New Investment Models Needed, Revisited: Structure Across the Whole Capital Stack
In March 2017 I wrote New Investment Models Needed to Solve Important Problems. The argument was simple: individually rational fund decisions were adding up to concentrated sector bets, overvaluation and hubris, and the strategy that looked conservative might actually be increasing risk. Two months later, in How to Stay Focused and Disciplined With Many Exciting Distractions, I looked at the same problem from inside the company: easy money and rising valuations make rapid expansion tempting, and it tends to end badly when conditions change.
Nine years later the pattern has filled in. The ten largest companies in the S&P 500 ended 2025 at more than 40% of the index, against about 27% at the 2000 dot-com peak (Tema ETFs, “The S&P 500 Has a Concentration Problem”). Private-market distributions have slowed to a trickle, and the question of an AI bubble has gone mainstream. This essay picks up the 2017 thread and goes further. Most critiques of finance stop at venture capital. But a company is shaped by every layer of capital it touches, and by every investor sitting beside it.
The principle underneath is the one Living Systems Investing starts from: structure dictates value. The form of the money decides what gets funded, how long it is allowed to live, whom it serves, and what it eventually becomes. Good people inside an incoherent structure still produce incoherent outcomes. Change the structure and the behavior follows, without anyone having to become a better person. That is the Nash thread I have been pulling since 2017. Is Chaos Really Bad? Are Win-Win Scenarios Really Possible? asked whether non-zero-sum outcomes were realistic. Best Outcome) argued that aligned, long-term investors could move small companies to a better equilibrium. And Nash Equilibrium, Complex Systems Investing, Phoenixes Rising (2025) carried that argument into today’s small-cap market.
Why the whole stack
Each layer of capital has its own clock, its own payoff and its own exit. A seed investor needs the next round’s markup. A venture fund needs a handful of outsized exits inside about ten years. A buyout fund needs a sale or a refinancing inside its hold period. A public shareholder may need nothing beyond next quarter. A family trust may need nothing for fifty years.
When those clocks and payoffs fit the living thing being funded, capital is a nutrient. When they don’t, capital is a forcing function, and the company is bent toward whatever the money needs. The founder rarely chooses this, and the investor rarely intends it. It is simply what the structure is built to do.
So the useful question is not “is this a good company?” It is “is this company held by a structure that lets it stay good?”
Walking the stack
For each layer, here is what it is built to do (stated fairly), where the structure pulls against coherence, and the tell to look for.
1. First money: friends, angels and “seed”
What it’s for: belief before evidence. Most real enterprises start here, and patient early money is one of the most generative things capital can do.
Where it pulls: the word seed has been inverted. In a living system a seed is small, alive and patiently tended. Its yield comes from growth that is actually earned. In much of today’s early-stage market, “seed” and “pre-seed” mean a large, hype-priced bet on scale, whose return is the next round’s markup. Once a company takes money priced for a unicorn outcome, every later decision is measured against that price.
The tell: a valuation that only makes sense if the company becomes something other than what it is.
2. Venture capital
What it’s for: funding extreme outcomes. Venture’s power-law math, with many bets and a few huge winners, financed the personal computer, the internet and the smartphone economy. That record is real.
Where it pulls: the same math that finds outliers filters out everything else. A fund that needs a few companies to return the whole fund cannot use an essential business that will be worth “only” $200 million. A ten-year fund life sets the company’s clock, whatever the company’s nature. And preference stacks (liquidation preferences, ratchets, participation rights) mean the headline valuation is often not the economic one. In 2017 I pointed to Bloomberg’s reporting on how these terms were inflating “billion-dollar” valuations (Frier and Newcomer, “The Fuzzy, Insane Math That’s Creating So Many Billion-Dollar Tech Companies,” Bloomberg, 2015). The terms protect the latest investor and quietly raise the risk for everyone else, the founder included.
Venture also changes slowly, for structural reasons rather than stubbornness. The venture investor Yohei Nakajima of Untapped Capital lays out the mechanics: ten-year funds, managers who run several cycles, and institutional limited partners who commit to four or five funds in a row, so the top turns over only every couple of decades (interview on the Odin podcast). That is why the LP, not the GP, is the leverage point.
The tell: growth targets set by the fund’s return math rather than by the market the company serves.
3. Growth, late-stage and layered vehicles
What it’s for: scaling proven companies and giving early holders liquidity while companies stay private longer.
Where it pulls: each round sets a new markup, and the markup becomes a debt the company owes to its own valuation. Layered vehicles, meaning SPVs inside SPVs, can stack entry fees and carry on top of each other until the end investor pays several times for the same exposure. The European investor Enrico Mellis (Animal), who runs a deal-by-deal model with no fees, describes late-stage layered SPVs charging entry fees of around 10% (Odin, “The Anti-Fund: One Investor’s Bet on SPVs”). The company feels this as pressure to justify a price it never needed.
The tell: a cap table that has grown more complicated than the business.
4. Private equity
What it’s for: buying established businesses, fixing operations and handing them on stronger. At its best it professionalizes neglected companies.
Where it pulls: in the leveraged version, the return engine is debt and the clock is the hold period. Fees are earned whether or not the company thrives. Marks are often model-based rather than market-tested. In recent years, as exits slowed, continuation vehicles and NAV-based lending let funds extend and borrow against what they hold instead of returning cash. Bain & Company’s 2026 Global Private Equity Report finds distributions to fund investors, as a share of net asset value, below 15% for four years running, an industry record. The rate was 14% in 2025, among the lowest levels since the 2008–09 financial crisis, with about $3.8 trillion in unsold companies still waiting for an exit (Bain press release, Feb 2026). Marty S., who sits on the family-office investment committee at High Plains Capital, adds his own analysis in Family Office Genesis, a LinkedIn series published under IVYFON and Resilient Alpha Media: top-quartile buyout funds near a 24% IRR over the past decade, the median fund only at parity with the S&P 500, and those on mark-to-model figures, not cash returned (Marty S., Family Office Genesis Part 5). Leverage turns a healthy operating business into a fragile one: it can survive anything except the debt.
The tell: cash flows that serve the capital structure first and the enterprise second.
5. Private credit
What it’s for: filling the gap left as banks retreated from middle-market lending. It brings real flexibility and often better-matched terms.
Where it pulls: opacity. Risk travels through contract terms (sale-leasebacks, termination clauses, residual value guarantees) that don’t show up on any screen until stress arrives. In fast-growing sectors such as AI compute, the financing can become circular: suppliers, customers and lenders all lean on the same expected growth. Ernest Chiang has traced how special-purpose vehicles keep data-center leverage off balance sheets (Off-Balance-Sheet AI), and Moody’s has questioned whether some of those deals belong off balance sheet at all (IFR). Three older ideas explain why this recurs. Stability breeds leverage, and leverage breeds instability (Hyman Minsky). Rising prices change the very fundamentals they claim to reflect (George Soros’s reflexivity). And every technological revolution passes through a speculative installation phase before the real gains spread (Carlota Perez).
The tell: you cannot say who holds the risk, or under what conditions it moves.
6. Public markets
What it’s for: price discovery, liquidity and broad ownership. Public markets are still the most transparent layer of the stack.
Where it pulls: passive flows reward size, not coherence. Money goes to whatever is already largest, which concentrates indexes in a few names and makes “the market” itself a concentrated bet. The quarterly clock punishes long reinvestment. And as research coverage thins, small and mid-cap companies, where many of the most coherent businesses actually live, go under-read and mispriced. In Phoenixes Rising I traced three causes: the decline of the brokerages that once covered small stocks, the dominance of passive investing, and predatory financing terms from short-horizon investors. For a patient reader, that neglect is the opportunity. For the system, it is a misallocation.
The tell: a company priced by its index weight rather than its structure.
7. The allocators: endowment-style portfolios and the advisory complex
What it’s for: prudent diversification for institutions and families with long horizons and fiduciary duties.
Where it pulls: the portfolio sold as diversified across public equities, venture and private equity is, structurally, one monoculture in three costumes: the same cheap-money, growth and liquidity regime, the same extractive-center logic, and a correlation that only shows up in a crisis. Polyculture is resilient because its parts are genuinely different. Planting the same crop in three fields and calling it variety is not polyculture. Conflicts of interest deepen the problem. In the same Family Office Genesis article, Marty S., who writes from inside the family-office world and is also involved in the IVYFON family-office network, traces one large investment consultant’s ownership back, through a wealth platform, to a private-equity firm: a private-equity-backed platform owning a consultant that recommends private-equity managers (Marty S., Family Office Genesis Part 5). He also reports a common complaint from his investor discussions, from angels to the largest family offices: as portfolios diversify, the paperwork and filings grow with every position while distributions shrink. Each added manager adds fees and governance load, with no added understanding.
The tell: many line items, one underlying bet.
8. Family offices
What it’s for: preserving wealth across generations, and increasingly investing directly.
Where it pulls: most great family fortunes were built through coherent concentration: one operating business the founder understood and ran. Research on founder-led firms supports what common sense suggests: René Fahlenbrach found that founder-CEO firms outperformed benchmarks from 1993 to 2002 (“Founder-CEOs, Investment Decisions, and Stock Market Performance,” Journal of Financial and Quantitative Analysis, 2009). Then the wealth moves into trusts, discretion moves to trustees whose mandate is prudent diversification, and the concentration that built the fortune is sold, with tax paid on the way out, to fund the multi-manager monoculture described above. Families lose the coherent concentration that worked and never gain genuine polyculture. Meanwhile direct investing has grown from about 44% of family-office deal volume in 2015 to about 70% in 2024 (PwC, Global Family Office Deals Study 2025), with each family rebuilding its own diligence, accounting and reporting machinery from scratch. Peter Lynch called the pattern “diworsification” in 1989. Marty S. calls it diversification drag, and his article “Elon Musk – Paul Allen, Convergent Visions – Divergent Outcomes” gives a sharp example (Marty S., Family Office Genesis Part 5). Paul Allen correctly identified broadband, online media, private spaceflight and AI, but spread his capital across more than 125 companies run by hired operators, while Elon Musk pursued many of the same theses concentrated in a few companies he ran himself. Diversifying beyond your capacity to understand and operate what you own becomes its own source of risk. (Concentration is not the same as coherence, though: a large paper fortune borrowed against is fragile in its own way.)
The tell: a book too spread out to hold an edge in any of it, and too expensive to run.
The pattern beneath the layers
Walk the whole stack and the same few structural failures repeat at every scale. That repetition is the point: these are not eight separate problems.
- Clock mismatch. The money’s timeline overrides the enterprise’s natural timeline. Living systems grow in seasons, while most capital runs on fund lives, hold periods and quarters.
- Extraction is a shape, not a choice. Fee-on-assets, carry on markups and leverage-driven returns pull value toward the center whatever the intentions of the people involved.
- Hidden correlation. Diversification on paper, monoculture in fact. The fragility stays invisible until the shock that reveals it.
- The center pulls the edge back. Coherent edge businesses, small, distributed and bespoke, get financed by structures that push them toward scale, a quick exit or a sale to the center. The financing quietly reverses what made them valuable.
- Inversion. The most dangerous form of incoherence is not an absence of coherence. It is a coherent form hollowed out and reversed while keeping its name: “seed” as a scale-bet, “diversification” as a monocrop, “family office” as a deal broker. The word carries the legitimacy of the original while the structure does the opposite.
Who else is at the table: the layer almost nobody reads
The LSI coherence template is a layered read. It looks at strategy, technology, the business model, current trends and forecasts, and the financial structure, and it asks whether they reinforce one another and the system the company lives in.
One layer gets far less attention than it deserves: the other investors.
I first wrote about this in the 2017 Nash essay. In micro- and small-cap markets, short-term, self-interested investor syndicates were imposing predatory financing terms that harmed both the companies and their other shareholders. My suggestion then still holds: a company should seek out long-term investors who share a systems view, because aligned incentives improve the outcome for everyone.
Your co-investors are your company’s future board conversations. Their clocks become your deadlines, their preferences decide what each exit is worth to you, and their needs determine whether the next round gets funded, the company gets forced into a sale, or a coherent business is allowed to stay what it is. Two identical companies with identical technology and strategy can have opposite futures because of who owns them.
The questions are structural, not personal:
- Whose clock is this company on? What does each investor need to happen, and by when?
- What is the preference stack? Who gets paid first, and what does the founder actually own at a realistic exit?
- What does each holder need in a downturn? Patience, a sale, or a recapitalization?
- Do they bring anything besides money? Contracts, customers, operating skill. One good test comes from the investor Ian Hogarth, as Saul Klein of Phoenix Court relates it: if this investor offered no money, would you still want them on your cap table, and how much equity would you give them? (Odin, “Only 21 in 20,000 funds can repeatedly pick winners”)
- Are the co-investors coherent with one another? A patient family trust and a fund in year eight of a ten-year life are not aligned, even when they own the same shares.
Conventional diligence reads the company. A layered read also reads the room around it, because the room decides what the company is allowed to become.
What new models need to do
This is not a product pitch. It is a set of design principles drawn from how resilient living systems actually organize themselves. Several already exist in the world; they simply aren’t the default.
- Match the clock to the living thing. Patient, stage-matched capital that fits what the enterprise is, rather than forcing it into a 12–18-month venture cadence or a five-year hold.
- Corporation, not fund, where it fits. Permanent or evergreen ownership, held to compound, instead of capital cycling from fund to fund on a fixed life. Some of the longest-lived enterprises in the world grew this way. Klein points to Sumitomo, roughly 450 years old, which moved from an apothecary and bookshop to a copper mine, then a port to ship the copper, warehouses at the port and insurance on the cargo: each business following from the needs of the last (Odin, “Only 21 in 20,000 funds can repeatedly pick winners”).
- Ownership that holds its shape. Steward ownership is a long European tradition, with companies held in trust for a mission, as at Bosch, Carlsberg, Zeiss and the Novo Nordisk Foundation. Cell structures give each unit its own legal boundary around a shared core, so a failure in one cannot spread to the rest: blast-radius thinking written into corporate law.
- A shared center with no stacked take. I worked through one version of this in Case Study for the Application of Complex Systems Investing Principles to the Medical Technology Industry, a hub-and-spoke model for medtech. Venture struggles with medical products: development is complex, FDA regulation is demanding, the user, payer and patient are often different parties, and insurance coverage requires costly clinical trials. In the model, a hub brings business development and investing together and holds the shared resources each company would otherwise rebuild from scratch. The spokes are individual products, such as devices and diagnostics, which can stay lean because they draw on the hub. Investors can participate at the hub, at a spoke, or both. And there are many routes to value besides one large public company: selling a spoke, packaging several spokes into a new company, taking a large spoke public, occasional distributions, and an evergreen structure in place of fund after fund. The case study argues this could lift the odds of success from roughly one in ten toward eight in ten. Scale that up and you get a hub of hubs. A fund of funds stacks fees and allocates passively. A hub of hubs would share one working center (services, standards, governance, pattern literacy) without a fee at every layer, and let learning travel between the parts. The center exists to serve the edge.
- Contracts and customers before capital. Many edge businesses need a buyer, a commercial bank or a supplier relationship more than they need equity. Klein tells of a founder with a $60 million proof-of-concept, part of a $300 million contract, being told by venture investors to “come back later,” when the deal belonged at a commercial bank (Odin, “Only 21 in 20,000 funds can repeatedly pick winners”).
- Concentrate coherently, or diversify into genuine polyculture. Both are resilient. What is fragile is monoculture sold as diversification, and paper concentration propped up by leverage.
- Yield that is real. Value that shows up as cash a business actually earns, not as a mark someone else has to believe.
None of this requires replacing the existing system. The barbell holds here too: large, centralized capital and small, distributed capital can coexist, and some problems genuinely need the big-center model. The point is that today’s structures make the coherent edge nearly impossible to finance well, and the edge is where a great deal of what matters next is being built.
Honest limits
Venture, private equity and index investing each solved real problems, and many people inside them are thoughtful and principled. This is not a moral argument. It is a structural one: a funding model shapes what it funds. The figures cited here (index concentration, distribution rates, median fund returns, family-office deal mix) come from the industry, consulting and research sources named alongside them. They are directional, and they will keep moving. The pattern is the durable part.
Mustard seed insights
- Structure dictates value. The form of the money decides what a company is allowed to become.
- Every layer has a clock. When the capital’s clock overrides the enterprise’s, the structure wins.
- Extraction is a shape, not a choice. You fix it by changing the payoff, not by asking people to be better.
- “Diversified” across public, venture and private equity is often one bet in three costumes.
- Read the room, not just the company. Co-investors are the company’s future, and almost nobody screens for their coherence.
- The axis is coherence, not concentration versus diversification.
- New models exist in pieces (steward ownership, cell structures, permanent capital, shared centers). What’s missing is putting them together around the edge.
This essay is one thread in a larger body of work. The Pythia Scrolls is where that work continues — reading coherence across public markets, private companies and crypto, theme by theme. If you’re investing your own capital directly and this way of reading structure resonates, have a look: The Pythia Scrolls. Founders and builders may find the Edge of Chaos Collective the better fit.
Related in the garden: Is Chaos Really Bad? Are Win-Win Scenarios Really Possible? (2017) · New Investment Models Needed to Solve Important Problems (2017) · How to Stay Focused and Disciplined With Many Exciting Distractions (2017) · Best Outcome) (2017) · Case Study: Complex Systems Investing Principles in Medical Technology (2025) · Nash Equilibrium, Complex Systems Investing, Phoenixes Rising (2025)
More on the principles behind this lens: Living Systems Investing principles
Subscribe via The Coherence Lab Letter here to get these garden essays by email.