Capital Allocation Process

Executive Summary

The Chief Investment Officer applies a disciplined, discretionary, and fundamental research process generally across oligopolistic industries. We maintain permanent coverage of a watchlist of predictable businesses within the CIO's circle of competence and wait for price-to-value dislocations within these known return streams. We often define the return stream based on two variables: owner's yield and conservative normalized growth. We define the owner's yield as normalized core operating profit (adding back non-cash D&A and GAAP stock compensation) minus interest expense, conservative maintenance capex, cash taxes, and the economic cost of stock-based compensation, divided by market cap, but also considering the enterprise value. We generally define conservative growth as the rough rate at which owner earnings will grow over a decade. By adding the owner's yield and the growth rate assumption we arrive at a conservative return stream we can expect the underlying business to produce over the long-term without multiple expansion and contraction. The watchlist grows steadily over time as the CIO's existing idea channels and Meridian, our proprietary screening system, surface companies that may fall within his circle of competence.

The CIO's time is allocated deliberately. Roughly 75% is spent on the portfolio — reading virtually all relevant consumable content on holdings, monitoring for tails that are widening or deepening, and continuously stress-testing existing theses. Roughly 20% is spent on the watchlist — a collection of hundreds of businesses within his circle of competence. The CIO tracks these surface level return streams and occasionally deep-dives into a business when its surface level return stream appears to be attractive relative to existing positions. When a watchlist business experiences a price dislocation that makes its return stream resoundingly superior to a current portfolio holding, the CIO rotates capital decisively.

New capital allocations typically occur one to three times a year, though deployment varies widely. Some years, like 2024, produce zero new positions; others, particularly during crises when correlations spike and indiscriminate selling creates broad mispricing, produce significantly more. Volatility is our friend. The watchlist contains far more businesses priced at attractive return streams than we have capital to deploy. The CIO spends his remaining time researching new businesses, some of which will fall within his circle of competence and enter the watchlist permanently.

Portfolio composition reflects bottom-up stock selection rather than top-down allocation. We believe the strategy can accommodate substantially larger assets than it manages today.

The Return Stream

Over the long term, a stock's return is generally a function of the owner's yield plus growth — provided that any reinvested capital earns an adequate return. Over shorter periods, multiple expansion and contraction dominate a stock's return. Individual businesses experience cyclical troughs that may last years. We generally underwrite to normalized owner's yield plus growth over a decade or more, requiring this to exceed, by a wide margin, the return we expect from the MSCI World index. The return streams offered to us by the market generally move up and down with global equity multiples and various other conditions. Over any specific 5–7 year period, actual returns will deviate—but the underlying economics, if underwritten correctly, assert themselves over time.

By and large, every opportunity is evaluated on a simple formula: Owner's Yield + Conservative Growth.

Owner's Yield

The framework centers on true distributable cash flow—what the owner can sustainably take out of the business while sufficiently reinvesting in the business to maintain its competitive position. Owner earnings is normalized core operating profit (adding back non-cash D&A and GAAP stock compensation) less interest expense, maintenance capital expenditures, cash taxes, and the economic cost of stock-based compensation, with inputs intentionally conservative to guard against overvaluation. If capital isn't being returned via dividends or buybacks, we discount the unreturned cash when ROIC is low or risky, or when management has a history of letting cash sit idle.

The analysis is intentionally approximate. Whether a company earns $120 million, $130 million, or $160 million matters less than whether the range is obviously attractive relative to price. If an opportunity requires a detailed model to look attractive, it is not attractive enough.

Conservative Growth

Growth assumptions are kept intentionally low to provide a margin of safety. The CIO doesn't build detailed financial models or rely on consensus estimates. Growth is underwritten independently using common-sense assumptions about industry structure and demand patterns. If consensus says 8%, and the overall industry is growing 8%, the CIO might end up assuming 3–4% to ensure the return stream remains attractive even if things go wrong.

Tail Risk

Risk is defined not as stock volatility but rather the long-term impairment of the underlying cash flow stream. The CIO analyzes this through two dimensions:

How bad is the worst case? A shallow tail means the return stream is dented but not destroyed—for example growth stalls, but the yield stays intact. A deep tail means the business faces bankruptcy or permanent capital loss.

How likely is the worst case? The CIO rejects any idea where the tail is both deep and wide, and favors shallow-tail profiles where the downside is bounded and survivable particularly when it comes to concentrated positions.

Illustrative Examples

Allocating capital to a deep and narrow tail and to a deep and wide tail were the two process failures that led to the codified changes discussed in Appendix III.

The distinction between shallow and deep often matters more than the expected return itself. A 20% expected return with a deep tail warrants a smaller position size than a 14% expected return with a shallow tail — because the range of outcomes, not just the midpoint, determines how much capital to commit.

Portfolio-Level Tail Risk

While the analysis is company-specific, the CIO also monitors how individual tails aggregate at the portfolio level. We view true diversification as having a diversified set of failure modes across the portfolio, not a certain number of businesses. At a surface level, a portfolio of fifty businesses across different industries can appear well-diversified — but it often takes second and third-order thinking to recognize that the underlying tail risks are more correlated than they look. The CIO watches for hidden concentrations and adjusts sizing or composition when exposure to a single failure mode gets too high.

Research intensity increases after buying. The CIO reads virtually all available relevant content on portfolio holdings daily, watching for tails that are widening or deepening.

The Research Process

Verification, Not Discovery

Unlike traditional buy-side firms that maintain coverage to predict quarterly earnings and model with exactitude, we maintain coverage to exploit material irrationality or ignorance. When a watchlist name experiences a significant price dislocation, we don't start research, but immediately initiate an efficient return stream and tail risk analysis. If the return stream is resoundingly superior to an existing portfolio position, we rotate capital decisively. When a position is exited but the thesis remains intact, the name returns to the watchlist. The fundamental work doesn't expire — a thesis on a dominant aircraft lessor or a global tobacco conglomerate is valid for years, not quarters, and the next dislocation may create another entry point.

Research Infrastructure

Primary sources include public filings (10-Ks, proxy statements, earnings transcripts), internet search, and AI-assisted research tools for rapid business model comprehension. The CIO's research infrastructure is intentionally simple, reflecting the conviction that informational edge is not the source of alpha. The edge derives from variant perception (seeing value where others see risk, dead money, a weak story, or something else), behavioral discipline, divergent thinking, hyper-logical thinking, second and third-order thinking, and ruthless capital allocation. Every position must continuously earn its place against the next best alternative.

Watchlist Deep Dives

For businesses in which the market seems to be offering a return stream that may earn it portfolio inclusion, the CIO conducts comprehensive tail risk stress testing. This isn't a discrete phase—it's continuous. Stress testing intensifies once a position is established and never concludes while capital remains deployed. The CIO generally rejects businesses where the worst case is unsurvivable.

Uncertainty about outcomes, however, often creates opportunity. When a business faces unclear near-term and medium-term results, institutional managers tend to avoid it; no one wants to tell an LP or investment committee "I don't know how this year will play out." That aversion to ambiguity often pushes prices below worst-case value. If the range of outcomes is unclear but the downside is bounded and survivable, and the stock is priced as if the worst case is certain, the opportunity may be compelling precisely because others cannot tolerate the uncertainty.

Management Quality

The CIO seeks businesses whose economics are strong enough that they don't depend on exceptional leadership to create value. For oligopolistic businesses management quality can marginally amplify or impair results, but it is not the pillar of an investment thesis.

The CIO prioritizes businesses where forward guidance reflects shareholder-friendly capital allocation, and views meaningful insider ownership, a focus on per-share value over empire-building, and a proven record of disciplined capital deployment as favorable but not essential characteristics. If an investment thesis depends on management doing something exceptional, it is not a thesis worth pursuing.

Thesis Simplicity

The CIO's orientation is that the core of the thesis must be simple enough to hold in your head and generally explainable in one page. This is a point of pride, not a limitation: the best ideas are simple, and complexity in thesis construction often signals the investment is not compelling enough. Investments must work with a wide margin of safety and error.

Emotional Detachment

A critical element of the research process is complete emotional detachment from investment candidates and positions. The CIO doesn't "like" or "dislike" investments; they are streams of cash flow with differing upside and downside probabilities with homo sapiens (individual animals) managing them. This orientation eliminates confirmation bias because there's no emotional attachment to businesses or people to validate. The analysis is purely mathematical: probability-weighted returns versus probability-weighted risks. This detachment is natural, not forced. The CIO is autistic—conferring reduced susceptibility to emotional contagion, social pressure, and the herd behavior that destroys returns.

Developed Market Mandate with Global Flexibility

The Partnership invests predominantly in developed markets. We have no particular bias to any one geography — core positions such as British American Tobacco in the United Kingdom, AerCap in Ireland and Sallie Mae in the United States demonstrate the willingness to deploy capital with high conviction both inside and outside the United States.

Our process recognizes that a company's geographic risk and currency exposure are often decoupled from its home jurisdiction or place of listing. We analyze the underlying economic reality of cash flows. A UK-listed entity deriving 90% of its cash flows in USD from North America is underwritten as a USD asset in a GBP wrapper.

This global mandate enables us to capitalize on broad mispricing globally caused by sector (country specific, region specific, or globally), subsector (country specific, region specific or globally), as well as national, regional, and global dislocations without boundaries.

All jurisdictions, including our home jurisdiction in the United States, are underwritten through four lenses:

Four Pillars of Jurisdictional Underwriting

Portfolio Construction

Position Sizing

Initial position sizes are generally determined by two factors: conviction level, reflecting the CIO's confidence in the accuracy of the return stream range, and downside protection, with larger positions generally sized based on the magnitude of potential permanent capital loss rather than upside potential. The CIO is willing to initiate at the maximum allowable size of 10% of NAV at cost when the facts support it. The Partnership doesn't employ a systematic approach of building positions over time. When an opportunity meets the standard, capital is deployed. That said, positions that ultimately become large often start smaller—and grow as the stock moves against us, becoming relatively more attractive while the thesis remains intact. Positions approaching the 10% limit generally reflect ideas where the CIO has high confidence in the return stream and believes the risk of permanent capital impairment is minimal. There are no equal-weighting or mechanical sizing rules; position sizes reflect judgment about relative attractiveness, with emphasis on limiting downside exposure rather than maximizing upside capture. Because the portfolio is concentrated in highly liquid businesses and turns over infrequently outside of crises and dislocations, there's minimal pressure to add or rotate into watchlist businesses as AUM grows. The CIO always has a watchlist consisting of hundreds of businesses within his circle of competence, varying in attractiveness, and ready to be deployed to if necessary. As assets grow, capital flows first to existing positions until liquidity limits are reached. New ideas, however attractive, generally wait until current positions are full. The best ideas earn every incremental dollar before new ones enter.

Sell Discipline

Each position is evaluated against two exit criteria:

• Thesis Invalidation. If the fundamental thesis is disproven—whether through deterioration in competitive position, adverse change in industry structure, or management actions inconsistent with shareholder interests—the position is generally exited because other opportunities become comparatively more attractive.
• Opportunity Cost. If a materially more attractive opportunity is identified and portfolio capacity is constrained, the CIO may exit or reduce a position that is less compelling on a relative basis.

The CIO doesn't employ stop-losses or mechanical selling rules based on price movement alone. Declines in price, absent fundamental deterioration, may represent opportunities to add rather than reasons to sell.

Scalability

We believe the strategy can accommodate substantially larger assets than it manages today. The Partnership invests predominantly in liquid, developed-market businesses. The opportunity set narrows as assets grow, as it does for any strategy.

The watchlist of businesses within the CIO's circle of competence consists of hundreds of businesses and is constantly growing, and our sixth or seventh idea is often comparable to what is in the portfolio.

Concentration limits applicable to the institutional portfolio (maximum 10% position size at cost) differ from the historical track record, where larger positions were permissible. This constraint will likely extend average hold periods. The impact on returns is expected to be modest — and may be positive, because we invest on return streams, not catalysts.

Most managers build theses around an earnings beat, a management change, or some identifiable event; once the catalyst plays out, they must rotate and source new ideas. Higher AUM multiplies that treadmill. Our strategy works differently. If we buy at 5x earnings and it rerates to 10x, the forward return is lower than at purchase — but the business may still offer an attractive return from normalized yield and growth, with optionality on further multiple expansion.

We view the requirement to hold as aligned with how the strategy works. Moreover, the watchlist contains a plethora of additional attractive return streams.

Appendix I: Adding Businesses to Circle of Competence

The CIO has no day-to-day operational, administrative, or capital formation responsibilities, enabling full focus on research, underwriting, and portfolio management. Businesses often enter the CIO's circle of competence through three channels:

Chain of Thought Research

A single catalyst—an article, price movement, or thematic observation—often triggers an iterative research process that surfaces multiple investment candidates. The CIO follows second- and third-order consequences across geographies, industries, and value chains, with the ultimate investment opportunity often distant from the original trigger. A commodity price movement may lead to analysis of mining companies, then regional alternatives, then related infrastructure, then customers of that infrastructure. The process is fluid and nonlinear, one catalyst can generate a collection of five to thirty analyzed companies in a single session. Some of these companies are within the CIO's circle of competence so enter the watchlist.

Experiential Observation

The CIO systematically captures investment insights from everyday consumer and industry interactions. This channel generates differentiated primary intelligence—qualitative insights that institutional investors confined to offices and conferences cannot access.

The edge is not exposure to these signals; everyone encounters them. The edge is recognition and action. How many Wall Street investors bought Lululemon apparel for a spouse or daughter before the stock compounded—and never connected the product experience to an investment thesis? How many student loan borrowers invest in equities but never researched Sallie Mae? How many iPhone users in 2019 wouldn't have traded their device for $5,000 cash—revealing extraordinary consumer surplus and switching costs—yet dismissed Apple as a commoditized hardware business at 10x earnings?

The CIO's structure, with minimal mandatory time in institutional settings, is deliberately designed to preserve access to these signals. But access is table stakes. The differentiation is pattern recognition: seeing a consumer interaction, following it to second and third-order investment consequences, and acting with conviction. There is no formal system for producing these insights, but these insights lead to new areas of competence each year.

Network and Content Monitoring

The CIO maintains ongoing dialogue with fundamental investors, industry executives, and research analysts. For existing holdings, the CIO reads virtually all publicly available relevant content on a near-daily basis. When an author demonstrates particularly insightful analysis, the CIO reviews their other published work, occasionally surfacing new businesses that fall within the CIO's circle of competence.

Appendix II: Meridian – Quantitative Screening

For seven years, the CIO has grown his circle of competence and thus the watchlist through his own idea channels discussed in Appendix I. However, the process was never codified, and the watchlist lived in the CIO's head. Meridian changes that. Every watchlist name now sits in a permanent system with the metrics most relevant to the CIO's process, eliminating the reliance on memory to track a dislocation or catch a name in an article at the right moment.

Beyond organizing the existing watchlist, Meridian screens the full universe of publicly traded equities, to surface return streams that may warrant the CIO's attention. For each company in the monitoring universe, it computes a rough expected return stream from two variables: owner's yield and conservative growth. It also gives a rough estimate for tail risk depth and width. The weights and sub-variables within each component are proprietary, editable, imperfect, and adjustable over time.

The system sorts the surface level return streams so that the CIO can spend spare time more efficiently researching businesses that may be mispriced and may fit his circle of competence.

In a contingency—CIO incapacity or transition—Meridian helps provide continuity. The full monitoring universe, the watchlist, every return stream estimate, and the CIO's annotated reasoning are all accessible within the system. The businesses we own and monitor are boring and predictable by design; their underwriting does not expire quickly. A thesis on a dominant aircraft lessor or a global tobacco conglomerate is generally valid for years, not quarters.

Meridian does not replace fundamental analysis. A name with the highest computed return stream in the universe can be passed on in seconds if the CIO sees risks the model can't capture. Conversely, the CIO may pursue a name that ranks lower because pattern recognition and judgment identify something the numbers miss. All movement from universe to watchlist and watchlist to portfolio is manually performed by the CIO.

Appendix III: Process Evolution

Over seven years, two investments stand out as process failures that generated explicit, codified lessons.

Qurate Retail (QRTEA): Balance Sheet Scrutiny

The investment resulted in a loss due to insufficient scrutiny of the company's debt load relative to its earnings trajectory. The capital structure left insufficient margin for error—and a confluence of adverse events (a warehouse fire, spiking freight rates, inflation, weakened demand) stressed the cash flow. The lesson: elevated leverage magnifies bad luck. The debt oversight was the process failure; the rest was misfortune that a stronger balance sheet would have survived.

Nemaura Medical (NMRD): Introductory Stage Exclusion

The investment in this company, which was in the introductory stage of its business and industry development, violated the core principle of investing in predictable businesses with demonstrable track records in mature industries.