Rebuilding GE Capital

The headcount is the argument

Asked in a few sentences what his firm actually does, Apollo Global Management's CEO answered by describing how the firm employs people rather than what it manages: "there are 4,000 other people at Apollo who do not carry an Apollo business card but instead originate credit. What does that mean? Well, think about what the old GE Capital was 15 or 20 years ago. GE Capital was thought of as the single best originator of senior secured private credit. They used their expertise in medical devices and aircraft to become a really good lender. That's essentially what we've done. We have a fleet finance business, an aircraft finance business, a securitization finance business."¹

Against roughly 3,000 people in asset management and 1,500 in retirement services who do carry the parent firm's name, the 4,000 people originating credit through owned specialty lenders means more than half the firm's total headcount works at companies not called Apollo at all.

What origination capacity is physically made of

The concept of origination capacity as the binding constraint on a firm like this is often stated abstractly, as capital being abundant while the ability to create assets is scarce. This is the concrete version of that claim: the capacity is not a bigger deal team, better screening technology, or stronger relationships in the ordinary sense, it is owned operating lenders staffed by people with real domain expertise in a specific asset type. Aircraft-lending capacity means owning an aircraft finance business staffed by people who actually understand aircraft, not a generalist credit team that occasionally underwrites planes.

That reframes how a coming shakeout in the wider industry is likely to play out. Origination capacity cannot be hired in a single quarter, but it can be acquired or built over years, since specialty lenders are plainly buyable even if the judgment inside them is not. The sharper version of the anti-consolidation argument, then, is that buying a competing manager's assets under management buys a deployment obligation, while buying an operating lender buys an actual capability, and only the second is worth doing.

A useful, and slightly uncomfortable, model

Using GE Capital as the reference point does double duty. As a model, it is the right analogy: domain expertise in things a parent company actually built and serviced, aircraft engines, medical devices, converted directly into a lending edge that a generalist bank could not replicate, since knowing what a used piece of equipment is actually worth is itself a form of credit analysis. As a warning, it is a strange banner to choose, since GE Capital was ultimately dismantled after the 2008 financial crisis precisely because a large, opaque lending book funded on short-term commercial paper and attached to an industrial parent turned into a systemic liability. The distinction available in defense is that GE Capital funded itself with commercial paper, a mismatch between short-term funding and long-term loans, while an insurance-anchored lender funds itself with long-dated annuity liabilities instead, meaning the lending model is similar while the funding model is closer to the opposite, though that is a distinction worth drawing out rather than one offered unprompted.

Where the credit-expertise claim gets tested

Pressed on whether the firm is genuinely better at picking credit than banks with a century of experience, Rowan's answer relocates the question rather than claiming superiority outright: "I wouldn't think so. Our credit record stands up to, I'd say, any bank out there... my guess is the same people who were there 100 years ago are no longer here, Nikolai. And these are all ultimately about people." The implicit argument is that institutional credit memory accumulates in the people who hold it, not in the institution itself, and people eventually leave, an observation that cuts both ways, since it applies equally to the firm's own operating lenders, staffed by people who could leave too.