Henry Ittleson was 37 years old in February 1908 when he incorporated a modestly named outfit in St. Louis: the Commercial Credit and Investment Company. His first product was receivables finance, advancing cash to small businesses against money they were already owed.

At the time, banks did not like receivables finance. They wanted simple loans against simple collateral, not a business that required constant monitoring and collections. Ittleson saw an opening: he would provide credit where banks would not.

In 1915, Ittleson moved the firm to New York and renamed it Commercial Investment Trust, but it became better known by its initials: CIT.

Not long after moving to New York, CIT struck an early agreement with Studebaker to finance automobiles, then broadened into installment finance for the consumer goods that defined the era—radios, furniture, and the new wave of household appliances like refrigerators and washing machines.

When the underwriting problem changed (from receivables to autos to appliances) CIT changed with it, widening its aperture while banks were slow to build the sales, servicing, and collection infrastructure required to compete.

CIT touted its speed and flexibility relative to banks as key differentiators. The story resonated with borrowers and enabled consistent growth for much of the 20th century.

In the 1980s, the company’s growth was turbocharged by the growth of the securitization market. Invented for the mortgage market in the 1970s, securitization technology broadened to include other types of assets in the 1980s; auto loans were first securitized in 1985 and credit card receivables followed in 1986.

CIT became an early-adopter (and prolific issuer) of asset-backed securities. Securitization changed the tempo of the company’s business model. Once it could sell pools of loans into the bond market, it didn’t need to hold them to maturity to grow. It could originate, distribute, and repeat—freeing up capital and funding capacity each time it securitized a pool.

In addition to being a heavy user of securitization, CIT became increasingly reliant on commercial paper as a funding source. Borrowing via commercial paper was cheap, enabling CIT to raise billions in capital at rock-bottom rates.

Cheap financing improved CIT’s Net Interest Margin (or NIM), the difference between its asset yield and its liability cost. While its funding was not quite as cheap as a bank’s, CIT didn’t have to abide by the regulatory oversight that comes with a bank charter.

It could lend to riskier borrowers at higher advance rates and hold less equity against its loans. As a result, CIT and other non-banks enjoyed similar (or better) returns on equity as banks.

CIT was riding high. It was a good time to be a non-bank lender. An early “golden age” before the term would be used to describe the private credit market in the 2020s. CIT reached peak assets of $90.2B in 2007, establishing itself as a legitimate force in the world of finance.

Then…the financial crisis happened. The liability structure that had powered CIT’s growth since the 1980s became a critical weakness. It’s wholesale funding model had to be rolled over every 30, 60 or 90 days. With credit markets shut, the firm was forced to draw $7.3B from a syndicate of banks to refinance maturing debt in March 2008.

When this wasn’t enough, CIT applied for and was approved as a Bank Holding Company, which enabled it to receive a $2.3B capital infusion from the US Treasury’s Troubled Asset Relief Program (TARP) on December 31, 2008.

The proud non-bank, in a moment of weakness, had to become the thing it had positioned itself against for most of its existence.

Ultimately, even the TARP money couldn’t save CIT. Losses on the assets side combined with a crisis of confidence on the liability side forced the company to file for bankruptcy on November 1, 2009.

CIT emerged from bankruptcy shortly after but shifted its funding model to a bank-style balance sheet where deposits became the dominant funding source. The firm merged with First Citizens BancShares in early 2022 and today operates as a division of First Citizens Bank.

Why tell the story of CIT at such length?

Because CIT illustrates a recurring theme in credit markets. The end state of many (most) non-banks is to become a bank—or to go bankrupt. With private credit being today’s iteration of the non-bank model, it is worth dwelling on whether this time will be different.

A Simple Framework: How Non-Banks End

Non-banks are usually founded by people who understand banks. Often they are ex-bankers. Or they built their careers lending alongside banks and grew frustrated watching good borrowers get turned away for reasons that felt bureaucratic rather than economic.

They start a non-bank for the same reason entrepreneurs start anything: to do a job incumbents won’t do.

Banks have constraints. Some are legal, many are cultural. They prefer clean collateral and clean paperwork. They dislike messy credits that require constant monitoring. They dislike products that require operational muscle: collections, servicing, repossession, workout teams. They especially dislike anything that invites regulatory oversight.

Non-banks pitch their speed and flexibility to borrowers. They claim they can underwrite with more nuance. They claim they can move when banks have to wait for committee meetings. And in good times, they are often right.

But non-banks have a key weakness relative to banks: their liability structure. Banks are mostly funded by deposits, while non-banks are mostly funded by wholesale markets.

Deposits are sticky. They don’t disappear overnight because a bond investor read a scary headline. The FDIC stands behind them. A bank with deposits is not immune to failure, but it has time to deal with credit issues. It can make mistakes and still survive long enough to correct them.

Non-banks build their businesses on liabilities that look stable until the day they aren’t. Commercial paper. Warehouse lines. Repo. Securitization execution. Whole-loan sales. The financing technology changes with the decade, but the core feature is the same: the funding must be rolled, and the market must be willing to roll it.

And markets can be fickle, especially when fear takes over.

That is why non-banks tend to end in one of four places.

First, non-banks become banks. Examples: CIT Group and GMAC (now Ally).

Sometimes this is voluntary, but often it isn’t. During the GFC, CIT and GMAC transitioned from non-banks into bank holding companies to survive. Both lost access to market funding and needed a liquidity backstop from the Federal Reserve, something they could only get by becoming a bank. Both received TARP funds, but only GMAC received permission to issue FDIC-insured debt. GMAC survived, rebranding as Ally. CIT went bankrupt, but restructured and became a bank anyway.

Second, non-banks get bought by banks. Examples: Associates First Capital, Household, Foothill Capital.

Sometimes non-banks are absorbed into banks not because they are struggling, but because they are crushing it (at least on paper). This was the thesis behind Citi buying Associates, HSBC buying Household and Norwest buying Foothill. The banks wanted access to non-prime lending or asset-based lending categories that they lacked. Unfortunately for the banks, the only acquisition of the three that worked was the Norwest - Foothill combination. Once integrated, it became clear that the high yields earned by Household and Associates were generated by aggressive fees and sales incentives that were likely predatory—and untenable within a regulated bank. Citigroup had to write off its entire $31.1bn investment in Associates First during the GFC.

Third, non-banks don’t become banks—but they discover, in a crisis, that they still need the banking system. Example: GE Capital.

This is the “public backstop” outcome, with the example being GE Capital. The non-bank may remain legally outside the perimeter, but during systemic stress it finds itself relying on emergency scaffolding built to keep funding markets from collapsing. Prior to the GFC, GE Capital was effectively a wholesale-funded shadow bank, with roughly $106 billion of commercial paper outstanding, an enormous short-term liability stack funding long-term leases and loans. When the CP market froze after Lehman, investors refused to roll GE’s paper, and the funding model broke. GE’s first lifeline was the Federal Reserve’s Commercial Paper Funding Facility (CPFF), where it placed roughly $16 billion of paper in October 2008 to replace private money-market funding that had evaporated. Then GE obtained an even more consequential backstop through the FDIC’s Temporary Liquidity Guarantee Program (TLGP): GE announced FDIC backing for up to $139 billion of GE Capital debt. It ultimately issued on the order of $50+ billion of FDIC-guaranteed notes in 2008–2009, government-insured wholesale funding that allowed GE to refinance when unsecured markets were effectively shut for non-bank issuers. GE Capital was such an important non-bank that it received a SIFI designation from the Fed (like the big banks).

Fourth, non-banks don’t do the other three options and eventually go bankrupt. Examples: Finova, Fremont, Conseco Finance.

Finova, Fremont, and Conseco Finance were real franchises in real niches, but they were built on wholesale funding—commercial paper, securitization execution, and credit lines that had to be continuously rolled. As we’ve seen, this fails when confidence breaks. The asset book can be “money good” over time but it doesn’t matter, because the company runs out of runway before it can prove it.

The four paths seem to suggest a similar arc. Booms produce non-banks. Busts turn them into banks; or into cautionary tales.

Private Credit: Non-Banks Without These Issues?

Private credit is the modern incarnation of the non-bank. Its leaders make the same pitch that CIT and other non-banks made in prior periods. Banks are constrained by the regulatory scrutiny that comes with taking deposits. Private credit is not. It can move quickly, structure creatively, and lend where banks have retreated.

To its credit, private credit has been built on a more solid foundation than its predecessor non-banks. Its growth from $200B in 2009 to $2T today has been powered, for the most part, by permanent capital (BDCs) or long-duration institutional commitments.

While private credit managers do make use of leverage, it is generally more modest than non-banks of the past. And generally termed out. As a result, private credit hasn’t run with a material asset-liability mismatch for much of its existence.

But fifteen years into private credit’s growth story, this is starting to change. Institutional allocations are maturing and private credit firms are finding growth harder to come by.

They are increasingly looking to the retail market for the next leg of expansion. Executives from Ares to Blackstone to Blue Owl speak openly about the opportunity. Retail capital is large, mostly untapped and eager for yield. It also tends to want liquidity.

You can see where this is going.

Private credit is being repackaged into interval funds with redemption windows, and now even ETFs that offer daily liquidity. These wrappers serve the goal of expanding the investor base. But they also reintroduce a familiar failure mode: pairing long-duration, hard-to-sell assets with liabilities that can walk out the door.

To be sure, private credit firms are not balanced on a knife’s edge to the same extent that groups like CIT or GE Capital were in their final years. Still, the trend towards liquidity is an unwelcome development if the goal is to make private credit something that lasts.

The last fifty years of non-bank history indicate liability design is crucial. Most non-banks eventually become banks, not by choice but by necessity, when a crisis forces them to seek stable capital and only banks can provide it.

If private credit is to chart a different course, it will have to resist the forces that have pulled so many before it back inside the perimeter.

Covenant Lite