The Bad Vintage Decade

The forecast

"The decade that has just passed I think is going to turn out to be very unusual, and I don't think will be great for investors. I have forecast and I've said publicly, I think returns in traditional private markets on existing vintages of funds done over the past decade will be low."¹ Marc Rowan, speaking in November 2024, telling a shareholder that his own industry's recent funds will disappoint.

The causal chain

Rowan's account is entirely macro and structural, and he does not blame manager quality. He lists roughly eight trillion dollars of United States monetary expansion following the financial crisis and then the pandemic; interest rates near zero, which made leverage nearly free; a procyclical rush into private markets, with capital arriving fastest exactly when conditions were most flattering; a claim that 60 percent of what was labeled private equity over the decade was actually growth investing, a materially different, higher-multiple, less-leveraged, more duration-sensitive activity wearing the same name; good companies bought at high entry prices, since he is careful to say the businesses themselves were often fine even where the price paid for them was not; capital structures priced at very low rates now refinancing at much higher ones; and holding periods extending, which compresses realized returns even when the eventual exit value is unchanged.¹

The 60 percent figure is the most interesting item and the least examined. If most of a decade's private equity was actually growth investing wearing a buyout label, the asset class's aggregate return will be driven by growth-multiple compression, and investors who believed they had bought cash-flow businesses with room for operational improvement had in fact bought duration instead.

Why the forecast is unusual

It is a concession against interest, made in the friendliest possible venue, to a shareholder, unprompted by any adverse market event, volunteered in answer to a neutral question about trends in the industry. It is also carefully bounded in three ways, and the bounding is where self-interest sits. The same treatment applies to public markets, so the argument is not simply leave private markets for something safer. Rowan claims private equity is not the same for everyone, and that a core group of firms did not succumb to the benefits of low-cost leverage and did not reflect it in their prices, implicitly placing his own firm in that group, a claim that is unverifiable from this account alone.¹ And the forward view stays positive: he is bearish on vintages already in the ground and bullish on demand ahead, from family offices, individuals, and retirement systems. Taken together, the shape is recognizable: the last decade was bad, my firm was the exception, and now is a good time to allocate. Every part of that may be true, and the shape is still the shape of a sales argument.

What it does not say

No number attaches to "low," against any benchmark, vintage year, or net-of-fee target. No date exists on which the forecast could be scored, since vintage returns resolve over ten to fifteen years. No evidence is offered for Apollo's own vintages beyond the assertion of membership in the disciplined group. And fees are explicitly ruled out as a variable in the same passage, described as remarkably stable over a very long period and expected to remain so, which means a decade of low gross returns paired with unchanged fees is a considerably worse story for the end investor than the framing admits.¹

The narrower, later version

The same forecast reappears roughly eighteen months later in a more specific, sector-scoped form, covered on SaaS Apocalypse: a claim that 30 percent of a decade of private equity capital went into enterprise software, and that in-the-ground returns there will be disastrous, an exit-multiple problem rather than a bankruptcy problem, attributed by then to AI reshaping the software business model rather than to interest rates and printing. The two versions are the same forecast at different resolutions roughly eighteen months apart. What is worth noticing is that the conclusion appears fixed before the reason was: in the earlier telling the bad decade is a monetary story, and in the later one the same bad decade is an AI story pointed at the sector that had just repriced. That is not necessarily a contradiction, since both mechanisms can operate on the same vintages, but it is a caution that a durable bearish call recruiting whichever explanation is topical at the time is harder to score than it looks.