Skip to content
Menu
Averaging Up
  • Home
  • About
  • Mental Models
  • The Infinite Investor
    • Preface & Table of Contents
    • Chapter I: The Lie of the Average
    • Chapter II: Ergodicity and Behavior
    • Chapter III: The Foundation
    • Chapter IV: The Four Buckets
    • Chapter V: The Lenses
    • Chapter VI: The Selection Process
    • Chapter VII: The Exceptions
    • Conclusion: From Finite to Infinite
Averaging Up

The Double Swell Freesurfer – When Two Waves Converge on the Same Reef

Posted on February 25, 2026February 25, 2026

🎧 Listen to the deep-dive discussion –  The AI and ETF Double Swell. (29:30 min)

https://averagingup.com/wp-content/uploads/2026/02/The_AI_and_ETF_Double_Swell.m4a

 

1. A Taxonomy of Structural Growth

In The Freesurfer, we introduced a formula for the rarest kind of business: one that grows without paying for its own growth. We called it A × B. A is the moat—the proprietary toll booth that the company owns and defends. B is the wave—an external force that expands the volume passing through the toll, at no cost to the company. The world creates B. The company merely collects.

Visa does not create digital commerce. It taxes each transaction as cash converts to card. MSCI does not push investors toward passive indexing. It collects a basis point on every dollar that migrates there on its own. The growth is structural, secular, and free.

But not all Freesurfers are the same species.

What follows is a discovery that emerged from observing how the artificial intelligence revolution intersects with these toll-booth businesses. The intersection revealed that the original formula—A × B—was only the first species. There are at least three.

* * *

Species 1: The Classic Freesurfer — A × B

One toll. One wave. The simplest and most elegant architecture.

The wave is external to the company. It is driven by forces the company does not control—technology adoption, demographic shifts, regulatory evolution. The company’s only job is to own the toll and not destroy it. The wave does the rest.

Visa and Mastercard are the purest specimens. Their toll (A) is the payment network—a duopoly protected by regulation, trust, and the switching cost of four billion cards in circulation. Their wave (B) is the secular migration of cash to digital payments. Every phone tap, every online checkout, every emerging-market consumer who opens a bank account for the first time adds volume to the network. Visa does not build the phones, does not open the bank accounts, does not create the e-commerce platforms. It simply collects its fraction of a percent on each transaction, in perpetuity.

The A × B structure explains why these businesses compound at extraordinary rates with almost no capital expenditure. There is nothing to build. There is nothing to maintain. There is only the toll, and the ever-expanding volume of traffic flowing through it.

* * *

Species 2: The Resilient Freesurfer — A × (B₁ + B₂)

Two tolls. Two independent waves. Or one toll collecting from two distinct sources of growth.

This species is rarer than the Classic, and structurally more resilient. If one wave weakens, the other continues. The company does not depend on a single secular trend—it straddles two.

Moody’s Corporation is the clearest example. Its toll (A) is the credit rating—a regulatory near-monopoly shared with S&P Global, where every significant debt issuance on Earth must pass through one of two agencies. Its first wave (B₁) is the structural growth of global debt—sovereign, corporate, municipal, emerging-market. As economies grow, as financial markets deepen, as the developing world integrates into the capital markets, the volume of debt requiring a rating expands. Moody’s does not create the debt. It taxes it at the gate.

Its second wave (B₂) arrived recently: artificial intelligence. Agentic AI systems—automated analysts, investment banking workflow tools, portfolio construction agents—consume credit data voraciously. Each automated workflow that touches debt issuance, credit risk, or fixed-income analysis pulls from the same proprietary ratings database. The AI does not replace the toll. It generates more queries through it.

But here is the critical distinction: B₂ does not amplify B₁. Artificial intelligence does not create more global debt. The AI accelerates the consumption of ratings data, but the fundamental driver of debt issuance remains what it has always been—interest rates, economic growth, corporate capital needs. If AI disappeared tomorrow, global debt would continue to grow. If global debt stagnated, AI would continue to consume historical ratings data. The two waves are additive, not multiplicative. They coexist without influencing each other.

This makes Moody’s more resilient than a Classic Freesurfer—it has a second engine—but not qualitatively different in kind.

* * *

Species 3: The Coupled or Double Swell Freesurfer — A × B₁↑B₂

One toll. Two waves. But unlike the Resilient species, the second wave amplifies the first. The two forces enter into a reinforcing loop—what systems theorists call a positive feedback loop, and what chemists call an autocatalytic reaction. The product of one process becomes the catalyst for the other, and vice versa.

This is the rarest and most powerful architecture. And to our knowledge, only one business exhibits it in its purest form.

Surfers call this a double swell — when two storms in distant, unrelated regions of the ocean send their energy toward the same reef at the same moment. The waves do not simply add up. They enter resonance. A two-meter swell meeting a two-meter swell produces not four meters but five or six — because the combined energy exceeds the sum of its parts. It is the rarest and most powerful condition in big wave surfing, and it is exactly what is happening in the financial markets today.

MSCI owns the indices that serve as the global lingua franca of investing—the common language that every fund manager, every ETF provider, every robo-advisor, and now every AI agent uses to define “the market.” Over $18 trillion in assets are benchmarked to MSCI indices. Over $2.3 trillion in ETF assets are directly linked to them. The toll (A) is the licensing fee—a basis point on every dollar indexed. The moat is the switching cost: you cannot move $18 trillion to a different set of indices without dislocating the entire global investment infrastructure.

The first wave (B₁) is the secular migration from active to passive investing. This migration has been underway for two decades. It is driven by an overwhelming body of evidence that most active managers underperform their benchmark after fees. Every year, more capital flows from active funds into index-tracking ETFs and mutual funds. Every dollar that migrates pays a basis point to MSCI. The company does not market passive investing. The world does it on its own.

The second wave (B₂) is artificial intelligence—specifically, the rise of agentic AI in financial services. AI-powered wealth management tools, robo-advisors, automated portfolio construction agents, and agentic workflow systems all consume MSCI index data as their foundational substrate. The MSCI index is to an AI portfolio agent what oxygen is to a combustion engine: the irreplaceable input without which nothing functions.

But here is what makes MSCI a different species: B₂ amplifies B₁.

Follow the causal chain. Agentic AI makes investing more accessible, more automated, more frictionless. A person who would never have consulted a financial advisor now has an AI agent managing their portfolio. What does that agent build with? Index funds. ETFs. The default building block of every algorithmic portfolio is the benchmark—and the benchmark is MSCI. Every new AI-powered investor is a new dollar indexed. Every new dollar indexed generates more data. More data feeds better AI models. Better AI models attract more users. More users generate more indexed dollars.

The loop closes on itself.

* * *

The Hybrid — S&P Global and the Double Engine

S&P Global does not fit cleanly into any single species. It is something more unusual: a business that houses two Freesurfer engines of different architectures within a single corporate structure.

The first engine is the credit rating duopoly. S&P Global and Moody’s together control approximately 80% of the global credit ratings market. Every significant debt issuance—sovereign, corporate, municipal—must pass through one of the two agencies. This is a regulatory near-monopoly with barriers to entry that are essentially insurmountable: no new entrant can manufacture the decades of institutional trust, the regulatory recognition, or the historical database that underpin a credible rating. The toll (A) is the rating fee. The first wave (B₁) is the structural growth of global debt. The second wave (B₂) is AI’s consumption of ratings data. But as with Moody’s, B₂ does not amplify B₁—artificial intelligence does not create more global debt. The two waves are additive. The ratings engine is a Resilient Freesurfer.

The second engine is S&P Dow Jones Indices—the proprietor of the S&P 500 itself. This is the index that defines “the American market” for the entire world. Every dollar that enters VOO, SPY, or any of the thousands of products tracking the S&P 500 pays a licensing fee to S&P Global. The toll (A) is the index license. The first wave (B₁) is the same passive migration that feeds MSCI. The second wave (B₂) is the same AI-driven automation of portfolio construction. And the coupling is identical: AI pushes more capital into passive vehicles, the default destination of which is the S&P 500, which generates more data that feeds AI systems, which push more capital into passive vehicles. The indices engine is a Coupled Freesurfer—a Double Swell.

S&P Global is therefore both species at once. Its ratings arm provides the resilience of two independent waves. Its indices arm provides the amplification of the reinforcing loop. The two engines are diversified not just in their waves but in their architecture—a form of structural redundancy that may make S&P Global the most resilient Freesurfer of the five. If the passive migration stalls, the debt wave continues. If debt issuance slows, the passive migration continues. And if both slow simultaneously, the AI consumption layer provides a third source of volume through both engines. It is a Freesurfer built for all weather.

Where MSCI is the purest specimen of the Double Swell—100% of its business rides the coupled wave—S&P Global is the most complete specimen—it captures every species within a single corporate boundary. For the investor, this raises a question that the taxonomy alone cannot answer: is it better to own the purest expression of the most powerful architecture, or the most diversified expression of all architectures at once? The answer depends on whether one values concentration or resilience—and on the price the market charges for each.

* * *

II. The Reinforcing Loop

In surfing, a double swell occurs when two storms in distant, unrelated regions of the ocean send their energy toward the same reef at the same time. Neither storm knows the other exists. But when their waves converge, something remarkable happens: a wave of two meters meeting another wave of two meters does not produce a four-meter wave. It produces a five or six-meter wave — because the energy of the two systems enters into resonance, and the resulting amplitude exceeds the sum of its parts. This is what big wave riders wait weeks to find: the rare day when two independent forces align on the same point and produce power that neither could generate alone. That is what is happening to MSCI.

In the language of systems thinking—the discipline pioneered by Jay Forrester at MIT and popularized by Donella Meadows and Peter Senge—the Coupled Freesurfer contains a reinforcing loop (often denoted “R” in system dynamics diagrams). A reinforcing loop is a circular chain of causation where each element amplifies the next, and the last element feeds back into the first.

The MSCI reinforcing loop:

More AI agents in finance → more automated portfolio construction → more dollars indexed to MSCI → more revenue and data for MSCI → MSCI data becomes more deeply embedded as AI infrastructure → more AI agents default to MSCI benchmarks → return to start.

The defining characteristic of a reinforcing loop is that it does not require an external push at each turn. Once triggered, each cycle provides the energy for the next. This is what makes the Coupled Freesurfer structurally different from the Classic or Resilient species. Visa’s wave (cash-to-digital) is a powerful secular force, but it is a one-directional current. MSCI’s coupled waves form a circle—each wave pulling the other forward.

The Coupling Phenomenon

In physics, coupling refers to the interaction between two independent oscillators that, when connected, produce an amplitude greater than the sum of their individual amplitudes. Two pendulums swinging independently will each reach a certain height. Connect them through a shared string, and under the right conditions, they can enter resonance—a state where the energy of one reinforces the energy of the other, producing oscillations that neither could achieve alone.

The passive-investing wave and the AI wave are two independent oscillators. Each has its own drivers. Passive investing is driven by fee compression, academic evidence on active management’s underperformance, and regulatory evolution. AI adoption is driven by Moore’s Law, venture capital investment, and the competitive dynamics of the technology industry. Neither wave was created to serve the other.

But they have entered into coupling. The AI wave, by automating investment decision-making, channels more capital into passive vehicles—which are the natural output of algorithmic portfolio construction. The passive wave, by growing the pool of indexed assets, generates more data and more demand for AI-powered analytics. The two oscillators are now connected through a shared string: the MSCI index.

MSCI is not merely the beneficiary of two tailwinds. It is the node where two independent forces converge and amplify each other. That is the structural definition of a Coupled Freesurfer.

* * *

III. Symbiosis in the Ocean

If we have used the metaphor of the ocean to describe the index—the vast body of water on which the Freesurfer rides—then the relationship between AI and passive indexing is best understood through marine biology.

Consider the relationship between the clownfish and the sea anemone. The clownfish lives among the anemone’s stinging tentacles, protected from predators. In return, the clownfish’s movements increase water circulation around the anemone, bringing it more oxygen and nutrients. Neither organism created the other. Neither depends on the other for existence. But together, they thrive in ways that neither could alone.

AI is the clownfish. The index ecosystem is the anemone. AI brings volume—more users, more queries, more automated portfolios flowing into indexed products. The index ecosystem provides the substrate—the structured, proprietary data that AI needs to function in financial services. Each organism benefits from the other’s activity.

And MSCI is the reef—the underlying structure on which the entire symbiosis rests. The reef existed before the clownfish arrived. It will exist after any particular symbiotic relationship fades. But the symbiosis accelerates the reef’s growth by channeling more biological activity through its structure.

The metaphor has limits, as all metaphors do. But it captures the essential insight: the Coupled Freesurfer is not a company riding a wave. It is the structure through which two independent ecosystems have become mutually reinforcing.

* * *

IV. The Sigmoid — Why This Is Not Infinite

In chemistry, an autocatalytic reaction—one whose product accelerates the reaction itself—follows a characteristic curve. It begins slowly (the latency phase), then accelerates dramatically (the exponential phase), then flattens as the substrate is consumed (the plateau). The resulting shape is called a sigmoid—an S-curve. Not an exponential. Not infinite.

The substrate for the Coupled Freesurfer’s reinforcing loop is the stock of actively managed capital. Today, passive investing represents approximately 50% of U.S. equity fund assets. The remaining 50% is the substrate—the raw material available for conversion. The reinforcing loop consumes this substrate: each year, a few percentage points of active capital migrates to passive, and AI is now accelerating that migration.

But the substrate is finite. And crucially, the system contains a natural balancing loop that prevents the reinforcing loop from consuming all of it.

The economists Grossman and Stiglitz described this mechanism in 1980: the less capital allocated to active management, the less efficient prices become, and the greater the opportunities for the remaining active managers. Alpha—the excess return available to skilled stock-pickers—has an inverse relationship with the amount of money under active management. As active capital shrinks, the reward for being active grows. At some point, the reward becomes sufficient to attract capital back toward active management. The system self-corrects before reaching an extreme.

In the language of Le Chatelier’s principle in chemistry: when a system at equilibrium is disturbed, it shifts in the direction that reduces the disturbance. The growth of passive investing disturbs the active-passive equilibrium. The market responds by increasing the profitability of active management, which slows or partially reverses the migration. The system oscillates toward a dynamic equilibrium—it does not collapse toward a corner solution.

Recent academic research supports this. Studies examining the effect of increased passive ownership on price efficiency have found that while information production about individual stocks declines (fewer analyst reports, less search activity), price efficiency remains broadly unchanged—because the remaining active participants intensify their arbitrage activity to compensate. The market’s price discovery mechanism degrades gradually, not catastrophically, and the incentive structure self-corrects.

What does this mean for the Coupled Freesurfer? It means the runway is long but bounded. We are likely in the early-to-middle section of the sigmoid—past the latency phase, entering the acceleration phase, far from the plateau. The substrate (active capital available for conversion) is still abundant. The catalyst (AI) has only just arrived. The coupling between the two waves has only recently become visible. The most powerful phase of the curve lies ahead—not behind.

But the Infinite Investor does not confuse a long runway with an infinite one. The sigmoid eventually flattens. The reinforcing loop eventually meets its balancing loop. And when it does, the nature of the investment changes. The Coupled Freesurfer is the most powerful architecture we have identified—but it operates within the laws of nature, not outside them.

* * *

V. The Complete Taxonomy

We can now classify the Freesurfer species. What began as a single formula—A × B—has revealed itself to be a family with distinct architectures, distinct risk profiles, and distinct compounding characteristics.

Species Architecture Defining Feature Pure Specimen
Classic A × B One toll, one structural wave Visa / Mastercard
Resilient A × (B₁ + B₂) Two independent waves, additive Moody’s (MCO)
Coupled A × B₁↑B₂ Two waves in reinforcing loop MSCI
Hybrid Mixed Resilient (ratings) + Coupled (indices) S&P Global (SPGI)

 

A note on S&P Global. SPGI is a fascinating hybrid because it houses two distinct Freesurfer engines within a single corporate structure. Its ratings business—the credit rating duopoly shared with Moody’s—operates as a Resilient Freesurfer: the growth of global debt (B₁) and AI consumption of ratings data (B₂) are additive but not coupled. Its index business—S&P Dow Jones Indices, the proprietor of the S&P 500 itself—operates as a Coupled Freesurfer: the passive migration (B₁) and AI-driven portfolio automation (B₂) reinforce each other through the same loop as MSCI. SPGI is both species at once. This makes it harder to classify but arguably more resilient than any single-species Freesurfer, because the two engines are diversified not just in their waves but in their architecture.

* * *

VI. The Proof in Real Time

On February 24, 2026, the coupling became visible.

Anthropic—the artificial intelligence company behind Claude—announced new MCP connectors for MSCI and FactSet, alongside partner-developed plugins from S&P Global and LSEG, integrated into its Claude Cowork platform for financial services. The announcement meant that AI agents could now pull MSCI index data, S&P Global Capital IQ Pro data, and FactSet analytics directly into automated financial workflows—equity research, portfolio construction, investment banking deal analysis—without human intermediation.

MSCI rose 2%. S&P Global climbed 3%. FactSet surged 6%. Moody’s gained over 2%.

The irony was rich. Three weeks earlier, the same companies had been punished by the market on fears that AI would replace financial data providers. Anthropic’s initial Claude Cowork launch in January had triggered a selloff that erased approximately one trillion dollars from software and financial data stocks. FactSet had dropped 10%. S&P Global, Moody’s, and LSEG had all seen sharp declines. The market’s first-order reaction was: AI threatens data providers.

The February 24 announcement was the second-order correction: AI does not replace data providers. AI consumes their data in quantities that humans never could. A human analyst makes perhaps fifty data queries per day. An agentic AI system can make five hundred. The toll per query is the same. The volume is ten times larger. Anthropic did not build its own indices. It did not create its own credit ratings. It plugged into MSCI and SPGI because those are the irreplaceable substrates of financial intelligence. The AI needs the data more than the data needs the AI.

What the market priced on February 24 was the surface reaction—a partnership announcement, a few percentage points of share-price recovery. What it has not yet priced is the structural implication: that the volume of data consumption by AI agents will grow non-linearly as agentic AI is adopted across the financial industry, and that this consumption is not merely additive to the existing passive-investing wave but coupled with it—each reinforcing the other in the sigmoid’s acceleration phase.

* * *

VII. The Darwinian Ocean

A natural concern arises: if the Coupled Freesurfer accelerates the migration to passive investing, does the ocean itself become fragile?

The S&P 500—the ocean on which these Freesurfers ride—is not a static body of water. It is a Darwinian machine. Each year, 20 to 30 companies are removed from the index and replaced by stronger ones. Kodak exits, Apple enters. General Electric falls, Tesla rises. The index renews itself through continuous selection. This selection mechanism operates independently of whether capital is allocated actively or passively. Even in a world of dominant passive flows, the index committee continues to cull the weak and admit the strong.

The ocean may experience more turbulent surface conditions as passive flows dominate—more pronounced drawdowns, sharper snap-backs, higher correlations among constituent stocks during stress events. Academic research confirms that stocks with higher passive ownership exhibit greater sensitivity to market-wide movements. The ride may become rougher.

But the ocean cannot evaporate. Its Darwinian composition ensures that it continues to represent the strongest enterprises in the economy. Its self-correcting equilibrium—the Grossman-Stiglitz mechanism, the Le Chatelier response—ensures that the active-passive balance finds a floor before systemic damage occurs. The index is not a fragile vessel. It is a robust, self-renewing ecosystem that has survived every crisis in its seven-decade history not by avoiding storms, but by selecting for the organisms that survive them.

The Coupled Freesurfer does not threaten the ocean. It accelerates the currents within it—currents that feed the reef, which feeds the symbiosis, which feeds the currents. The circle turns.

* * *

VIII. Conclusion

We began with one formula: A × B. We end with three species and a discovery.

The Classic Freesurfer—Visa, Mastercard—surfs a single structural wave. It is simple, powerful, and enduring. The Resilient Freesurfer—Moody’s—surfs two independent waves. It is more robust, because the failure of one wave does not silence the other. The Coupled Freesurfer—MSCI, and the index business within S&P Global—surfs two waves that amplify each other through a reinforcing loop. It is the most powerful architecture we have identified, and it is extraordinarily rare.

The five businesses that pass the Freesurfer test—Visa, Mastercard, S&P Global, Moody’s, and MSCI—are the five toll booths positioned on the structural highways of the global financial system: payments, credit, and indexation. They are not interchangeable. Each occupies a distinct position in the taxonomy, with distinct compounding characteristics and distinct risk profiles.

The sigmoid reminds us that this is not infinite. The reinforcing loop will meet its balancing loop. The substrate will be consumed. The market’s self-correcting mechanisms will find a new equilibrium between active and passive capital. But we are early in the acceleration phase of the curve—past the slow beginning, before the plateau. The runway is measured in years, likely a decade or more.

The Coupled Freesurfer does not promise infinity. It promises the most favorable stretch of a bounded curve—the part where the acceleration is greatest and the exhaustion is farthest away.

The Infinite Investor does not ride waves forever. The Infinite Investor knows which phase of the wave carries the most energy—and stands on the board at exactly the right time.

 

This post is part of the Averaging Up — Mental Models series. For the original framework, read The Freesurfer at averagingup.com.

Recent Comments

No comments to show.

Archives

  • February 2026
  • January 2026
  • December 2025
  • January 2025
  • December 2024
  • September 2023
  • July 2023
  • June 2023

Categories

  • Capital Allocation
  • Capital Allocation
  • Compounding
  • Growth Investing
  • Sizing
  • Uncategorized
©2026 Averaging Up | Powered by SuperbThemes