The Rise of Private Credit: Why Banks Lost the Middle Market

Fifteen years ago, if you ran a $200 million EBITDA company and needed a unitranche loan, you called your bank. Today you call Ares, Blackstone, Blue Owl, or Golub. The bank is still on the rolodex, but mostly to provide the revolver and a clearing account. Private credit — direct lending by non-bank institutions to mid-market companies — has gone from a curiosity to one of the largest asset classes in alternatives, and it has happened almost entirely outside of public scrutiny. This is a brief history of how that shift happened, what it actually looks like under the hood, and why the next cycle is the one that will tell us whether the model works at scale.

The Regulatory Push

Basel III and the Dodd-Frank framework did two things to mid-market lending. First, they raised the capital that banks had to hold against any non-investment-grade loan, particularly leveraged loans. Second, they imposed leverage limits on the loans themselves — the famous "6x EBITDA" guidance that the OCC issued in 2013. The combined effect was that banks could either originate the loan and earn a thin fee, or hold it and tie up expensive capital. Most chose to originate, syndicate, and exit — and then eventually stopped originating altogether for the squishier end of the market.

The vacuum was filled by asset managers — first the BDCs (business development companies) that had been around since the 1980s but had never quite scaled, then by private credit funds inside the big alternative asset managers, and eventually by insurance-affiliated platforms that married long-duration insurance liabilities to long-duration credit assets.

What Private Credit Actually Is

The default mental model is "private equity but for debt," which is roughly right but misses important structure. A typical direct loan is a senior secured first-lien unitranche term loan, six to seven years in maturity, floating rate (SOFR + 500-700), held by one or two lenders, and with a covenant package that is closer to old-school bank loans than to broadly syndicated leveraged loans.

The borrower is usually a sponsor-backed mid-market company — call it $30-150 million of EBITDA — being financed for an LBO, an acquisition, a recapitalisation, or growth. The lender is a fund whose LPs are pension plans, sovereign wealth funds, insurance companies, and increasingly retail wealth platforms.

The economics are simple. Borrowers pay 200-300 bps more than they would on a comparable broadly syndicated loan. In exchange they get speed, certainty of close, fewer lenders to negotiate with, and a lender who will be in the loan for the full life rather than trading in and out. Lenders earn that spread plus origination fees and call protection. LPs get 9-12% net IRRs on a strategy that is supposedly less volatile than equity.

The Three Models

Not all private credit is the same. There are three meaningfully different models operating today.

Pure direct lending — what most people mean when they say "private credit." Loans held by funds, marked at fair value (which usually means par unless something is wrong), and distributed to LPs. This is the largest segment and the cleanest analog to bank lending.

Insurance-affiliated lending — driven by Apollo, KKR, Blackstone, and others who acquired or built insurance platforms (Athene, Global Atlantic, Resolution Life, etc.) and use long-duration annuity liabilities to fund long-duration credit. The insurance balance sheet provides incredibly stable funding at a low cost, which lets the affiliated asset manager originate at scale and capture the spread between annuity costs and loan yields.

BDC structure — publicly listed vehicles that fundraise from retail investors and lend to the same mid-market borrowers. BDCs have been around for forty years but the modern era started when the big alts platforms started using them as permanent capital vehicles. They pay a high dividend, trade like a hybrid stock-bond, and are subject to disclosure rules that make them the only meaningfully transparent corner of private credit.

The Yield Story

For most LPs, the appeal of private credit is yield. With base rates at 5%+ and credit spreads of 500-700, an unlevered direct loan portfolio yields north of 10% gross. Apply some modest leverage at the fund level and the LP nets 9-12%. Compare that to a broadly syndicated loan portfolio yielding 8-9% with much higher mark-to-market volatility, and the trade is obvious — at least on paper.

The on-paper part is doing real work in that sentence. Private credit loans are valued by the manager every quarter using a fair-value methodology that has very wide latitude. Public credit gets marked to market every minute. A direct lending fund showing a max drawdown of 3% in 2020 was not "less volatile than the BSL market" — it was less marked. The economic loss took a year to show up in the marks, and even then it was smaller than what an honest mark-to-market would have shown at the trough.

The Coming Stress Test

The first real stress test for the modern private credit market is going to come during the next default cycle. Defaults in mid-market lending have been suspiciously low — under 2% annualised through 2024 — but that is partly because the loans are amend-and-extended at the first sign of trouble, and partly because the underlying borrowers have been able to refinance or be acquired before default became inevitable. In a real recession, neither of those escape hatches works.

When defaults rise meaningfully, three things will be tested simultaneously. One: the recovery rates on first-lien unitranche loans, which the industry assumes are 70-80% but which have never been tested at scale. Two: the LP redemption mechanics on semi-liquid retail-facing vehicles, which assume that gates and queues will hold but which have never seen a real run. Three: the marks themselves, which will face pressure from auditors and from LPs who want to compare quarterly reports to public credit indices that are clearly showing losses.

What Borrowers Get

For borrowers, the value of private credit is real and durable. Speed matters in M&A. Certainty of close matters when you are bidding against other sponsors. Having one or two lenders rather than thirty matters when you need to amend a covenant. None of these benefits goes away if the model gets stressed — they are structural advantages of bilateral lending over syndicated lending.

What might change is the pricing. The 200-300 bps premium that private credit currently charges is partly compensation for the structural advantages and partly an arbitrage that exists because the supply of willing lenders is still smaller than the demand for the product. As more capital flows in, the premium compresses. Already in 2024-2025 we have seen unitranche pricing for high-quality borrowers tighten from SOFR + 600 to SOFR + 475. The arbitrage is closing.

What This Means for the Cycle

Private credit is going to be the most-watched corner of the credit markets in the next downturn for the simple reason that nobody really knows what is in the boxes. The funds are private. The loans are private. The marks are private. The LPs are mostly institutional and have their own reasons to prefer steady marks to honest ones. Everything about the structure is designed to absorb shocks quietly.

That can work for a long time. It can also fail spectacularly when the wrong combination of factors hits at the same time — a default cycle, an LP liquidity squeeze, an auditor pushback on marks, and a regulatory inquiry about retail vehicles. The market is now large enough that any of these on their own would be a meaningful event, and the combination would be a real moment.

I am not predicting that. I am noting that the next two years will tell us whether private credit is a structural improvement on bank lending or a regulatory arbitrage that worked until it did not. The honest answer is that both can be true, and which one matters depends on what the macro environment does next.