discussing with the precision it deserves. It involves two of the largest initial public offerings in financial history, approximately $200 billion that needs to come from somewhere, and a rotation dynamic that will force selling in the stocks most retail investors are currently holding. Not because those stocks have deteriorated, not because earnings have disappointed, but because of a structural liquidity event that the market's architecture makes essentially inevitable.
I want to explain this mechanism completely, not because I enjoy creating anxiety about market conditions, but because understanding the structure of how large capital movements work is the foundation of any rational investment decision in the current environment. And the investors who understand this specific mechanism before it fully develops will be positioned very differently from those who understand it after the fact.
Let me start with what Ray Dalio actually said, because the way it has been reported misses the most important part of his argument. Dalio has spent 50 years managing capital through every major market cycle in the modern era. He did not warn that artificial intelligence is a bad technology or that the companies building AI infrastructure will fail. He said something more specific and more analytically important. He said that every great technological transition produces a bubble and that bubbles end not when the technology fails but when paper wealth needs to convert into cash. The pricking of a bubble in his precise formulation is the converting of wealth into money. That distinction between paper wealth and and actual liquidity is the key to understanding what is happening right now.
Uh his bubble indicators, he noted, are approaching levels last seen in 1929 and in 2000. Both of those years ended in severe market corrections. Both were followed by periods of extraordinary opportunity for investors who were liquid and uh and prepared and both were characterized in their final stages by exactly the dynamic we are observing now, an acceleration of new issuances at the moment of maximum optimism creating a structural demand for liquidity that the existing market cannot supply without significant price adjustment.
Now, let me show you the specific numbers because the magnitude matters. SpaceX is preparing to list at a valuation of approximately $1.7 trillion with an initial offering expected to raise approximately $75 billion. Anthropic has filed confidentially with the SEC at a valuation approaching $1 trillion targeting a an offering later this year that will require comparable capital deployment. Two offerings combined capital requirement in the range of 130 to 200 billion dollars in a single calendar year.
To understand why this number is significant, consider the context. The total capital raised across all initial public offerings in the United States since 2022, a 4-year period, is smaller than what these two offerings alone will require. The money has to come from somewhere, and in a market where institutional capital is fully deployed, where the NAAIM exposure index shows um professional money managers at approximately 97% invested, you uh the somewhere is not new cash sitting on the sidelines, it is existing positions being liquidated to fund the new purchases. This is the mechanism most analysis is missing entirely.
When SpaceX lists on a major index, which happens automatically um within days of its market debut, every index fund, every ETF that tracks that index, is required by its mandate to hold SpaceX in proportion to its market capitalization. The purchase is not discretionary, it is structurally mandated, and the capital to fund that purchase comes from proportional reductions in every other position in the index, not because those companies have become less valuable, because the index weights must be rebalanced to accommodate the new entrant. This is structural, mandatory, mechanical selling pressure on every existing index constituent. At a scale of $1.7 trillion in new market capitalization entering the index, the selling pressure on existing holdings is material and certain. It is not a question of if, it is a question of magnitude and timing.
Um but the structural index uh rebalancing is only the first layer of the mechanism. The second layer is what I would describe as the proxy trade unwinding, and it is the layer that explains specifically why uh the stocks most retail investors currently hold are most vulnerable. For the past several years, investors who wanted exposure to the AI economy had a limited set of options. You could not buy open AI directly, you could not buy SpaceX directly, you could not own Anthropic before its public listing. So, investors did the logical thing. They bought the companies most directly exposed to the AI opportunity that were available in public markets. Nvidia for compute, Microsoft for enterprise AI integration, Alphabet for AI search and infrastructure, Amazon for cloud. These companies became, in effect, the proxy trades for a set of private companies that were generating enormous attention and investor enthusiasm. The proxy trade works until the underlying asset becomes directly available. When SpaceX lists, the investor who owned aerospace defense companies as a space proxy can own SpaceX directly. When Anthropic lists, the investor who owned Microsoft as an AI infrastructure proxy can own Anthropic directly. The proxy trade does not unwind all at once. It unwinds gradually, systematically, as the price relationship between the proxy and the underlying asset is recalibrated by the market's collective judgment. But, the direction is consistent. Capital flows from the proxy into the direct exposure. And the proxy experiences selling pressure that is independent of its own fundamental performance.
This is, um, in my analytical view, what Broadcom's behavior illustrates with unusual clarity. Uh, Broadcom, uh, reported earnings that were, uh, by any objective measure, exceptional. Revenue grew 48% year-over-year. Semiconductor revenue doubled. Earnings per share exceeded analyst expectations. The company guided strongly by the conventional logic of equity markets when and some there were there were were improving fundamentals driving improving prices meet turning archer you know Broadcom should have risen on these results it fell 12% in a single session. The explanation is not that the results were bad. The explanation is that the results were priced in and then some the stock had already appreciated 40% in the preceding months as as institutional investors positioned for exactly the kind of result that materialized when when the result confirmed the the expectation the investors who had positioned in advance did what rational investors always do at the moment of confirmation. They took the profit. They did not take the profit because Broadcom deteriorated. They took it because the capital has a better use specifically the AI infrastructure companies that are about to list and that represent direct rather than proxy exposure to the theme that Broadcom partially expressed. This behavior good earnings falling price is one of the most reliable signals of a late cycle market. It tells you that the marginal buyer has already bought that the positive news is already reflected in the price and that the next move requires either a further acceleration of the narrative or a source of new capital. In the current environment neither condition is easily satisfied.
Now I need to add the layer that makes this more complex than a simple IPO liquidity story. And this is the layer that most market analysis is treating as a separate story when it is in fact part of the same system, inflation and oil. The Federal Reserve's capacity to provide the liquidity backstop that has sustained equity valuations across every market stress event of the past 15 years depends on a single condition, that providing that liquidity does not accelerate inflation to a level that becomes politically and economically unacceptable. In 2020 that condition was easily satisfied because the pandemic had caused a sharp deflationary shock. The Fed could print aggressively because there was no inflation risk. In 2023 that condition was harder to satisfy but still manageable. In 2026 with consumer prices running well above target and an oil shock from the Iran conflict driving energy costs higher, the condition is actively failing. Goldman Sachs has modeled oil reaching $150 per barrel in their adverse scenario. That number is not a certainty but it represents a plausible outcome if the Strait of Hormuz disruption continues at its current level. And at $150 uh per barrel the e the uh inflationary uh pressure on the entire uh economy and on transportation costs, on manufacturing input costs, on consumer energy expenditure makes the Federal Reserve's ability to provide accommodative monetary policy in response to a market decline essentially impossible. The Fed would be forced to choose between supporting the market and controlling inflation. And the political and institutional imperatives of controlling inflation in an environment where voters are experiencing real purchasing power erosion would likely win.
This is the specific scenario in which the correction I have described becomes a liquidity event rather than a valuation adjustment. The difference is significant. A valuation adjustment is painful but orderly. Prices fall to levels that better reflect um discounted future earnings at higher discount rates. Investors absorb losses and the system resets to a more sustainable equilibrium. A liquidity event is disorderly. Forced selling creates price declines that trigger more forced selling, margin calls compound the downward pressure and and assets fall to prices that significantly undershoot their fundamental value. The historical record of every major oil shock preceding a market correction is not coincidental. The 1973 oil embargo preceded the worst US stock market decline since the Great Depression. The 1979 oil shock um preceded a period of financial repression and market stagnation that lasted years. The 2008 oil price spike to $147 per barrel in July of that year preceded and contributed to um the acceleration of the financial crisis in September. Oil shocks matter because they create inflationary pressure that constrains the policy response at exactly the moment when a policy response is most needed.
Now, let me explain the timeline that the structural mechanics imply because vague warnings about future market stress are less useful than a specific um analytical framework for thinking about when and how the dynamics I have uh described or are likely to manifest. SpaceX's listing targeting mid-June marks the beginning of the capital absorption phase. In the weeks surrounding the listing, institutional capital will be redeploying from existing positions into the offering. The index rebalancing effect will generate structural selling in existing constituents. The proxy trade unwinding will create specific pressure on the companies that were most used as AI and space economy proxies. This phase is likely to be characterized by elevated volatility, sector-specific weakness in the most crowded AI-adjacent positions, and apparently paradoxical behavior where strong fundamental results generate weak price responses.
The 6-month lockup period on SpaceX insider holdings means that the full selling potential from pre-IPO shareholders cannot materialize until approximately December 2026. If Anthropic lists in October as targeted, its lockup would expire approximately April 2027. The period from December 2026 through April 2027 therefore represents the window of maximum selling pressure from the IPO cycle. The window when insider selling, index rebalancing, and the natural reassessment of AI company valuations based on their early public earnings results converge. This timeline also overlaps with the US midterm election cycle, which has historically generated policy uncertainty and risk-off sentiment in the periods immediately surrounding the elections. The convergence of IPO lockup expirations, earnings reality testing for newly public AI companies, and political uncertainty creates a window of elevated market stress risk that is more specific than the generic warning that markets are expensive.
The second order effect that I believe is most under appreciated is what this dynamic means for the companies and sectors that are not directly involved in the AI IPO cycle. When large amounts of capital rotate from diversified portfolios into concentrated IPO purchases, the assets that are sold are not randomly selected. They are the assets that are most liquid, most widely held, and most easily monetized without creating a market impact that defeats the purpose of the sale. These are disproportionately the large cap growth stocks that anchor most retail and institutional equity portfolios. The selling pressure therefore falls most heavily on the most widely owned assets, which means the correction, when it comes, will be felt most intensely by investors who believe their portfolios are well diversified precisely because they hold the most liquid and widely owned equities.
The counterintuitive implication is that the assets that benefit from this dynamic are not the obvious ones. The obvious defensive trade moving from growth to value, from technology to to utilities or consumer staples, provide some protection but is also subject to the general liquidity contraction I described. The assets that genuinely benefit are those with cash flows anchored to physical scarcity rather than financial sentiment, the commodity producers, the energy infrastructure operators, the gold mining companies whose revenues scale with asset prices that are determined by physical supply and demand rather than by the multiple the market is willing to pay for promised future earnings. Um the structural case for these assets is not that they are exciting or that they will produce spectacular um short-term returns. The case is that in an environment where financial asset valuations are compressing, where the multiple on future earnings is declining cuz risk risk free rates are rising and the liquidity backstop is is constrained, the assets that do not depend on a high multiple to deliver their value are the ones that preserve capital.
I want to address a specific behavioral error that I observe consistently in how investors respond to the kind of analysis I have just provided. The error is the conflation of timing precision with analytical validity. I have described a mechanism, IPO-driven capital rotation, proxy trade unwinding, index rebalancing pressure, oil constrained monetary policy, and I have described a window late 2026 through early 2027 when these forces are most likely to converge. That description is analytical, not predictive. The mechanism is real and the forces are operating now. The timing is a range, not a date.
Um the investors who respond to this analysis by waiting for the precise moment of maximum market stress before adjusting their positioning will, in my observation, be the investors who adjust too late. The forces I have described operate gradually and then suddenly. The gradual phase um the the capital rotation, the uh the proxy trade unwinding, the structural selling is already observable in the Broadcom price behavior I described, the sudden phase, the moment when the cumulative effect of these forces becomes visible visible as a market level event is not predictable with the precision required to time a portfolio adjustment to the day or the week. The more durable positioning responses to evaluate each asset in a portfolio against a specific question. Does this asset's value depend on the continuation of conditions, cheap money, high multiples, unlimited liquidity backstop, strong growth relative to discount rate that need mechanism I have described is actively undermining. If the answer is yes, the rational response is to reduce that exposure gradually and systematically, not to hold it until the moment of maximum stress.
And the corresponding question for positioning into the stress event is which assets generate cash flows from physical economic activity that is relatively independent of financial market valuations. The companies producing the copper, uranium, and energy that the that the physical economy requires regardless of what happens to equity multiples. The gold mining companies whose revenue scales with the price of the asset that sovereign institutions are accumulating as a dollar alternative. The infrastructure companies with contracted backlogs whose revenues are committed regardless of the sentiment cycle in equity markets. These are not exciting positions. They do not generate the kind of returns that accompany a late cycle AI proxy trade in its final months. What they provide is something more valuable in the current environment, exposure to assets whose fundamental value is determined by physical scarcity and contractual commitments rather than by the market's willingness to pay for promised future earnings at a multiple that assumes the continuation of conditions that are changing.
The $200 billion that the IPO cycle will absorb will come from somewhere. The structural mechanics of where it comes from are not a matter of speculation. They are determined by the architecture of the equity market and the behavior of institutional capital under the specific conditions of the current cycle. The investors who understand the mechanism are positioned to respond rationally. The investors who do not will respond emotionally at the moment of maximum stress in exactly the way that transfers wealth from the unprepared to the prepared. That is the precise opportunity and the precise risk that the current moment presents.