That's the number that keeps nagging at me. Eighty percent of all venture capital funds come from institutional investors — pensions, endowments, sovereign wealth funds. Not from Silicon Valley visionaries with a gut feeling. From teachers' retirement accounts and university endowments. So Daniel sent us this one, and it's a good one: what's actually the difference between venture capital and private equity, how does the GP/LP structure work, where does the money actually come from, and who are the limited partners — the people everyone in the industry knows about and nobody outside it ever talks about. He also wants to know which came first, and how the two models diverge in strategy, leverage, and returns.
The limited partner angle is genuinely the most underexplored part of this whole world. You can read a hundred articles about Sequoia or Blackstone and come away with almost no understanding of who actually funded them, how those people get paid, or what leverage they hold.
Or don't hold, which is its own interesting story. By the way, today's episode is brought to you by Claude Sonnet four point six — that's the model writing our script today.
Good taste in topics. So let's actually get into this, because I think the place to start is the thing that trips people up immediately, which is the assumption that VC and PE are basically the same thing with different vibes. They're not. The structure looks similar on the surface — both run as limited partnerships, both have GPs managing the fund and LPs providing the capital — but the underlying logic is almost opposite in some important ways.
The conflation is understandable. Both involve large pools of private capital, both operate outside public markets, both charge you roughly the same fees. But the investment thesis, the risk profile, the use of debt, the type of company they're targeting — all of that diverges pretty sharply. So before we get into the mechanics, can you just give listeners the one-sentence version of each?
Venture capital is early-stage, high-risk, minority stakes in startups, betting that a small number of companies will return the whole fund many times over. Private equity is typically mature companies, often majority control, using operational improvements and sometimes significant debt to generate more predictable returns. The timelines overlap, the fee structures rhyme, but the actual work is completely different.
One of those predates the other by quite a bit, which I think surprises people.
Private equity, in its broadest sense, goes back to nineteenth-century merchant banking. The modern leveraged buyout model really crystallized in the nineteen eighties — KKR's buyout of RJR Nabisco in nineteen eighty-eight being the canonical example. Venture capital as a distinct asset class came later, emerging after World War Two specifically to fund technology startups. The investment in Fairchild Semiconductor in nineteen fifty-seven is often cited as one of the earliest recognizable VC deals in the modern sense.
PE is the older sibling, and VC is the younger one who got all the press coverage.
And VC got the press coverage partly because the companies it funded became household names — Intel, Apple, Google, eventually — while PE was doing things like buying grocery chains and industrial manufacturers, which are important but harder to make into a TED Talk.
Right, nobody's making a documentary about the leveraged buyout of a regional hospital network. Okay, so let's get into the actual structure, because this is where it gets interesting and also where most coverage just completely glosses over the mechanics.
The GP/LP structure is the foundation of everything. Both VC and PE funds are organized as limited partnerships — that's the legal entity. The general partners, the GPs, are the fund managers. They make the investment decisions, they sit on boards, they manage the portfolio. The limited partners are the investors who provide the actual capital. And the "limited" part is not incidental — it's a legal designation that limits their liability to the amount they've committed. They cannot lose more than they put in, and critically, they have no say in day-to-day operations.
Which is a feature, not a bug, for most of them.
For most of them, yes. The LP commits a certain amount of capital, let's say a pension fund commits two hundred million dollars to a fund. But they don't wire that money on day one. What happens instead is the GP issues capital calls over time — as they identify deals and need to deploy capital, they call down portions of that commitment. So the LP needs to have liquidity management processes in place, because you might get a capital call notice with thirty days to fund it.
If you miss a capital call, the consequences are not pleasant.
No, they're not. You can face significant penalties — dilution of your interest, sometimes forced sale of your stake at a discount. It's one of the less-discussed operational risks of being an LP, particularly for smaller institutions that might have cash flow constraints.
The LP is not just writing a check and forgetting about it. They have ongoing obligations.
Ongoing obligations, ongoing monitoring responsibilities, and — this is the misconception I want to push back on — they are not passive. The word "limited" makes people think they just sit there. But sophisticated LPs, particularly large pension funds and endowments, have dedicated private markets teams. They conduct due diligence on fund managers before committing. They negotiate terms. They request co-investment rights. They sit on limited partner advisory committees, which are formal governance bodies that can weigh in on certain fund decisions — conflicts of interest, fund extensions, valuation disputes.
CalPERS is the example I always think of here. The California Public Employees' Retirement System — one of the largest pension funds in the world, something like five hundred billion dollars in assets under management. They allocate meaningfully to both VC and PE. And they're not just passively receiving statements. They have entire teams evaluating fund managers, they've publicly pushed back on fee structures, they've been vocal about wanting better terms.
CalPERS is a great example precisely because they're large enough to have real leverage in negotiations. If you're a mid-sized VC fund and CalPERS wants to commit two hundred million, you're going to listen when they ask for a fee discount or co-investment rights on your best deals. Smaller LPs don't have that leverage, which creates a real stratification within the LP universe.
That stratification matters for understanding where the money actually comes from, because it's not uniform across fund types and fund sizes.
So the typical LP landscape — and this cuts across both VC and PE, with some differences we'll get into — includes pension funds, university and foundation endowments, sovereign wealth funds, family offices, and high-net-worth individuals. Pensions and endowments dominate the large, established funds. Sovereign wealth funds — Norway's Government Pension Fund, Singapore's GIC, Abu Dhabi Investment Authority — have become increasingly significant players, particularly in large PE. Family offices tend to be more flexible, more willing to back emerging managers.
That flexibility of family offices is interesting because it maps onto a real gap in the market. Institutional LPs like pensions almost categorically refuse to back first-time fund managers. The due diligence burden, the track record requirement, the committee approval process — it's all calibrated for established managers with five or more funds behind them.
Which creates this chicken-and-egg problem for emerging managers. You need LP capital to build a track record, but you can't get LP capital without a track record. Family offices and high-net-worth individuals are often the ones who bridge that gap. They're willing to back a first-time fund manager with a compelling thesis, partly because the check size is smaller relative to their total portfolio, and partly because the decision-making process is faster — sometimes it's literally one person deciding.
There's something almost romantic about that, in a finance-nerd kind of way. The entire venture ecosystem has, at various points, depended on a relatively small number of wealthy individuals who were willing to take a flier on an unproven manager.
Those bets have occasionally worked out. The early backers of Sequoia, of Kleiner Perkins, of Andreessen Horowitz — the people who committed to those funds before the track records were established — generated extraordinary returns. But for every one of those, there are dozens of first-time funds that returned nothing or close to nothing.
Which brings us to the fee structure, because how GPs get paid is deeply connected to how risk and reward are allocated across the whole system.
The standard model is what the industry calls "two and twenty." Two percent management fee on committed capital per year, and twenty percent carried interest on profits above a hurdle rate. The management fee is meant to cover operating costs — salaries, office, travel, due diligence expenses. On a five-hundred-million-dollar fund, two percent is ten million dollars a year. That's real money, and it's one of the critiques of the model — GPs can generate comfortable lives from management fees alone even if the fund underperforms.
Which does create a tension with the whole "aligned incentives" pitch that GPs give LPs.
The theoretical alignment comes from carried interest — the twenty percent of profits. The idea is that GPs only make serious money if the fund generates serious returns for LPs. But there's an important detail that often gets lost: carried interest is only paid after LPs have received back their invested capital plus the hurdle rate, which is typically eight to ten percent annual returns. So the GP doesn't get to start clipping twenty percent of gains until the LP has been made whole and has earned at least eight percent annually.
That hurdle rate is doing a lot of work in that sentence.
It really is. And in practice, many funds don't clear it. The distribution of returns in both VC and PE is highly skewed — the top quartile of funds generates the vast majority of the industry's aggregate returns. If you're an LP in a median or below-median fund, you may get your capital back, you may even get the hurdle rate, but you're not generating the kind of returns that justify the illiquidity and complexity of being in a private fund versus just buying an index.
The illiquidity piece is underappreciated. When you commit to a ten-year fund, you are locking up capital for a decade. There's no secondary market in the way there's a public stock market — or rather, there is a secondary market, but it's thin, it's discounted, and it exists specifically because some LPs need to exit before the fund terminates.
The fund lifecycle is worth walking through explicitly. Both VC and PE funds typically run ten to twelve years. The first three or so years are the investment period — the GP is actively sourcing and making investments, calling down LP capital as they deploy. The middle years are portfolio management — working with existing investments, helping companies grow or improve, sometimes making follow-on investments. The final years are the harvesting period — the GP is exiting positions, returning capital to LPs.
The GP doesn't just sit on their hands during the harvesting period. They're usually out raising their next fund, which creates its own set of incentive dynamics.
Right, because the GP's incentive during the harvesting period is to show strong distributions to LPs, which makes fundraising for the next fund easier. There's been academic work suggesting that GPs sometimes time exits strategically around their fundraising cycle, which is not necessarily in the best long-term interest of LPs but is very much in the interest of the GP's next fund close.
This is where the distinction between VC and PE starts to matter quite a bit in terms of how those exits happen and what the return profile looks like.
In venture capital, the model is fundamentally about power law returns. You're making twenty, thirty, forty investments in a fund, knowing that most of them will fail or return only modest multiples. The entire fund economics depend on one or two investments returning ten times, fifty times, a hundred times the invested capital. Sequoia's investment in WhatsApp returned something like fifty times their invested capital when Facebook acquired it for nineteen billion dollars in twenty fourteen. One deal like that can return a whole fund.
Which means VC fund managers are not optimizing for consistency. They're optimizing for the right tail of the distribution.
And that changes everything about how they select investments, how they support portfolio companies, how they think about follow-on capital. You want to double down on the winners and let the losers go. In PE, the model is different. You're typically buying a majority stake in a mature company, often using significant debt — leverage — to amplify your equity returns. The debt is serviced by the company's own cash flows. You then spend three to five years improving operations, cutting costs, expanding margins, maybe doing bolt-on acquisitions, and then you exit — either through a sale to another PE firm, a strategic acquirer, or an initial public offering.
The debt piece is the thing that gets PE the most criticism and also generates the most misunderstanding. Because leverage is not inherently predatory — it's a financial tool. But it does mean that if the company hits a rough patch, the debt burden can become existential in a way it wouldn't for an all-equity investment.
That's a real risk. The leveraged buyout wave of the nineteen eighties produced some spectacular returns and also some spectacular bankruptcies. The Toys R Us bankruptcy in twenty seventeen is a frequently cited modern example — the company was bought by a consortium of PE firms including KKR and Bain Capital in two thousand five, loaded with significant debt, and when the retail environment shifted, the debt service left no room to invest in the business to compete with Amazon.
Blackstone is the name that comes up most often when people think about large-scale PE, and their model is worth understanding as a contrast to something like Sequoia on the VC side.
Blackstone manages something in the range of one trillion dollars across its various strategies — PE, real estate, credit, hedge funds. Their PE strategy targets large, established companies across sectors like healthcare, manufacturing, financial services. They're looking for companies where they can apply operational expertise, industry knowledge, and their network to drive value creation. It's a very different intellectual exercise than what Sequoia does, which is finding a twenty-three-year-old in a garage who's building something that doesn't exist yet and might be the next trillion-dollar company.
Or might be nothing. That's the other outcome.
Which is the whole bet. And the fact that those two models — one hunting for outliers, one engineering consistent improvement — both use the same GP/LP legal wrapper is something that gets lost in how people talk about this space.
It's the same chassis with completely different engines. Both are limited partnerships. Both have GPs who manage and LPs who provide capital. Both charge management fees and carry. But the underlying logic of what you're doing with the money, and why, is fundamentally different.
For listeners who are coming to this fresh — VC is early-stage, minority stakes, high failure tolerance, chasing the outlier. PE is mature companies, often majority control, operational improvement, leverage as a tool, more predictable return targets.
That's a fair compression. And the GP/LP structure is foundational to both because it solves a specific problem: how do you pool capital from a large number of investors, give one team the authority to deploy it without needing approval on every deal, and still protect those investors from liability beyond what they put in? The limited partnership structure does all three. LPs have capped downside — they lose their committed capital in a worst case, but they're not personally liable for fund debts. GPs have full authority to act. And the whole thing is governed by a partnership agreement that runs to hundreds of pages.
The liability cap is not a small thing. It's what makes institutional participation possible at all. A pension fund cannot expose its beneficiaries to unlimited liability. That's not a theoretical concern, it's a structural requirement.
Which is why the conflation of VC and PE bothers me a little, because it obscures the actual decision an LP is making when they allocate to one versus the other. The risk profile, the return timeline, the liquidity characteristics — they're different enough that treating them as interchangeable is a real analytical error.
And that brings us to the LP side of the table. Who are they, and what are they actually trying to achieve with these allocations?
The canonical LP is a large pension fund. CalPERS — the California Public Employees' Retirement System — is one of the most studied examples because they're a public fund and their allocations are disclosed. They manage something in the range of five hundred billion dollars in assets, and they allocate a meaningful portion to private markets, both VC and PE. The logic is straightforward: pension funds have long-dated liabilities — they owe money to retirees decades from now — so they can tolerate illiquidity in exchange for higher expected returns. A ten-year fund lockup is manageable if your liability horizon is thirty years.
CalPERS has been public about the fact that their PE allocations have historically outperformed their public equity allocations, which is part of why the asset class attracts institutional capital in the first place.
Though the measurement question is complicated. PE funds report returns using a metric called the internal rate of return, which is a time-weighted measure of cash flows. The problem is that it's sensitive to the timing of those cash flows in ways that make direct comparison to public market indices tricky. There's a whole literature on whether PE actually outperforms public markets on a risk-adjusted basis once you account for leverage and illiquidity, and the honest answer is: it depends on the fund, the vintage year, and exactly how you do the comparison.
Vintage year is a term that sounds like wine, and I enjoy that.
It's exactly like wine, actually. A fund raised in two thousand seven, just before the financial crisis, had a very different deployment environment than one raised in two thousand ten when assets were cheap and credit was available. The year you raise and start deploying capital matters enormously for eventual returns, and LPs who have been in the asset class for decades think explicitly in terms of vintage year diversification — spreading commitments across multiple funds raised in different years to smooth out that exposure.
Beyond pensions, who else is in the LP mix?
University endowments are the other major institutional category. Harvard, Yale, Stanford — their endowments run into the tens of billions and have historically had significant private markets allocations. Yale under David Swensen, who ran their endowment for decades until his death in twenty twenty-one, essentially pioneered the "endowment model" — heavy allocation to alternative assets including PE and VC. Yale's private equity allocation at various points exceeded thirty percent of total assets, which is aggressive by any standard.
The endowment model got widely copied, sometimes by institutions that had neither the staff nor the relationships to execute it well.
That's a real critique. Swensen himself said as much — the model works if you have access to top-quartile managers, which requires long relationships, a strong brand as an LP, and the analytical capacity to evaluate funds. A small endowment that tries to replicate Yale's allocation without those advantages ends up in mediocre funds with high fees and poor liquidity. The model isn't portable to everyone.
Then you have sovereign wealth funds.
Sovereign wealth funds are fascinating because they operate at a different scale and with different constraints than pensions or endowments. Norway's Government Pension Fund Global is the largest in the world — over a trillion and a half dollars — though they've historically been more public-markets-focused. The Abu Dhabi Investment Authority, GIC in Singapore, Temasek — these are the ones with deep private markets programs. They can write very large checks, which gives them access to the largest funds and sometimes co-investment rights alongside the fund.
Co-investment rights are worth flagging, because that's one of the ways LPs push back on the two-and-twenty model.
A co-investment is when an LP invests directly alongside the fund in a specific deal, usually with no management fee and no carry on that portion of the capital. So if Blackstone is doing a buyout and CalPERS is a fund LP, CalPERS might negotiate the right to put in an additional hundred million dollars directly into that deal at the same terms as the fund but without paying fees on it. Over time, a sophisticated LP can meaningfully reduce their effective fee burden by doing a lot of co-investments.
Which implies LPs are not passive in the way people assume. They're negotiating terms, demanding co-investment rights, doing their own due diligence on deals.
The myth of the passive LP is one of the things that frustrates people inside the industry. Large institutional LPs have dedicated private markets teams. They're doing independent analysis of sectors, of managers, of individual deals where they have co-investment rights. They're sitting on LP advisory committees where they get visibility into fund operations and can flag conflicts of interest. They're not just writing checks and waiting.
Though the governance rights are still asymmetric. The GP runs the fund.
The partnership agreement is clear on that — LPs cannot direct investment decisions without potentially losing their liability protection. There's a legal concept called "control" — if an LP exercises too much control over fund operations, they can be reclassified as a general partner and lose the liability cap. So the passivity has a legal dimension, not just a structural one. But within those constraints, sophisticated LPs are quite active in shaping the relationship.
Then family offices sit at a different point on that spectrum entirely.
Family offices are the most varied category. You have single-family offices managing the wealth of one ultra-high-net-worth family — these can range from a hundred million to tens of billions — and multi-family offices pooling capital from several families. They tend to be more flexible than institutional LPs. They can move faster, they can back first-time managers that a pension would never touch, they can take positions in smaller funds where the minimum check size is too small for CalPERS to bother with. And because the decision-making is often closer to the principal — the family patriarch or a small investment committee — they can be entrepreneurial in their allocation approach.
Which is why, if you're a first-time fund manager, family offices are often the first calls you make.
They're frequently the seed LPs for emerging managers. And there's a logic to it — family offices are often built on wealth that came from entrepreneurship, so there's a cultural affinity with the risk profile of backing a new manager. Whereas a pension fund has fiduciary obligations to beneficiaries that make backing an unproven GP difficult to justify, regardless of how compelling the pitch is.
The fiduciary constraint is the thing that shapes so much of institutional LP behavior that never gets discussed.
It's the invisible hand in the room. Every allocation decision a pension fund makes has to be defensible to a board, to regulators, potentially to beneficiaries. That creates a strong pull toward established managers with long track records, even if the expected return from a first-time fund with a great team might actually be higher. The career risk for the pension CIO of backing a fund that blows up is much more salient than the upside of backing the next great manager early.
There's a whole behavioral finance literature on that asymmetry and I'm glad someone else is living it and not me.
Which brings us to what the money is actually doing once it lands inside these funds. Because VC and PE are not just different in who they back — the underlying mechanics of how they generate returns are different.
Structurally different, yeah. A PE fund like Blackstone is typically acquiring majority control of a mature business — manufacturing, healthcare services, logistics — and the core thesis is operational. They're going to cut costs, improve margins, maybe bolt on acquisitions, and sell the business in five to seven years at a higher multiple than they paid. The value creation is real, but it's incremental. It's not "this company didn't exist three years ago and now it's worth forty billion dollars.
Whereas Sequoia's model is predicated on exactly that kind of discontinuity.
Sequoia, or any major VC, is writing checks into companies where the base case is failure. The power law is brutal — a small number of investments return the fund, and the rest either die or return modest multiples that don't move the needle. If you look at how Sequoia's returns have been generated historically, a disproportionate share comes from a handful of positions. Early Stripe, early Airbnb, early WhatsApp. The rest of the portfolio is largely noise from a returns perspective.
Which is why the VC model requires you to swing hard and swing often. Caution is actually a losing strategy.
It's counterintuitive if you're coming from a traditional finance background. In PE, underwriting discipline matters enormously — you're using leverage, so a bad deal can blow up the fund. In VC, the bigger risk is being too conservative and missing the companies that actually matter. Fred Wilson at Union Square Ventures has talked about this — the pattern of great VC funds is that they don't avoid losses, they generate outsized wins.
Let's talk about the leverage point, because that's one of the cleaner mechanical differences. PE funds are borrowing to buy.
The leveraged buyout structure is central to PE in a way it simply isn't to VC. When Blackstone acquires a company, they might put in thirty or forty percent equity and finance the rest with debt — secured against the target company's assets and cash flows. That debt amplifies returns when the deal works: if the company's enterprise value goes up twenty percent, the equity return is much higher than twenty percent because you only put in a fraction of the purchase price. But it cuts the other way too. A company that's carrying heavy debt load has less margin for error operationally.
Interest rates matter enormously for whether the math works.
The leveraged buyout boom of the nineteen eighties — which is really where modern PE as an industry took shape — was partly a function of the credit environment and the availability of high-yield debt, what people called junk bonds at the time. KKR's buyout of RJR Nabisco in nineteen eighty-eight was the canonical deal of that era. Twenty-five billion dollars, massive leverage, generated enormous controversy about what PE was doing to American companies. That debate has never fully gone away.
The asset stripping critique.
Which has some validity and some overreach. The honest version is that PE firms have sometimes loaded companies with debt that ultimately destroyed them, and sometimes the operational improvements are real and the companies come out stronger. The evidence is mixed depending on the sector and the manager. Healthcare PE in particular has attracted serious scrutiny in the last several years because the operational playbook that works in manufacturing doesn't necessarily translate cleanly to patient care.
The incentive structures collide in uncomfortable ways.
But the point is that VC is structurally insulated from this critique because it's not using leverage at the fund level. A VC fund invests equity into early-stage companies. Those companies might take on venture debt, but the fund itself isn't borrowing to make investments. The risk is equity risk — total loss on individual positions — not the compounding risk of leverage on top of operational problems.
The return profiles end up looking quite different.
PE targets something like low-to-mid teens net returns annually — consistent, institutional-grade, with a Sharpe ratio that justifies the illiquidity premium over public equity. VC targets are theoretically higher, but the variance is enormous. A top-quartile VC fund might return three or four times committed capital over ten years. A median VC fund might barely return capital after fees. The dispersion between top and bottom quartile is much wider in VC than in PE, which is one reason LP manager selection matters so much more in VC.
Why getting into the top funds is self-reinforcing. The best founders want to work with the best VCs, which generates the best returns, which attracts the most capital.
The access problem in VC is real. Andreessen Horowitz, Benchmark, Sequoia — these funds are not capacity-constrained by capital. They're oversubscribed. The LPs who get allocation are the ones with long relationships or something valuable to offer beyond capital: networks, operating expertise, a brand that founders find attractive. Which circles back to why emerging manager selection is so hard. The established funds have structural advantages that compound over time.
Historically, PE came first. VC is the younger sibling in this story.
By quite a bit. The lineage of PE goes back to nineteenth-century merchant banking — private capital financing industrial companies before public markets were the default exit. Modern leveraged buyouts as a distinct practice took shape in the nineteen seventies and eighties. Venture capital in its recognizable modern form is a post-World War Two phenomenon. The funding of Fairchild Semiconductor in the late nineteen fifties is often cited as an early landmark — the model of pooling institutional capital to back high-risk technology companies was novel at the time.
PE had decades to build infrastructure, relationships, and legal frameworks before VC even had a name.
Which partly explains why the LP base for PE is larger and more institutionalized. Pensions have been allocating to PE strategies since the nineteen seventies and eighties. VC as a serious institutional asset class is more recent — the returns from the dot-com era and then the mobile era convinced a broader set of LPs that it was worth the volatility. Before that it was a fairly niche allocation even among sophisticated endowments.
The Yale endowment model helped legitimize it.
Swensen putting substantial capital into VC at Yale gave other endowments institutional cover to do the same. And when those allocations generated strong returns through the nineties and two thousands, the asset class grew. The eighty percent institutional LP figure we mentioned at the top of the episode — that's the downstream consequence of several decades of that legitimization process.
There's something almost circular about it. Institutional capital flows to VC, which funds the tech companies, whose IPOs generate the returns, which justify more institutional capital flowing to VC.
The flywheel is real. And it has geographic concentration effects too — the capital, the managers, and the deal flow cluster in a small number of markets. Sand Hill Road in the Bay Area, a handful of firms in New York, some presence in Boston. The LP capital that comes from pensions in Ohio or pension funds in Europe ends up concentrated in a very small ecosystem geographically. Whether that's optimal or just path-dependent is a interesting question.
I suspect it's mostly path-dependent and people have rationalized it as optimal after the fact.
That's probably the more honest read.
What does any of this mean for someone who's not running a pension fund or sitting on an endowment committee? Because we've gone pretty deep into the mechanics and I want to make sure we surface something actionable.
The GP/LP structure is actually the most useful lens even if you're a complete outsider to these markets. Because the moment you understand that GPs are compensated primarily through carried interest — that twenty percent of profits — you understand where their attention actually goes. They are not optimizing for your liquidity. They are not optimizing for steady income. They are optimizing for a large exit event that triggers carry. That shapes everything about how a fund behaves, what kinds of companies it pushes toward, how it handles portfolio companies that are performing okay but not great.
Okay but not great is the death zone.
A company returning one point five times capital is almost worse than a write-off from a GP incentive standpoint, because it consumed years of management attention and returned almost nothing above fees. The carry math means GPs want their portfolio companies to either go big or go home relatively quickly.
Which is not always what the company actually needs.
Not always, no. And if you're an LP evaluating a fund, understanding that tension is essential. You want to know how the GP has handled the middle of the portfolio historically — not just the wins and the write-offs, but the companies that took longer than expected, where the GP had to make hard calls about follow-on capital versus cutting losses.
The second thing is less about fund mechanics and more about recognizing that LPs are not passengers in this ecosystem. They have real leverage, and in PE especially, the more sophisticated LPs have started using it.
The co-investment trend is the clearest example. Large LPs negotiating the right to invest directly alongside the fund in specific deals — at reduced or zero fees — is now fairly standard practice for major PE relationships. CalPERS does it. The big sovereign wealth funds do it. It's a meaningful way to improve net returns without paying the two and twenty on every dollar deployed. And it only happens because LPs pushed for it.
The practical takeaway there is: if you're ever in a position to evaluate a fund relationship, co-investment rights are worth asking about explicitly. They're not charity, they're a negotiating point.
The third thing, which I think most people haven't done: look up how your pension is actually invested. This is public information for most defined benefit plans in the United States. CalPERS publishes its full private markets portfolio. Many state pension funds file detailed reports. You can find out what percentage of your retirement assets are allocated to PE, to VC, to real assets. And you can look at the managers they've selected and whether those are top-quartile relationships or not.
Most people have no idea their retirement money is sitting in a Blackstone fund or backing a VC strategy in San Francisco.
And it matters, because those allocations are illiquid. If the pension needs to rebalance, it can't just sell the PE position. The J-curve means returns look negative for the first few years of a fund's life before they turn positive. Understanding that your pension's private markets allocation is going to look bad on paper for years before it looks good is actually relevant to how you should think about pension fund health and solvency reporting.
The J-curve is the thing that makes pension fund annual reports almost deliberately misleading if you don't know to look for it.
It's not malicious, it's structural. But the effect is the same — early vintages look worse than they'll ultimately perform, and if you're reading a snapshot you'll systematically underestimate the eventual return. And that divergence raises a bigger question about where these models are headed.
That structural question — whether these two models keep diverging or start converging as the capital pools get larger and the strategies get more sophisticated — is worth sitting with for a minute.
There are real pressures pushing in both directions. On the convergence side, you have large PE firms that have launched VC arms — Blackstone has done it, KKR has done it — and large VC firms that have started doing growth equity deals that look more like PE than traditional venture. The lines at the edges are blurring.
The core mechanics still pull them apart. Leverage versus equity risk, minority stakes versus majority control, power law versus consistent returns. You can't really hybridize those without ending up with something that's neither.
The more interesting pressure point to me is sovereign wealth funds. We touched on Norway earlier, but the Gulf funds — Abu Dhabi Investment Authority, Saudi Arabia's Public Investment Fund — are now deploying at a scale that dwarfs most institutional LPs. When a single LP can write a check large enough to anchor an entire fund, the GP/LP power dynamic shifts considerably.
They're not passive capital anymore. They're setting terms.
They have strategic interests that go beyond financial returns — economic diversification, technology transfer, building domestic ecosystems. That's a different kind of LP than a pension fund that just wants to beat a public markets benchmark by two hundred basis points. How GPs manage those competing LP interests over the next decade is unresolved.
Good place to leave it. Thanks to Hilbert Flumingtop for producing, and to Modal for keeping the infrastructure running behind the scenes. This has been My Weird Prompts. Leave us a review if this one was useful — it helps. We'll see you next time.
