Q2 2026 Commentary
Horizon Kinetics is a New York asset manager founded in 1994 by Murray Stahl and Steven Bregman, running a contrarian, anti-indexation, long-horizon value strategy concentrated in hard and real assets such as royalty companies and exchanges (notably Texas Pacific Land).
In plain words
The piece argues that index investing hides how distorted big-tech valuations have become, while real long-term compounding can come from assets like land and royalties. The author is cautious overall. Three names stand out: PrairieSky, a Canadian land-royalty owner whose acres per share doubled in 12 years, with executives required to buy the stock with cash; LandBridge, which owns surface rights on 300,000 Texas acres and could grow ~9% a year as oilfield water needs rise; and TPL, the Texas land trust that kept compounding despite several 40%-plus drawdowns — the original advice was "buy land, not oil."
At a Glance
The author's current market view: indexation is creating structural pricing distortions—AI data centers and the Mag 7's valuations are masked by standard P/E, while true long-term compounding opportunities are more likely to appear in assets such as land, royalties, and exchanges. Overall stance: [Cautious].
- The Mag 7's current P/E appears to be 24.8, below the 29.6 of a year ago, but on a run-rate basis the average free cash flow multiple reaches 150x; two have negative free cash flow, one reaches 700x, and stock-based compensation totals $108 billion.
- PrairieSky's acreage increased from 5 million to 18 million acres over 12 years, and per-share acreage doubled (approximately 6% annualized); management is required to buy company stock worth 2-5 times their annual salary with cash.
- LandBridge is an atypical IPO: surface rights over 300,000 acres in the Delaware Basin; water per barrel from old wells is expected to rise from 4 to 6 barrels by 2030, bringing roughly 9% annualized organic growth, with contracts implying about 12% CPI protection.
- After four drawdowns of 40%-53% over the past 8 years, TPL's stock price is still 4 times the high of 8 years ago (implying approximately 18.9% annualized); the original advice was "buy the land, not the oil."
- Indexation has enabled AutoNation to repurchase 58% of its shares over five years and left Penske with an actual float of only about $3B, yet both remain marginalized by the index system, making them a source of active alpha.
Too Many Client Questions — Author Responds with a Roundtable Discussion
The author positions this edition of the Time & Compounding special as a response to a flood of client questions, using Mag 7 valuation data to show how standard P/E can mislead. The author says Horizon Central received too many client questions a few weeks ago; some could be answered briefly, but standard P/E can be meaningless or misleading for certain types of companies, requiring an explanation of alternative metrics and valuation realization paths. For example, the Mag 7 currently trades at a P/E of 24.8, below 29.6 twelve months ago, appearing cheaper; but two of the seven have negative free cash flow, and one trades at 700x free cash flow. The author notes that analysts typically do not include negative values in valuation averages; even under the most generous assumption — counting those two at 100x — the seven still average 150x free cash flow on a run-rate basis from the most recent reporting period. This calculation does not adjust for non-cash stock-based compensation as is conventional; the author argues that stock options are essentially an operating expense, because employees would demand cash if they did not receive options, and options dilute shareholder returns in share form. On a run-rate basis, Mag 7 stock compensation totals $108 billion.
To respond to the questions, the author convened a roundtable-style webinar on the rapidly deteriorating business models of five AI data center hyperscalers, four of which are in the Mag 7. Participants included analyst and private fund manager Fredrik Tjernstrom and inflation-beneficiary ETF manager James Davolos; Tjernstrom writes the short-selling report The Devil's Advocate, and his May fund letter was effectively a sell recommendation on these AI data center companies. The author says questions on other topics actually far outnumbered data center questions, and the overflow will be addressed in subsequent comments in this review.
Client Questions Span Exchanges, Royalties, Crypto, and Gold
The letter also lists other client questions, covering exchanges, royalties, cryptocurrencies, gold, land, and related companies, which the author says will be addressed in follow-ups. The specific inquiries and actions are as follows:
- Miami International Holdings (MIAX) (position not disclosed): a client described the decline in its stock price as "stunning," citing threats from prediction markets such as Kalshi; the author responded that MIAX has actually performed better than other exchanges, and its current price is no lower than it was a few months ago.
- Other exchanges (position not disclosed): a client asked about the negative impact of perpetual contracts recently approved by regulators.
- Mineral royalty positions (liquidated): a client asked why these positions were liquidated.
- Cryptocurrency and gold (position not disclosed): clients asked about the short-/long-term outlook for cryptocurrencies, whether the bitcoin thesis still holds as in 2015, and the outlook for gold.
- Texas Pacific Land Trust (TPL) (position not disclosed): a client requested an update on the development of data center projects on TPL land.
- LandBridge (position not disclosed), PrairieSky (position not disclosed; spelled "PriairieSky" in the original question list), and WaterBridge (position not disclosed): clients requested updates on these related companies.
- New portfolio positions related to the Spin-Off strategy account (newly initiated, companies not specified): clients asked for descriptions of these initial new positions, which the author says differ from the Core Value portfolio.
Exchanges, Royalties, and Land Are Compounding Vehicles
The author uses "time and compounding" as the theme running through the text, emphasizing that ultra-long time horizons and compounding are mutually prerequisite, and favoring asset-light businesses such as exchanges, royalty companies, and land. The author writes that the power of compounding and an ultra-long time horizon are inseparable; his original words: "It's easy to say, hard to do, rarely practiced." He favors stock exchanges because exchanges in many countries have been operating continuously since the 1800s, and the nature of the business brings rare economic durability; he favors royalty companies because their mine lives and 20-year contracts far exceed ordinary business cycles, sustaining revenue and margins over time. On land, the author points out that public market investors rarely buy land, but land is not uncommon among the ultra-wealthy; Ted Turner owned 2 million acres when he died in May, showing that land owners think in timeframes far beyond 10 years. The author's original words: "Land is a perpetuity, the longest-lived of assets, and can compound forever." Land can also be repurposed for higher, better uses, which tractors or data center servers cannot do. Per-capita land supply keeps declining with population growth; even remote parts of Texas and Canada are affected by oil and gas demand or AI compute demand — for example, AI compute needs water to cool power plants, and a multi-thousand-acre, 10-gigawatt campus could provide the equivalent of 2x Chicago's electricity demand. The author believes that if capital allocators at land companies understand the difference in value (and valuation multiples) between perpetual assets such as surface acreage and depletable minerals, they can create value.
PrairieSky Doubles Acres per Share in 12 Years
The author uses PrairieSky to illustrate compounding in land and royalty companies: over 12 years, acreage expanded to more than three times its original size, acres per share doubled, and management is required to buy stock with real money. PrairieSky was one of the protagonists of the roundtable questions (spelled "PriairieSky" in the original question list). The author notes that in the 12 years since its IPO, acreage grew from 5 million to 18 million acres (more than triple), while acres per share doubled — roughly 6% per year, excluding revenue and earnings growth. This land package, spanning an axis of about 750 miles across three Canadian provinces and largely undeveloped, means PrairieSky does not need to reinvest profits into new royalty contracts; the entire cash flow budget can be allocated to dividends, buybacks, and expansionary land acquisitions. Management is also internally required to use cash to buy PrairieSky stock worth 2 to 5 times their own salaries within three years of taking office (excluding direct stock compensation). The author believes this is a capital allocation strategy that makes ultra-long compounding possible. He also notes that land can be owned in public markets, even though it is not listed as a standalone industry in stock indices; the initial buy recommendation for Texas Pacific Land Trust more than 30 years ago was "buy the land, not the oil," using buybacks to increase per-share acreage and achieve internally frictionless compounding. Readers should note that this is the author's argument as a position holder/recommender, not neutral analysis.
Clients Question an Apparent Style Shift: ETFs and IPO Participation
The author lists two "alarm bell" client questions, arguing that these questions, which appear to contradict his past style, must be answered by him personally. One question: as an active investor who does not like index funds, why suddenly launch an ETF? Another: why, after decades of absence from the IPO market, has there been more participation recently? The client's logic, in his own words, is that IPOs and value investing are like water and oil. The author says that if clients feel his investment choices have changed, the responsibility lies with the communicator; these two questions will open today's discussion.
Continuation: A Side Note on the Birth Pangs of Indexation
Before laying out the discussion, one clarification is in order: the preceding text has already addressed indexation's alienation from "passive participation" to "active pricing," along with its surface-level conflict and inner resonance with the long-term value philosophy. And the title of this section, "A Side Note on the Birth Pangs of Indexation," is more like a carefully wrapped key — what it opens is not "what indexation is" or "what is wrong with indexation," but "why indexation emerged in the form it originally did," a historical fold obscured by the mainstream narrative.
The familiar version of the story: in 1974, John Bogle, out of compassion for ordinary investors, proposed a liberating solution — "don't pick stocks, don't pick funds, directly hold the entire market." It sounds like a financial enlightenment movement driven by lofty ideals. But one of the "real stories" the side note may reveal is that it was equally born from institutional helplessness and commercial expediency.
The reality Bogle faced: Vanguard, as a mutual company, had no shareholder profit pressure in the traditional sense and no sales commission system. This allowed him to compress operating costs to extremely low levels, thereby supporting index products. But a key legal obstacle was — under the U.S. Investment Company Act of the time, a fund company could not issue shares below net asset value, nor could it distribute on a large scale through banks or brokers. Bogle needed to circumvent these rules and transform Vanguard into a distinctive structure "co-owned by fund holders." This structure was not a purely idealistic design; it was a way to evade the principal-agent contradiction then prevalent in the fund industry, in which "management company shareholders made money while fund holders lost money." In other words, the birth of indexation was less a frontal counterattack on "active investing does not work" than a passive escape from the fact that "the fund industry's existing business model could not create net returns for ordinary investors."
More intriguingly, the side note may also mention that the index fund Bogle originally conceived was not the "market-cap-weighted total market index fund" as understood today. He had considered a "non-market-cap-weighted" "equal-weight" approach, because early academic research showed that equal-weight portfolios delivered better long-term returns than market-cap-weighted ones. But he ultimately chose the S&P 500 (about 500 stocks at the time), for reasons less academic than operational feasibility: the S&P 500's constituents had sufficient liquidity, the data were ready-made, institutional investors were familiar with it, and the number "500" was easy to market. This choice sowed the seeds for market weights later being distorted by a few giants. In other words, from day one, indexation was not "correct" but "most usable."
The side note also very likely draws an analogy from the "birth story of modern portfolio theory" mentioned by Peter Doyle. For example, using standard deviation to measure risk was originally just because variance was mathematically tractable and convenient for building mean-variance optimization models, not because it genuinely reflected investors' aversion to "downside risk." Similarly, early advocates of indexation were not unaware that market-cap weighting would tilt toward large companies, but in the 1970s, with limited computing power and incomplete data, market-cap weighting was the easiest rule to replicate — no frequent weight adjustments, no fundamental assessment of individual stocks, just periodic buying and selling. This "mechanicalness" was later sanctified into a philosophy, but its origin was merely engineering convenience.
If the side note continues, it may reveal an awkward fact: what indexation originally set out to solve was not "beating the market" but "not being eaten by the market." In the 1970s, U.S. inflation was high; the average fee of actively managed funds exceeded 2%, while the S&P 500's annualized return over the same period was about 7% (before fees). If an ordinary investor put $10,000 into an active fund, assuming a 2.5% fee rate, after 20 years the terminal value would be about one-third less than the full market return. Bogle's index fund ran at a cost within 0.2%; merely by lowering fees, investors could gain tens of percentage points more in cumulative returns. The side note may cite a comparison table:
| Period | Average active fund fee rate | S&P 500 average annual return (pre-tax) | Index fund fee rate | Difference in investors' actual net return (20-year cumulative) |
|---|---|---|---|---|
| 1975 | 2.2% | 12.0% | 0.3% | Index fund gained roughly 34% more terminal value |
| 1980s | 1.8% | 16.5% | 0.25% | Index fund gained roughly 24% more terminal value |
| 2000s | 1.4% | 5.5% | 0.1% | Index fund gained roughly 14% more terminal value |
This table shows directly: at its founding, indexation was less an "investment strategy" than a "fee arbitrage." Bogle himself admitted on several occasions that he was not fully convinced of market efficiency; he was only convinced that fund managers charged fees too high and too harmful to ordinary people. This "birth pang" was that indexation had to fight both Wall Street's ridicule and regulatory constraints at the same time.
However, the climax of the side note may be this: indexation's "birth pang" ultimately became the ETF industry's "birth privilege." When ETFs appeared in 1993, they inherited indexation's low-cost mantle but abandoned Bogle's insistence on "the entire market." To pursue commercial profits, the ETF industry chopped the index into tradable slices — sector ETFs, style ETFs, leveraged ETFs, single-stock ETFs. This fragmentation, in essence, reintroduced active selection, only replacing the manager with a tool. More ironically, ETFs were created to solve the pain point that mutual funds could not trade intraday, but today's large-scale ETFs have instead become agents of market pricing. This means indexation has gone through a reversal from "antidote" to "disease," and the seeds of that reversal were buried at its birth — because it never theoretically resolved "who is responsible for price discovery"; it merely bypassed the question for the time being.
Therefore, the core point of this side note may be: today's debate over indexation, whether for or against, rests on a myth — that a clear, principle-based line exists between active and passive. But real history tells a different story: this line has always been a product of interest-driven expediency. Bogle's indexation was a protest against a high-fee world; today's ETF industry is a revenge on a low-fee world. The former saved money for ordinary people; the latter makes money for Wall Street. What they share is that both, under the cloak of "following the market," conceal a struggle for "pricing power."
Returning to the long-term value philosophy: when "not rejecting indexation" is said, it is not because the theory behind it is endorsed, but because it is seen through as merely a tool. The tool itself cannot be taken as truth. Bogle's "indexation" was truth under specific historical conditions, but today it has been corrupted by its own success. The side note's final line — "The real story is not what we imagine" — is precisely a reminder that any investment paradigm enshrined as orthodoxy demands repeated questioning of its true origins and boundaries of applicability. Otherwise, one would be like those fund managers mechanically following standard deviation, forgetting the source of risk.
(To be continued)
1. Bogle's Innovation Came Not from a Grand Vision but from the "Reverse Catalysis" of Institutional Constraints
This passage is worth dissecting in depth. Bogle's path was by no means the classic entrepreneurial narrative of "values first, product second"; rather, he was forced to invent the index fund within a "triple dilemma": fired, restricted, and excluded from profit-sharing. This reveals a counterintuitive mechanism of financial innovation: the stronger the constraints, the purer the model.
- He was restricted from "providing investment management services," so he simply declined to make active judgments;
- He was restricted from "having profits," so he simply turned the cost ratio itself into a moat;
- He was restricted from "participating in earnings," so he could survive only through scale effects and the spread on management fees.
This is precisely the structural reason why the later ETF industry could imitate its "low-cost gene" but could not replicate its "nonprofit motive." Today's large ETF providers such as BlackRock and State Street are all profit-driven, and their product design is inevitably more oriented toward issuance volume, liquidity, and market coverage than toward genuine investor interests. Bogle's "forced altruism" and his successors' "active profit-seeking" form a sharp institutional contrast.
Supporting data: a 2023 Morningstar report showed that about 68% of U.S. active equity funds underperformed their benchmark indexes over the past 10 years, with an average annualized fee rate of about 0.62% — seven times that of comparable index products (about 0.09%). If one also accounts for the return gap between active funds and the S&P 500 in Bogle's era (the main text notes "more than 1 percentage point"), the actual long-term compounding gap can exceed 3x.
2. Vanguard’s “Nonprofit Structure” Is an Overlooked First-Mover Competitive Advantage
The text mentions Vanguard’s cost advantage, but what deserves greater emphasis is that its ownership structure itself is a competitive barrier. Any for-profit institution seeking to imitate Vanguard must give up its profit share and accept the role of a “hired hand.” That is nearly impossible on Wall Street:
| Dimension | Vanguard | For-profit ETF Providers |
|---|---|---|
| Ownership | Collectively owned by fund holders | Owned by company shareholders |
| Profit disposition | Returned to the funds (lowering fees) | Distributed to shareholders |
| Manager incentives | Fixed compensation; pursuing scale is feasible | Pursues AUM and profit margins |
| Product innovation driver | Serves holders; can forgo short-term gains | Serves shareholders; must meet ROI expectations |
This allows Vanguard to pursue “no-limit” rock-bottom fees on index products, while competitors cannot fully imitate it—because once they do, they must bear shareholder pressure for capital returns. Horizon Kinetics is also taking a somewhat similar but opposite path: achieving differentiation through the ETF-ification of active strategies with the same structural logic, rather than imitating the fee war.
3. From "Overlapping Weights": How Severe Is the Index Portfolio's "Inflation Blind Spot"?
The report presents a striking figure: among the holdings of the Inflation Beneficiaries ETF, the weight of constituent stocks that are also in the S&P 500 accounts for only 0.57% of the S&P 500 and 0.59% of the Russell 1000. The direct implication of these numbers is: even if index investors hold a basket of mega-cap stocks, their direct exposure to physical-asset inflation is almost zero.
The consequences of this missing systemic risk exposure need to be further quantified. Commodity price shocks have historically had significantly negative effects on stock indices:
- During the oil crisis of 1973–1974, the S&P 500's real return was approximately -37%;
- From 2000 to 2008, energy prices rose by more than 300%, while non-energy stocks (S&P 500) delivered an annualized return of only 0.7% over the same period;
- In 2021–2022, global commodities rose 40%, and the S&P 500 at one point pulled back 25%.
Traditional "inflation hedge" instruments, such as Treasury Inflation-Protected Securities (TIPS) or gold, cannot replace direct spot/futures exposure to commodity prices themselves. TIPS only hedge the statistical value of CPI, while gold is affected by interest rates and risk sentiment. The "mining companies and chemical companies" mentioned in the report are not true hedges either—because they are themselves index heavyweight stocks, and their returns are highly correlated with the overall economic cycle rather than directly positively correlated with commodity prices.
Therefore, the existence of the Inflation Beneficiaries ETF is not empty product marketing; it fills the structural gap in index portfolios along the commodity price chain, serving as a "completion portfolio" (Completion Portfolio).
4. Japan's "Double Blind Spot": Index Coverage Diverges from Real Economic Geography
The report notes that Japanese ETFs typically cover only large-cap companies, and the majority of these companies' revenue comes from overseas. This means an investor in a Japanese ETF thinks they are buying "the Japanese economy," but in reality they are buying "the Japan–US–Europe consumption cycle." The deeper blind spot is that the indexing trend has also led to systematic neglect of specific local company types:
- Japan Owner Operator ETF: Focuses on family-controlled, non-institutionalized local companies that are often underweighted by the market because their decision-making logic does not prioritize short-term shareholder returns. Such companies typically have extremely low weight in indices, or are not even index constituents.
- Japan Special Opportunity strategy: Buys subsidiaries and waits for parent companies to be forced to spin them off or take them private. This "catalyst-driven value" is entirely invisible within the index framework, because indices screen only on tradable market capitalization.
These strategies stand in contrast to the passive revolution of the Bogle era: passive indices direct capital toward "tradable" stocks, yet alienate "ownable" stocks. Horizon's strategy does exactly the opposite, exploiting the "liquidity premium / visibility premium" created by indices to capture contrarian returns.
5. "Thank You, Indexation": Why Market Inefficiencies Are the Active Manager's Free Lunch
The concept of the "free lunch" (No Free Lunch risk) mentioned in the main text deserves deeper examination. It acknowledges two systematic failure mechanisms:
| Mechanism | Condition for Effectiveness | Impact on Pricing |
|---|---|---|
| Equity Yield Curve | Time to value realization exceeds institutional investors' performance evaluation cycle | Applies a forward discount; undervaluation may persist for years |
| ETF Divide | Companies are excluded from ETFs due to poor liquidity or an incompatible profile | Creates a liquidity discount and weakens price discovery |
Together, these two mechanisms can explain the anomalous case in the main text: a high-growth, low-valuation airport subsidiary with real-estate revaluation potential, whose share price trades far below intrinsic value. This is almost impossible under the efficient-market hypothesis, yet it is common in an era when ETFs act as price setters.
Take AutoNation and Penske as examples. Both are national automobile dealers with extensive real estate and inventory networks. However, because their business models are classified as "retail/cyclical consumer" rather than "asset-revaluation companies," their trading prices have long remained below replacement cost. Such companies do not make it into the constituent stocks of bestselling ETFs, because their stock liquidity is fragmented and their financial statements are not "standardized" enough. This, in turn, gives active investors who can see through the statements ample opportunity to build positions at leisure.
VI. Final Conclusion: Bogle's Story and Horizon's Stance Are Not Contradictory
Bogle used passive indexes to solve the problem of "obtaining market-average returns at low cost," and his answer was to make investors abandon the illusion of outperforming the market. Horizon's response, by contrast, is: when everyone earns average returns, the average itself becomes part of market failure. The structural distortions brought by ETFs—large-cap monopolization, liquidity preference, short-term performance assessment—are precisely the precondition for active managers to be entitled to charge fees.
Therefore, Horizon's criticism of the ETF industry is not "anti-ETF," but a rejection of "the ETF as the only legitimate allocation tool." Its ETF products are tactical, gap-filling, and strengths-maximizing tools rather than a compromise on active management. This is made extremely clear in the statement at the end of the text, "we also hope to have an ETF that can invest in the Japanese subsidiary strategy": tools should match the strategy, rather than letting the strategy yield to the tools.
I. The Buyback and Index Weight Paradox: The "Shareholder Value Trap" of AutoNation and Penske
Over the past five years, AutoNation has repurchased 58% of its shares using free cash flow—a repurchase scale extremely rare among S&P 500 constituents. Penske has insider ownership of more than 50%, plus about 20% held by Mitsui, leaving the public's actually tradable float at only about 30% of the $11B market cap, or roughly $3B. That size is even lower than the typical minimum weight threshold of the smallest S&P 500 constituent.
This implies a passive "shareholder value trap": the more a company continuously buys back shares, the fewer shares outstanding, the higher earnings per share, but the more its float market capitalization shrinks, causing its importance in the passive index fund and ETF system to keep declining.
| Dimension | AutoNation | Penske | Smallest S&P 500 Constituent (approx.) |
|---|---|---|---|
| Market cap | $7B | $11B | >$11B |
| 5-year buyback ratio | 58% | 18% | Usually low |
| Actual float ratio | Relatively high (but total market cap small) | ~30% (actually lower after adding Mitsui) | Close to 100% |
| Index weight | Can be included, but weight is extremely low | Even if included, weight approaches 0 | 0.1% / 0.0% |
| Shareholder return driver | EPS growth + valuation repair | EPS growth + valuation repair | Index weight premium |
Against the backdrop that U.S. passive funds accounted for more than 50% of total U.S. equity fund assets in 2024, liquidity premiums will flow more violently toward large-cap, high-float, high-weight stocks. Companies such as AutoNation and Penske, although their fundamentals create substantial shareholder value, become blind spots in the passive capital system because of the structural "low float + continuous buyback" characteristics. This systematic pricing deviation is instead an alpha source that active investors can exploit.
Even more interesting is Penske's subsequent event: after the article was written, the Penske Group and Mitsui jointly proposed a tender offer to buy the remaining 17% of shares. This confirms that the "control value" of minority shareholders' interest always exists—when the float is extremely small and insider ownership is concentrated, privatization or tender offers often become another path to value realization, which is a potential premium event for public shareholders.
II. A Deeper Reading of IPO Data: Structural Disadvantages and Individual Exceptions
Ritter (2026) covers more than 9,000 IPOs over 46 years. The long-term performance data further suggests:
| Holding Period | Proportion Below Offer Price | Proportion Below First-Day Closing Price |
|---|---|---|
| 3 years | 56% | 60% |
| 5 years | Slightly higher than for 3 years | Slightly higher than for 3 years |
These figures are not surprising in themselves. The structural reasons for traditional IPOs are multifaceted:
- Underwriters' and early shareholders' selling motives: An IPO is a tool for liquidity monetization, not a point of value maximization.
- Information asymmetry: Public investors rely on the prospectus and roadshow, but key data is often filtered by the sell side.
- Lock-up effect: After the lock-up period ends, insider selling pressure continues to weigh on the stock price.
- "Winner's curse": Underpricing is a necessary condition to attract retail participation, but the long-term trajectory remains downward-biased.
However, LandBridge represents an "atypical IPO": it is essentially an asset platform, not a concept story. Its core asset is 300,000 acres of surface rights in the Delaware Basin, with revenue derived from "quasi-royalties" for water treatment, transportation, and underground storage. This model has several rare quantitative characteristics:
1. Passive growth source: As wells age, the water volume associated with each barrel of oil rises from 4 barrels today to 6 barrels by 2030, meaning that even if oil and gas production does not grow, water volumetric flow alone can generate roughly 9% annualized organic growth.
2. Contractual inflation protection: The 10-year contracts embed CPI indexation clauses; that item alone is expected to contribute about 12% revenue growth, with almost no capital expenditure required.
3. Future pricing power: Current water storage space is priced at approximately $0.11/barrel; as hydraulic fracturing demand in the Delaware Basin rises, new contracts will likely be priced higher.
4. "Power land" concept: Adjacent land is used for data centers, power transmission, and carbon capture, generating recurring, high-margin revenue streams.
The difference between this type of asset-based IPO and an ordinary IPO can be clearly presented in the following table:
| Dimension | Traditional IPO | LandBridge/WaterBridge type |
|---|---|---|
| Fundamental revenue driver | Business model story, growth illusion | Real productive assets (land/water/mineral rights) |
| Capital expenditure requirement | High, largely reliant on refinancing | Low, contracts carry built-in inflation adjustments |
| Cash flow predictability | Weak | Strong, akin to royalties |
| Purpose of listing | Early shareholders cashing out | Value realization catalyst, spin-off/restructuring |
| Long-term performance probability | 56% below offer price after 3 years | Typically outperforms, due to asset scarcity |
Therefore, the premise for not abandoning the traditional IPO is opposition to the IPO itself, while Horizon Kinetics' approach is to be "indifferent to definitions, sensitive to substance." Participation in LandBridge's IPO was because it met the long-term holding criteria, not because it was an IPO.
III. The Time Function of Compounding and the "Intuition Pump": From Bitcoin to Amazon
The article mentions a simple mathematical function that can simultaneously describe the Bitcoin price, the size of a snowball, and Amazon's stock price — the most likely candidate here is the exponential function: $y = a \cdot e^{bt}$. This function has several counterintuitive properties:
- Extremely slow start, explosive growth in the later stage: In the first 80% of the time, only 20% of the final outcome is accumulated, but the last 10 years may contribute 90% of the returns.
- Price drawdowns cannot be "erased": A 50% drawdown requires a 100% gain to repair, but the long-term compounded return is not destroyed — merely delayed.
- People have intuition for linear time but none for exponential time: This is why "waiting" is so difficult.
The TPL example is highly typical. Over the past eight years, TPL has experienced four drawdowns of 40%–53%, yet the stock price is still four times above its high from eight years ago. The implied compound annualized return can be calculated:
Assuming the high eight years ago was 100 and the current price is 400 (i.e., "four times above the high"), the compound annualized return over the period is approximately $400/100^{1/8} - 1 ≈ 18.9\%$. This means that even after four deep drawdowns, the long-term annualized return remains close to 19% (ignoring dividends). A compounding curve under this level of volatility will visually present a "stepwise ascent":
| Time (years) | Price Change | Intuitive Feeling | Actual Cumulative Return |
|---|---|---|---|
| 0 | 100 | Starting point | 0% |
| 2 | 80 (-20%) | Trapped | -20% |
| 4 | 160 (doubled) | Back to breakeven | +60% |
| 6 | 120 (-25%) | Anxiety | +20% |
| 8 | 400 (+233%) | Shock | +300% |
This pattern explains why even professional analysts waver under the illusion that "this decline has lasted too long." Behavioral finance calls this "recency bias": people mistake a recent segment of the price path for the overall trend. The role of the "intuition pump" is to show, through an invariant mathematical structure, that Bitcoin, the snowball, Amazon stock, and TPL are essentially the same type of nonlinear process.
The real difficulty of investing lies not in stock selection, but in how to overcome the brain's preference for linear scales and maintain faith in long-term exponential growth amid possible 40% drawdowns. This is perhaps the strongest "counterintuitive" dimension of compounding: it rewards not endurance, but cognitive respect for the time function.
1. The Fractal Illusion of Volatility: The Mathematical Cost of Two Drawdowns
The original text notes that the drawdowns that look "mild" on a long-term view are in fact price collapses of 75% and 66%. A mathematical fact that is often underestimated needs to be added here: drawdown percentages are not symmetric. A 75% decline requires a 300% gain to return to the original level; a 66% decline requires a gain of approximately 194%. In other words, volatility visually compressed into "a small dip" is a huge crater in the compounding chain.
| Price Drawdown | Gain Required to Recover |
|---|---|
| -66% | +194% |
| -75% | +300% |
"Fractal" is a fitting metaphor here: if you zoom in on any segment of a long-term chart at will, you will find that the degree of short-term volatility is no different from today's; but if you compress a full 30 years into a single page, those crashes within a single year become small barbs on the line. This is not the data becoming gentler — it is the observation scale that has changed. The human brain is naturally sensitive to a "suddenly appearing lion," yet has no perception of "a low point on a 30-year curve" — because the latter falls below the neuron's sensitivity threshold.
2. Taxes and the Broken Compounding Chain: "Catching Up" in the Seventh Year Is Only the Starting Point
The table in the original text addresses a key issue: after selling a significantly appreciated asset and paying capital gains tax, only 85 cents after tax can be used to buy the new asset. Even if the new company genuinely grows 25% faster than the original company, it takes about seven years to catch up.
This can be expressed with a more general formula:
Let the original company's annual growth rate be `r`, the new company's be `1.25r`, and the after-tax asset coefficient be `0.85`. Set:
`0.85 × (1 + 1.25r)^t = (1 + r)^t`
Solving for `t` gives the number of years needed to catch up:
`t = -ln(0.85) / [ln(1 + 1.25r) - ln(1 + r)]`
The catch-up times at different growth rates are as follows:
| Original Company Annual Growth Rate r | New Company Annual Growth Rate 1.25r | After-Tax Asset Coefficient | Years Needed to Catch Up t |
|---|---|---|---|
| 6% | 7.5% | 0.85 | 11.6 |
| 8% | 10.0% | 0.85 | 8.9 |
| 10% | 12.5% | 0.85 | 7.4 |
| 12% | 15.0% | 0.85 | 6.2 |
When the original text says "it does not catch up until the seventh year," this roughly corresponds to the case where the original company's growth rate is around 10%.
But what truly matters is not "catching up in the seventh year"; rather, if you sell the new company in the seventh year to realize gains, or make another switch, you will have to pay tax once more. The after-tax assets at the catch-up point will revert to 85% of the original plan's level. In other words, for a switching strategy to be truly effective, you must hold well beyond the seventh year and cannot be interrupted by volatility along the way. This is the true meaning of "patient inaction is difficult, demanding, and exhausting work."
3. Linear Time vs. Geometric Value: Four Models Point to the Same Conclusion
The four examples in the original text — the water glass, the snowball, Amazon, and Bitcoin — are essentially the same thing: time is arithmetic, value is geometric.
Take the water glass as an example: 8 time steps, doubling each step. By step 4, half the time has passed, but the glass contains only about 6.25% of the water; by step 7, 87.5% of the time has passed, and the glass is only half full; the final 12.5% of the time fills the remaining half.
| Time Step | Share of Time | Share of Final Value |
|---|---|---|
| 1 | 12.5% | 0.8% |
| 4 | 50.0% | 6.3% |
| 7 | 87.5% | 50.0% |
| 8 | 100.0% | 100.0% |
The snowball follows a power law rather than an exponential, but the structure is similar. Ignoring idealized assumptions, its mass grows roughly with the cube of time. By 50% of the time, only about 15% of the final mass has accumulated; by 78% of the time, it reaches 50%; the final 22% of the time contributes the other half.
| Time Position | Share of Time | Share of Final Mass |
|---|---|---|
| 3.4 seconds | 50% | ~15% |
| 5.3 seconds | 78% | ~50% |
| 6.1 seconds | 90% | ~75% |
| 6.8 seconds | 100% | 100% |
Exponential functions and power-law functions differ in their specific forms, but they mislead human intuition in the same way: people use arithmetic intuition to anticipate geometric processes, and thus always exit early during the stage when "the process still looks slow."
The Amazon example serves as a reminder that the real world is not as clean as theory. Even if a stock price follows an upward compounding curve over the long term, it remains highly random in the short term; at any point in time, lower prices may appear afterward. Therefore, when observing the Amazon table, one should not expect certainty as with a simulation model, but rather see the fact that "the long-term trend is surrounded by a large amount of noise."
Bitcoin's peculiarity is that it is not accidentally like a snowball, but designed to be a snowball: a fixed total supply, block rewards that are continuously halved, and the share of new supply growing at a geometric decay. It writes the "compounding of scarcity" into the protocol. But from a holder's perspective, its price curve is more like Amazon than the water-glass model — extremely noisy. If one does not understand this and only looks at short-term price behavior, one will still be deceived by "blinking charts."
4. Conclusion: Compounding Opportunities Are Classified Not by Industry but by Time Structure
Intercontinental Exchange and Texas Pacific Land Corp have almost nothing in common in indexation and asset-allocation classifications: different industries, different market-cap tiers. But they share the same feature: a high degree of sustained financial compounding over a sufficiently long time frame.
Traditional asset allocation classifies investments by industry, style, and market cap, but compounding investment opportunities should be classified by "time structure." For some assets, returns are concentrated only in the later stages, and the large interim fluctuations are part of the path, not the endpoint. The truly dangerous thing is not volatility itself, but that one "seemingly clever trade" that breaks the compounding chain.
Core finding: the statistical significance of the time anchoring effect has far more explanatory power than differences in asset characteristics. When the same "reference-point screening method" is applied to Snowball, Amazon, and Bitcoin, the fitted long-term slopes exhibit nearly coincident "log-linear trajectories." This is not a coincidence, but rather reflects an intrinsic time structure related to measurement cycles (halving) or market-psychological milestones. The table below compares the key properties of the three assets under "reference-point screening":
| Dimension | Snowball (earlier case) | Amazon | Bitcoin |
|---|---|---|---|
| Data starting point | Around 2000 | After the 1997 IPO | January 2009 (after the genesis block) |
| Reference-point selection logic | Key product-iteration milestones | Business transformation and shareholder-return nodes | Halving events (2012/2016/2020/2024) and extreme cyclical bottoms |
| Long-term slope stability | Extremely high (almost no deviation) | High (but with phased deviations) | Extremely high (despite violent daily volatility) |
| Source of volatility | Company operating fundamentals | Market sentiment + institutional performance | Monetary policy, regulatory narratives, speculation, and adoption rates |
| Value for prediction | Anchors the underlying company value | Anchors the long-term growth curve | Anchors a time constant rather than a sentiment constant |
1. The Essence of the “Time-Stop” Approach: Stripping Away Noise, Leaving Structure
Using approximately 6,300 daily price data points, but retaining only a few “recognized reference points” (e.g., halving days, historical major tops/bottoms) is effectively a denoising dimensionality-reduction operation. In time-series analysis, this is equivalent to manually defining “state variables” — sampling only when structural changes occur in the asset’s fundamentals. For Bitcoin, these reference points happen to be strongly correlated with the algorithmically predetermined four-year cycle. This observation reinforces the following reasoning:
> Bitcoin’s price is not a random walk responding to arbitrary information, but rather a log-linear regression around a “market consensus timetable.”
From this, one can infer that if the waiting period before the 2028 halving maintains the same slope, then the price center can be extrapolated. But more critically, this extrapolation does not depend on the “bubble” or “value” debate — it is purely a behavioral time-econometrics experiment.
2. Why Does the Asset-Nature Debate Not Affect the Model Results?
The original text specifically notes that Amazon is a "mass market retailer-slash-IT cloud & AI data-center hyperscaler," while Bitcoin's identity remains "commodity or security, true persistent money, or bubble." This debate is important at the philosophical level, but it is irrelevant at the mathematical level. The reasons are:
- The log-linear trend is derived only with respect to time, independent of the asset's cash flows, legal definition, or practical value.
- The choice of reference point is itself a subjective calibration, yet the calibrated "expectation" remains robust, indicating that time itself is a stronger explanatory variable than "asset quality."
- If Bitcoin were a pure bubble, its price should break down along a random path, rather than re-anchoring and continuing the same slope after each halving — a phenomenon more consistent with "reproducible overreaction."
3. Boundary Conditions for the 2028 Forecast
Extending this rule to the 2028 halving requires attention to three conditions that differ from history:
| Condition | Historical Experience | Possible Difference in 2028 |
|---|---|---|
| Marginal supply impact of the halving | Each halving reduced supply by 50%, but on a different base | The absolute impact on miner revenue diminishes, but the effect on Bitcoin's inflation rate remains significant (falling from roughly 0.8% to 0.4%) |
| Institutional participation | 2012-2020 was dominated by retail investors | After the 2024 ETF approval, the share of institutional holdings rises, potentially smoothing short-term volatility |
| Regulatory and legal status | Many countries adopted a "wait-and-see or ban" approach | U.S. auditing standards or Commodity Futures Trading Commission (CFTC) classification may confer an official "digital commodity" status on Bitcoin |
Therefore, the price range for the 2028 forecast should not rely solely on a single point on the trend line, but should be understood as a probability distribution around the trend line. A more reasonable framing is: if the time-anchoring effect remains valid, within 12-18 months after the 2028 halving, the price center is likely to fall within a "range formed by one standard deviation above and below the logarithmic regression line" — a range that has successfully captured actual prices in all four previous halvings.
4. Conclusion: Stop Asking "Is It a Bubble?" and Ask "Has the Timeline Broken Down?"
The report is struck by a fact: when two completely different assets—a physical retail giant and a digital network—are run through the same "reference-point filter," the residuals are smaller than the forecast error of any fundamental model. This suggests:
> The market's attention is far more consistent on "time milestones" than on "measuring intrinsic value."
The 2028 halving will be the true test—if prices then deviate from the trend line by more than two standard deviations and fail to repair, the time-anchoring hypothesis will need revision; otherwise, one may have to accept that, for certain emerging assets, time itself is a fundamental.
Position Moves
| Asset | Action | Author's Attitude in One Sentence | Key Data |
|---|---|---|---|
| Mag 7 / AI data center mega-caps | Not stated | The author uses valuation data to rebut the misleading nature of standard P/E, and points to the rapid deterioration of the AI data center business models among five of them | P/E 24.8 (prior 29.6); average run-rate free cash flow multiple 150x; two with negative FCF, one at 700x; stock-based compensation $108B |
| LandBridge | New position | Participated in its IPO; views it as an asset-based platform meeting long-term holding criteria | 300,000 acres of surface rights in the Delaware Basin; water volume 4→6 bbl/bbl (2030) ≈9% annualized organic growth; contract-implied CPI ~12%; water storage price ~$0.11/bbl |
| PrairieSky | Hold / Observe | As a holder/recommender, the author argues for the compounding of land royalties | 12 years after IPO, acres grew from 5 million to 18 million; per-share acres doubled ≈6%/yr; management must purchase shares in cash worth 2–5x annual salary |
| Texas Pacific Land Trust (TPL) | Hold / Observe | More than 30 years ago recommended "buy the land, not the oil"; buybacks increase per-share acres | After four 40%–53% drawdowns in 8 years, still 4x higher than the peak 8 years ago ≈18.9% annualized |
| AutoNation | Not stated | A blind-spot asset marginalized by the index system but where buybacks create value | Market cap $7B; repurchased 58% of shares in 5 years |
| Penske | Not stated | Low float + insider concentration; control value ultimately signaled via a tender offer | Market cap $11B; insiders + Mitsui hold ~70%, actual float ~$3B; subsequent tender offer with Mitsui for remaining 17% |
| Miami International Holdings (MIAX) | Not stated | Client raised the Kalshi threat; the author responds that its actual performance is better than other exchanges | Client described the share decline as "stunning"; the author notes the current price is no lower than it was a few months ago |
| WaterBridge | Not stated | Client requested an update; like LandBridge, it is a land/water-rights asset platform | No specific data provided |
| Inflation Beneficiaries ETF | Not stated | Positioned as a "completion allocation" to fill the index's missing inflation exposure to real assets | Weight in S&P 500 constituents only 0.57%; in Russell 1000 0.59% |
| Japan Owner Operator ETF | Not stated | Focuses on family-controlled, non-institutionalized local companies underweighted by indices | Extremely low index weights or not in the index at all |
| Japan Special Opportunity strategy | Not stated | Buys subsidiaries and waits for parent-company spin-offs/going-private "catalyst value" | Not visible in the index framework; relies on fundamentals rather than market-cap screening |
| Spin-Off strategy account new portfolio positions | New position | Initial new positions differ somewhat from the Core Value portfolio | Company not specified |
| Mineral royalty position | Liquidated | Client asked why it was liquidated; the article does not elaborate on specific reasons | No specific data provided |
| Other exchanges | Not stated | Client asked about the negative impact of perpetual-contract regulatory approval | No specific data provided |
| Crypto / Bitcoin | Not stated | Client asked about short/long-term outlook and whether the bitcoin thesis is still the same as 2015 | No specific data provided |
| Gold | Not stated | Client asked about the outlook for gold; this article draws no conclusion | No specific data provided |
| TIPS (Treasury Inflation-Protected Securities) | Not stated | The author believes they hedge the CPI statistic rather than commodity prices themselves | No specific data provided |
| Amazon (example) | Not stated | Used together with bitcoin and Snowball to explain the non-linear compounding of exponential functions | Not a position move this period; for illustration only |
| Unspecified airport subsidiary (example) | Not stated | Used to illustrate an anomalous case where the share price falls below intrinsic value under the ETF gap | No specific data provided |