The Long Horizon and the Short Cycle

I. There is a particular feeling that comes with running a fund whose timeline doesn't match the work you want to fund. You watch your clock and your portfolio's clock running at different speeds, and the gap accumulates over the course of a fund cycle. The conventional response, which is to adjust the structure, find creative LP arrangements, push outcomes faster, accept smaller positions in slower-moving ventures, produces a quality of decision-making that feels off even when each individual decision looks defensible. You are choosing the wrong ventures, or the right ventures at the wrong stage, or the right ventures at the right stage with terms that will turn out to be wrong by year five. You can see the mismatch from the inside. Your LPs can see it from the outside. The investment committee discusses it occasionally, in oblique terms. Your peers at other funds are running the same calculations. None of you have figured out what to do about it, because the answer requires a different kind of vehicle than the one you have, and the vehicles that might fit are still being designed. That experience is the subject of this essay. Worth being precise about what the experience actually is, because it is easy to mislabel. It feels like a pipeline problem, because the ventures that fit the conventional thesis don't tend to be the ventures doing the most interesting work. It feels like a diligence problem, because the analytical frameworks trained on technology-company trajectories keep producing assessments that don't match the actual risk profile of a regenerative farm or a multi-generational land trust. It feels like a GP/LP alignment problem, because the conversations about return expectations involve too much translation and too little shared vocabulary. But none of those descriptions reach the actual source. The source is structural: the instruments available were built for a different kind of work, and when you put them up against this kind of work, the fit problems are not incidental but architectural. The map was drawn for a different territory. The map is not wrong. That's the thing worth insisting on, for the same reason the previous essay in this series insisted that the original architects of voluntary carbon markets were serious people solving a real problem. Conventional venture finance was built for a genuine purpose, works well for the businesses it was designed for, and has created substantial value over decades of application. The 5 to 10 year fund cycle, the GP/LP structure, the exit-driven return profile, the portfolio-construction logic: these were not design errors. They were solutions to a specific problem. The problem worth examining is what happens when that architecture is carried into territory the map wasn't made for. A family office principal who has allocated to three regenerative funds across ten years, watched all three struggle with exit timelines, and is now looking at a fourth allocation is not looking for a better version of the same fund structure. She is looking for a map that actually shows her where she is. This essay attempts to describe what the territory looks like, what has been built to navigate it, and where the cartography is still genuinely incomplete. II. The venture-fund model that became dominant through the 1980s and 1990s was a solution to a specific problem: how to finance companies that had no assets to collateralize and no revenue to service debt, but that had a credible shot at rapid growth if they could attract patient equity. The 10-year fund with a 2-year extension, the 2-and-20 fee economics, the portfolio construction across a dozen or more positions, the IRR as the primary performance metric: each of these features was calibrated to that specific challenge. The model worked remarkably well for the business category it was designed for. Technology companies with steep scale curves, capital-light operating models, network-effect platforms, and short paths to market validation could mature within a fund cycle. The LP's illiquidity was compensated by the scale of eventual returns when exits materialized through IPO or acquisition. A firm that funded a company in year one and had a public exit by year eight could generate the return multiples that justified the structure. The math worked across a diversified portfolio because the technology company's primary relationship was with capital markets: quantifiable returns, creatable liquidity, alignment between the investor's financial stake and the enterprise's core operating logic. If you've worked inside this model, you've felt how well it functions when the conditions fit. Henry Hansmann's framework for analyzing ownership structures asks a deceptively simple question: what is the lowest-cost ownership form for a given pattern of transactions? Investor ownership wins where the relevant transactions are with capital markets, where returns can be quantified, liquidity can be created, and the primary relationship of the enterprise is with people holding financial claims against its future performance. That's the technology venture. The investor's financial stake aligns closely with the venture's core purpose; the ownership form fits the transaction pattern. The misalignment emerges when the same framework is carried into enterprises whose primary transactions are not with capital markets but with biological systems, land, time, and the multi-generational relationships that regenerative work depends on. Those enterprises have a completely different transaction pattern. Investor ownership is not wrong for them in any moral sense; it is a category mismatch in Hansmann's sense. The architecture optimizes for a feature, liquidity, that is in tension with what makes the work itself durable. III. The specific claim worth making carefully is not that the portfolio's clock runs slower than yours. It is that it runs on a different mechanism entirely: one set by biological and ecological systems, not by technology development cycles or market adoption rates. The numbers in the agronomic literature are worth sitting with. Soil organic carbon under best-practice regenerative management rebuilds at roughly 0.2 to 0.5 percent annually. A depleted agricultural soil starting at 1 percent organic matter and targeting 4 percent , a common threshold for meaningful soil health improvement, needs 20 to 40 years of consistent management to get there. Perennial tree crops have their own curve: hazelnuts reach first commercial harvest at 4 to 7 years and full production at 12 to 20 ; chestnuts are 8 to 12 years to first harvest and 20 or more to full production; a well-managed silvopasture system takes 10 to 20 years to approach the stocking densities and species complexity that produce premium grass-fed products at scale. These are not projections subject to product-market-fit adjustments. They are biological facts. A diligence process designed to evaluate a company's five-year trajectory has no framework for evaluating a soil's twenty-year one, and the absence of that framework is not a gap in the analyst's preparation. It is a gap in the instrument itself. Land-tenure and governance arrangements have their own long curve, and it is institutional rather than biological. A multi-stakeholder land trust designed to survive its founders is, by definition, designed to mature past any individual's working life. A covenant arrangement binding a family, a community, and a steward to each other across generations is not intended to reach exit velocity in eight years; it is intended to still be functioning in eight decades. The conventional due-diligence question, what does the exit look like, does not apply to this category of work, and the fact that it keeps being asked is itself a symptom of the architectural mismatch. The category error runs in both directions. Trying to force regenerative ventures into the 5 to 10 year fund cycle produces the predictable results: pressure to harvest early, pressure to substitute revenue metrics for the soil health or biodiversity outcomes the venture was designed to produce, pressure to restructure covenant arrangements to make them legible to investors whose mandates require it. What gets optimized away in the process of making a regenerative venture fit the vehicle is usually the thing that made it worth funding. These pressures don't just produce bad outcomes for the venture. They often produce bad outcomes for the investor too, because the shortcuts that close the gap between your clock and the portfolio's clock tend to destroy the thing the portfolio's clock was supposed to be measuring. The venture reaches an exit the investor can underwrite, and what made the venture worth underwriting has been engineered away. A reader with experience in long-duration capital will have formed an objection by now. Timberland investment management organizations, agricultural REITs, and infrastructure funds routinely operate on hold periods of 20 to 30 years ; they exist precisely to match patient capital to slow-maturing assets. The question is not whether long-duration instruments exist. The question is what they are designed to hold. A timberland TIMO holds board-feet per acre. An agricultural REIT holds rental income from land. An infrastructure fund holds toll-equivalent cash flows from physical assets. Each of these instruments is designed to hold a fungible, measurable, exit-ready claim. The regenerative venture's primary value is in the things those instruments are explicitly designed not to hold: the specific soil's carbon trajectory, the specific steward's relationship with the land, the specific community's multi-generational tenure arrangement. Long-duration capital solves the timeline problem. It does not solve the fungibility problem, which is the deeper one. A longer fund cycle helps at the margin but doesn't solve the architectural problem. A 20-year fund with otherwise conventional structure is still an exit-oriented vehicle, still carries an IRR-primary measurement framework, still asks the wrong question about when and how the work reaches a liquidity event. Extending the timeline without changing the architecture is a palliative, not a solution. The architecture itself needs redesigning, and what follows is a survey of the attempts. IV. This is the section most readers find themselves returning to. Not because it proposes an answer, but because if you've been inside one of these structures, or circling around one, or watching a peer try to make one work, what's usually missing is a complete picture of what else is running and what it has actually done. The survey is not exhaustive; the territory is still being mapped. What follows is what's visible from a reasonably careful look. Evergreen funds. The simplest structural modification: remove the fixed termination date. Evergreen funds recycle capital over longer time horizons without the forced-exit pressure of conventional fund cycles. Generation Investment Management's long-horizon approach and Bridges Fund Management's permanent capital structures have operated on something close to this model. What evergreen structures solve is real: the clock pressure that produces early harvest and covenant-shortcutting decisions largely disappears. What they don't solve is the capital formation challenge. Raising LP commitments into a vehicle with no defined liquidity event is substantially harder than raising into a conventional fund with a projected return profile. Most evergreen attempts have defaulted back to 15-year quasi-evergreen structures, which preserve some of the pressure they were designed to remove. What GPs inside these structures consistently report is that the harder problem isn't the fund's clock but the LP's. An LP whose own mandate runs on a 10-year horizon cannot commit to an indefinite vehicle regardless of the GP's willingness to hold. True indefinite-horizon vehicles exist but remain rare, and the LP base willing to commit to them is narrower than the regenerative capital thesis requires. Perpetual purpose trusts. The most architecturally ambitious structural form. Ownership is held in perpetuity for a defined purpose, with no individual beneficiary who can liquidate the position. Patagonia's 2022 transition is the most publicly discussed recent example: Yvon Chouinard restructured the company so that the Holdfast Collective holds the voting shares in trust for the company's environmental mission, while the equity shares sit in a purpose trust. Profits go to the mission, not to any individual owner. The structure is designed to survive the founder without the conventional succession pressures that tend to convert family-built conservation enterprises into ordinary commercial ones within two generations. The form is not new. Bosch's foundation structure dates to 1937 and has operated through World War II, post-war division, reunification, and a century of technological change. The Carlsberg Foundation, founded 1876, holds control of the Carlsberg Group with an explicit purpose lock. The Novo Nordisk Foundation controls the parent of the insulin manufacturer under similar terms. Colin Mayer's work on purpose-driven enterprise provides the theoretical grounding for why this structure changes the operative logic throughout the organization: when purpose is the foundation of ownership rather than a constraint on shareholder return, the governance calculus at every level of the enterprise changes. That argument was made before the Patagonia transition made it famous, and has been borne out in the cases that have run the longest. What perpetual purpose trusts don't solve is capital formation for growth. The trust can't raise growth equity in the conventional sense because there is no equity return for an investor to claim. Operating capital comes from the enterprise's own revenues; growth capital is constrained to what the enterprise can generate internally or raise through debt. For ventures still on the upward part of their biological maturation curve, this is a binding structural constraint. The Patagonia case works because the company was already generating substantial revenues before the transition. A regenerative venture in year three cannot take this path. Steward-ownership and cooperative governance. A different approach to the same problem: decoupling decision-making rights from capital appreciation. In steward-ownership models, the people running the enterprise hold the authority to direct it without holding its liquidation value. The Mondragón cooperatives, founded in 1956 in the Basque Country, remain the most studied example at scale: worker-owned, worker-governed, with financial stakes that accumulate over working careers but cannot be extracted on exit in ways that would destabilize the enterprise. The Carl Zeiss Foundation, also a purpose trust in structure but operationally closer to the steward model, has maintained its governance architecture since 1889. What steward-ownership solves is the alignment problem that conventional investor ownership creates: the steward's interests are aligned with the enterprise's long-term health rather than with a liquidity event. What it doesn't solve is the cost of capital. Investors who hold no equity upside have no reason to accept below-market returns; the enterprise pays for its independence through higher cost of debt or through slower growth than equity-funded competitors can achieve. What practitioners inside these structures will tell you, and rarely put in writing, is that the subtler cost is governance overhead: the continuous work of re-explaining the arrangement's logic to new employees, new partners, and new stewards who weren't present when the original terms were set. The model is durable, but it requires deliberate maintenance of a kind that conventional ownership structures don't ask for. Land trusts and conservation easements. For land-based ventures specifically, separating land tenure from operational investment is the structural move that most consistently changes the long-horizon calculus. A conservation easement placed on land before operational investment begins removes the land's appreciated value from the exit calculation and makes the ecological outcomes durable across ownership transitions. Indigenous-led land trusts in North America, operating under tribal governance frameworks that predate and supersede individual ownership, have maintained multi-generational stewardship across multiple waves of external pressure. Mexico's ejido structure, with its communal and individual hybrid tenure, has persisted through multiple administrations and multiple reform waves since the 1917 Constitution. What land trust structures solve is the land question. They don't solve the operating-capital question. A venture on trust-held land still needs equity or debt to build processing infrastructure, hire staff, and carry the biological maturation period before cash flows arrive. Both questions need structural answers; most current solutions address one without the other. Hybrid and emerging structures. The most interesting recent experiments combine elements from multiple categories: investor-owned operating entities sitting on mission-locked land, holding companies with multiple structural layers creating different risk and return profiles for different capital types, B Corp certification combined with purpose trust ownership of a controlling stake. Certain Brazilian family-land arrangements have attempted one variant: multi-generational family ownership of the land underneath a regenerative operating enterprise, with outside investment at the operating layer, structured to prevent the land from being included in the operational exit. Colombian agroforestry trust experiments are attempting similar layering in a different legal environment. These hybrid structures are promising precisely because they try to match different structural forms to different parts of the capital stack: patient, purpose-locked capital at the land layer; more conventional investment at the operating layer. The challenge is legal infrastructure cost and the complexity of maintaining multiple structural forms across generational transitions and economic stress. Most of these experiments haven't yet been tested through a major liquidity event or a significant political disruption. The design looks right; whether the construction holds is an empirical question that will take decades to answer. V. The survey above maps what has been built. What it doesn't map is what remains open, and the gaps are where most practitioners in this space are actually working. Naming them honestly is more useful than pretending they have been closed. Capital formation for trust-locked ventures. The perpetual purpose trust removes the exit incentive that makes conventional equity investment attractive. A trust-locked venture cannot offer equity returns; it can only offer debt service and, in some structures, revenue-share arrangements capped below what equity would have returned. No structural innovation has yet produced a path for trust-locked ventures to raise growth capital at conventional timescales without compromising the trust structure. Some ventures have solved this through slow, revenue-funded growth; others through philanthropic capital accepting below-market returns; others through debt with patient covenants. None of these paths scales easily or replicably. The most active current work on this gap is happening in revenue-based financing structures and in CDFI-backed debt instruments with patient covenant packages, not in equity innovation, which suggests the field has implicitly accepted that trust-locked ventures and equity capital may be structurally incompatible and is working around that incompatibility rather than through it. Generational transitions in LP relationships. The first generation of regenerative-fund LPs, the family offices and foundation endowments that committed to long-horizon theses in the 2010s, is now approaching principal transition. Next-generation decision-makers at those institutions are reviewing commitments made under a prior generation's thesis, and bringing different time horizons, different risk tolerances, and different definitions of what counts as an outcome. Whether multi-generational LP commitments survive multi-generational LP transitions has not yet been tested at scale. The structures haven't been built for it, and the early evidence from family-office generational transitions in conventional asset management is not especially encouraging. The most promising partial response is happening in family foundations that have embedded the regenerative thesis into their investment policy statement at the governance level rather than in a side pocket, making the commitment harder to revisit without a formal governance process. Cross-jurisdictional structures. Most working long-horizon structures are jurisdiction-specific. A US perpetual purpose trust, a German Stiftung, a Mexican ejido: each is well-adapted to its own legal context and poorly portable across borders. A regenerative venture operating across multiple Latin American countries, which describes a meaningful fraction of the work in this space, faces structural friction that no current architecture handles cleanly. The legal costs and complexity of maintaining different structural forms across jurisdictions are real constraints, and they are constraints that fall most heavily on the ventures and regions where the regenerative work is most needed. The most tractable current approach is modular: a holding entity in a stable jurisdiction sitting above jurisdiction-specific operating entities, with the structural purpose-lock at the capital layer. What the modular experiments have clarified is that cross-jurisdictional complexity isn't primarily a legal cost problem. It is a legitimacy problem. A structure imposed from a foreign jurisdiction on a local operating reality creates friction in the relationships that make the stewardship work, not just in the paperwork. The modular architecture tries to solve this by preserving local legitimacy at the operating layer while holding the purpose-lock at the capital layer. That separation is the key design insight the experiments are testing; whether it holds across generational transitions is the open question. Liquidity for steward-owners. In steward-ownership models, the steward's claim is operational rather than financial. They cannot sell the venture; their financial stake runs through salary and the ongoing health of the enterprise. This is structurally sound as long as the steward's life circumstances are stable. It produces genuine strain when health, family, or geographic transitions arise that conventional ownership would have addressed through a partial sale or recapitalization. The steward-ownership field has not yet developed liquid secondary markets for steward positions, and the absence of such markets is a real friction on adoption among the founders who might otherwise choose the model. The most active current experiments involve structured secondary transactions in which a patient buyer, typically a foundation or a family office with a long-horizon mandate, acquires the steward's position at a negotiated value, providing liquidity without disrupting the enterprise's governance architecture. These transactions are negotiated individually, slow, and expensive, but they are happening, and the accumulation of case precedent is beginning to make the next one slightly less difficult than the last. VI. The work of designing the next generation of structures is collectively underway. It is being done by lawyers writing novel trust documents in jurisdictions without settled case law. By family-office principals making allocations into vehicles whose structural soundness will only be tested in two decades. By founders betting quietly that what they are building will hold long enough for the work itself to mature. By LPs accepting return profiles that don't fit any standard mandate. None of them are working from a finished playbook. The reader who opened this essay with the feeling of two clocks running at different speeds is not left with a way to synchronize them. That turns out to be the wrong goal. The clocks measure different things: one measures capital deployment and return, the other measures biological maturation and institutional durability. They are not designed to run at the same speed, and the attempts to force synchronization are where most of the damage described in Section III comes from. What the structural experiments in Section IV are actually trying to build is more modest and more durable than synchronization: a vehicle designed from the beginning to hold both rates of change at once, that doesn't require the portfolio's clock to run faster or the fund's clock to run slower, but that treats the gap between them as a design parameter rather than a problem to be engineered away. Nobody has fully built that vehicle yet. But the people closest to building it are the ones who stopped trying to close the gap and started designing for it. If you've been feeling the gap from inside a fund structure, and you've made it to the end of this essay, you are already doing that work.