Private capital is not one machine. The name tells you the kind of bet, and so the conditions that kind needs to hold; it does not tell you whether they actually hold. That is the rest of the machine, and you have to read it.

Visual companions: the structure diagram → and the worked example → (best on a wide screen)

A private equity return is built from five things: the price paid to buy a company, the price it is later sold for, how much its earnings grew in between, how much of the purchase was funded with debt, and how long the whole thing took. A strategy is a way of working those five. A buyout leans on debt. A growth investment leans on earnings. A lending strategy gives up the upside for a claim that gets paid first. Each is a bet that a particular configuration of surrounding conditions holds.

This is a claim about a particular kind of capital: negotiated and structured for one company, run through a vehicle whose own return is the position’s return, with rights used to steer the outcome. Where that holds, the reading below holds. Where it does not, it does not, and the last section says exactly where the edge runs.

The name tells you which kind of bet it is, and so which conditions it needs. It does not tell you whether those conditions hold. Both are the machine: you get the first for free and confirm the second yourself.

A strategy can change while keeping its name

A controlling buyout is a financing structure as much as an operating one. Cheap, plentiful debt is what turns a doubling of earnings into three times the money rather than two. That extra turn is the leverage, and it depends on a debt market deep enough to supply it.

A fund’s vintage is its configuration fixed at one moment. Two vintages of the same strategy can be different machines, because the debt market, the entry pricing, and the exit window each one deployed into were not the same. The name on the fund tells you the kind, and the conditions that kind needs. The vintage tells you which ones it actually got.

Take that debt away and nothing about the label changes. The deal is still called a buyout, still underwritten as a buyout, still sized against a buyout’s expected return. But the equity now has to fund almost the whole purchase, the extra turn is gone, and the entire return has to come from growing the business. The strategy has quietly become an operating-improvement play wearing a buyout’s name. Anyone still pricing it to the old number is pricing a configuration that no longer exists. The same deal, with each condition moved one at a time, is worked through in numbers in the companion worked example.

The same thing holds across every strategy, at the name. A minority stake that relies on negotiated protections is a real position where those protections are enforceable and an empty one where they are not, with the same documents either way. A loan repaid from a company’s cash flow is a different instrument from a loan repaid only if the company can raise its next round of equity, and both are called lending.

And the configuration the name leaves out is not only this market over time but which market, and which setup, it is in at all. A buyout where acquisition debt cannot be had at any price was never the leveraged machine its name implies.

Two kinds of condition

The conditions that turn a strategy into a different machine come from two different places, and they have to be read differently.

Most are read from the world the position runs in: whether debt is available to carry the deal, whether fresh capital arrives to finish the build, whether real buyers compete for the asset at exit, whether a paper right holds when it is contested. Call these the field conditions. They are not chosen; they are the market the position is dropped into. Two of them work differently from the rest: whether there are enough good targets to be selective (asset density) and whether advisers and operators can be hired deal by deal rather than carried in-house (intermediation depth) do not pick out any one strategy. They set the price and reliability of moving every other term, taxing or lifting all of them at once.

A few are read instead from whoever holds and runs the position (the fund or vehicle, not the investors behind it). Two matter most. There is the clock: whether the funder needs its cash back on a schedule the deal’s own realisation point will meet, or on one that arrives first. And there are the objectives: whether the funder wants only the return, or also something the five terms never counted.

The asymmetry between the two is the part worth holding onto. A missing field condition can foreclose the machine outright. There is no version of a leveraged buyout in a market where acquisition debt cannot be had at any price; the strategy simply cannot be run there, whatever its merits somewhere else. A funder condition never forecloses like that. The position still runs. What it does is change what the position is being run toward. A clock that needs cash at year two of a four-year build sets the return by the calendar rather than the deal. An objective that includes something past the financial return (money that will hold past the point a financial seller would sell, or pay what a financial buyer would not) means the same five terms are no longer the whole scorecard. The machine still turns; it is just no longer being scored only on the number it was built to produce.

So: field conditions decide whether the strategy can run in this market at all. Funder conditions decide whether the funder still runs it to the return its name implies. The name records neither.

The full apparatus — every condition in both tiers, the strategies each one generates, and the work-arounds that try to escape them — is mapped in the companion structure diagram.

Escaping a condition only moves the risk

Some structures exist to get around a condition that a market does not supply, and they are worth reading carefully, because none of them actually removes the condition. Each relocates it to a single narrower point that the structure now depends on entirely.

Securitisation manufactures an enforceable claim where the broader legal environment will not enforce one, by ring-fencing a specific stream of cash. It works, but the whole position now rests on that ring-fence holding when it is tested, which is the same enforcement condition moved down one level rather than escaped. A permanent-hold vehicle removes the need for an exit that does not exist, and pays for it with the permanent illiquidity of its own capital. In every case the condition does not disappear. It concentrates at whatever point the escape now depends on. When a structure looks like it has solved a problem the market could not, that point is exactly where the risk has gone, and exactly where it is least likely to be priced.

Whether a condition can be escaped at all does not sort by tier. Some field conditions have a work-around and some have none; a funder condition like the clock has one too, in the continuation vehicle that lets a fund become its own buyer when its finite life runs out before a real one appears. The lesson is the same in every case: the escape buys relief at one point by loading the risk onto another.

Where the apparatus stops

All of this applies to one kind of capital, and not past its edge. Three things have to be true together. The position is negotiated and structured for the specific company, not taken off a market. Its own return flows to the people who funded it, rather than being kept as a spread. And rights are used to steer the outcome, not merely to monitor it. Where all three hold, the reading above holds.

Where they do not, it does not. Public shares, index exposure, and standardised balance-sheet lending sit outside the class, not because the institutions behind them are different, but because the capital is deployed differently. The line runs by deployment mode, not by the kind of firm. A bank running a negotiated, governed, single-name workout is inside it; a private-capital firm buying an index is outside. Ask the membership question first, because the rest of the apparatus only applies once a position is inside the class at all.

Three questions to ask

The name tells you what each position needs. These ask whether it holds.

  • The position: do the conditions its name depends on still hold, or has it become something else the name still fits?
  • The frame: is the market you are judging it against the one you assume, or a different configuration you have mistaken for the standard?
  • The escape: where a structure looks like it escaped a condition the market could not supply, what single point now carries it, and does the approach still hold if that point does not?

None of these needs information a practitioner does not already have. Each needs only checking whether the conditions the name implies actually hold.