How Wall Street Banks "Masked" A Record $350 Billion In Subprime Exposure

The latest monthly auto-loan data from the Fitch Auto ABS Index showed something very troubling: subprime delinquencies 60 days or more past due on the secondary market rose to 5.76% in February, the highest they’ve been in 22 years, or since 1996, and blowing past the highs hit during the 2008 financial crisis. At this rate, a record high print is assured in a month or two.

The data seemed confusing, almost a misprint to Hylton Heard, Senior Director at Fitch Ratings who said that "it’s interesting that [smaller deep subprime] issuers continue to drive delinquencies on the index in an unemployment environment of around 4%, low oil prices, low interest rates — even though they are rising — and a positive economic story overall." In other words, there is no logical reason why in a economy as strong as this one, subprime delinquencies should be soaring.

As Brian Ford of Kroll Bond Rating Agency added "the securitized universe is a small subset of the overall universe of auto loans, but it still gives you a pretty good indication of the health of the consumer." If that is the case, the US consumer is doing far worse than the near-record S&P and 4.0% unemployment rate would suggest.

Making matters worse, rising interest rates have made interest payment on subprime loans increasingly unserviceable for those who are currently contractually locked up - hence the surge in delinquency rates - or those US consumers with a FICO score below 620 who are contemplating taking out a new loan to buy a car, something we touched on three weeks ago in "Subprime Auto Bubble Bursts As "Buyers Are Suddenly Missing From Showrooms."

Yet while the subprime bubble is clearly bursting on the demand side, courtesy of rising rates and a 3M USD Libor above 2.3%, Wall Street has remained relatively sanguine on the broader financial implications for one reason: there is, allegedly, no concentration in exposure among bank lenders, and thus no concern of a rerun of the subprime bubble bursting on Wall Street similar to 2007/2008. To be sure, following the financial crisis, big banks appeared to have learned their lesson and turned rather conservative when it comes to potential subprime borrowers, insisting on such things as income verification and other details, which has made room for specialty lenders with fewer such compunctions.

Meanwhile, as we highlighted two weeks ago, Wall Street's exodus from the sector meant a gaping market share opportunity, which led to the emergence of scores of "nonbanks": smaller, specialized subprime auto lenders, some backed by private equity firms looking for leverage upon leverage. These lenders make money by borrowing from big banks to fund various high-interest loans to subprime customers. The interest margin, or difference between the rates banks charge those lenders and the rates lenders charge their subprime customers (often in the double digits) is their profit.

However, even here the breaking point appears to have been reached: as we pointed out in early April, three of them – Summit Financial Corp, Spring Tree Lending, and Pelican Auto Finance – recently collapsed into bankruptcy or were shut down. Allegations of fraud and misrepresentations are swirling through the bankruptcy filings.

Still, the fact that exposures were largely siloed-off from conventional banks meant that no matter how bad things get - or will get - would suggest there is no risk of a systemic crisis. Or is there?

After all, these specialized, direct lenders who hand out auto loans, revolving consumer loans, payday loans, and mortgages to America's "subprime" get their funding from somewhere. That somewhere, it will is come as no surprise, is Wall Street which while not making loans directly to subprime borrowers, is on the hook for hundreds of billions in debt it has issued to fund the specialized subprime borrowers which then turn around and do lend out the money to subprime borrowers, at far higher rates of both interest and delinquency.

Call it one degree of separation between banks and subprime borrowers.

Consider the case of Exeter Finance, which was acquired by Blackstone in 2011, and which is one of America's larger specialty subprime lenders. As the WSJ rhetorically asked recently, "where does Exeter get the money to make subprime auto loans?" The answer: "From Wells Fargo and Citigroup. They have helped lend Exeter $1.4 billion for that very purpose."

And therein lies the rub: while banks have been wise to "mask" their exposure to the subprime sector, their unquenchable desire to lend out money and collect interest means that Wall Street may well be on the hook when the next subprime bursts. In fact, according to FDIC filings and WSJ calculations, bank loans to Exeter and other nonbank financial firms quietly increased sixfold between 2010 and 2017 to a record high of nearly $345 billion. They are now one of the largest categories of bank loans to companies.

Of course, banks will push back and say their new approach of lending to the nonbank lenders is safer than dealing directly with consumers with bad credit and companies with shaky balance sheets. Yet, as the WSJ observes, the funding relationships mean that banks are deeply intertwined with the riskier loans they say they swore off after the financial crisis.

Meanwhile, there are numerous vivid examples from the crisis days demonstrating just how incestuous the subprime funding relationships are: one such case was the Montgomery, Ala.-based Colonial Bank, which became one of the largest bank failures of the era after a nonbank mortgage lender misappropriated more than $1.4 billion from its credit facility with the bank, according to the DOJ.

And yet, flooded with cheap money, banks were eager to delude themselves that just because they were no longer directly facing subprime customers, the risk was gone: “It’s very easy for people to deceive themselves over whether risk has migrated,” said Marcus Stanley, policy director at Americans for Financial Reform, a nonprofit organization that advocates for tougher financial regulation.

Comfortably wrapped in this delusion, banks launched on a historic lending spree, which as noted above, sent bank loans outstanding to nonbank, subprime lenders, from $50 billion at the start of the decade, to $345 billion at the end of 2017.

To be sure, what banks' total indirect exposure to subprime loans – not just auto loans, but also subprime mortgages, and subprime consumer loans – is, as Wolf Richter notes, a bit of a mystery, although one can use regulatory filings to put together the bigger pieces of the puzzle. According to FDIC reports, bank loans to nonbanks lenders have soared, and here are the top contenders:

  • Wells Fargo: $81 billion, up from $13.4 billion in 2010
  • Citigroup: $30 billion, up from $4.1 billion in 2010
  • Bank of America: $30 billion, up from $2.8 billion in 2010
  • JP Morgan: $28 billion, up from $10.4 billion in 2010
  • Goldman Sachs: $22 billion
  • Morgan Stanley: $16 billion

And visually:

One can see why, during the period of record low rates, banks would jump over each other to lend money to the nonbanks whose net interest margin was a number the big banks could only dream of, burdened down by regulation.

In the case of the abovementioned Exeter, the company initially tapped the $1.4 billion line of credit to extend billions of dollars of loans since its 2006 founding. Then Barclays and Deutsche Bank joined Wells Fargo and Citigroup to provide the facility. It eventually bundles its loans into securities and sells them to private investors, using the proceeds to pay back the banks, in addition to paying them fees.

Of course, before the financial crisis, there was no need for nonbanks: the typical subprime-auto customer would have walked into a retail branch of a Wells Fargo or Citigroup and gotten a loan directly, and within minutes. However, after crisis, regulators ended the party and Wells Fargo closed its subprime-lending subsidiary in 2010 and dialed back from auto lending more broadly in 2016. Citigroup sold much of its auto lending unit.

This resulted in a cash bonanza for nonbanks who were generally unregulated, as well as a feeding - or rather lending - frenzy for banks:

By 2016, loans to nonbanks grew to the fourth-largest category of bank lending to companies, up from the 11th in 2012. Around that time, officials from the Office of the Comptroller of the Currency reviewed the exposure at more than a dozen banks, according to a person familiar with the matter.

The regulators looked at the types of nonbanks the banks were lending to, whether those loans were properly secured by collateral and whether there were any concentrations of risk, the person said. At the time, the OCC found the exposure manageable.

it also meant that in addition to a line of willing bank lenders, there were countless PE firms desperate to take these firms private. As Richter notes, among the PE firms that plowed into the auto loan subprime businesses was Blackstone, which acquired a majority stake in Exeter Finance in 2011. The result was a churn in the corner office, and Exeter cycled through three CEOs. As of September 2017, Exeter charged off about 9% of its loans, according to S&P Global, cited by the Wall Street Journal. At the same time, lender Wells Fargo’s own "pristine" auto-loan portfolio experienced charge-offs of only 1%, suggesting that while the profit impact is magnified through leverage, the risk exposure is artificially reduced.

But why are alarm bells not going off yet?

Well, banks traditionally require nonbanks to commit the loans they make as collateral for the bank loan. And they will only lend the nonbanks an amount equivalent to a portion of the collateral—meaning a much higher-than-expected share of the loans would have to go bad for the bank to lose money.

Of course, the lent money still ends up with people with poor credit. The typical Exeter customer, for example, has a FICO score of around 570 on a range of 300 to 850. And, as the chart at the top of this article show, the delinquency rates on subprime loans are finally soaring, and have even surpassed losses encountered during the financial crisis.

Meanwhile, nonbanks like Exeter do everything in their power to take their exposure off the books, and securitize as much as possible: it will create subprime asset-backed securities that it tranches and sells in slices as bonds to investors, while potentially hanging on to the riskiest junk-rated slices that take the first losses. At the same time, there is intense interest in the higher-rated slices. Exeter then uses the proceeds to pay down the credit lines, which creates room to fund new business, grow the company and provide even more loans to subprime consumers.

Banks, in the meantime, remain exposed via the original credit line to Exeter and its loans. This works really well, and the fees and spreads are really sweet... until consumers begin to default more than anticipated, which is the case now, as we showed above:

While banks have provisioned for losses, and typically will lend out at low Loans-To-Value, the truth is that nobody knows just how bad recovery rates will be this cycle: some, such as JPMorgan, have estimated that during the next recession (or even before it), recovery rates may be the lowest ever observed, which would suggest that banks and their $345 billion in loans to nonbanks, is woefully overexposed to what is coming.

The truth, however, is that until we go through the end of the credit cycle (which as both Morgan Stanley and Bank of America now warn, is fast approaching), nobody really knows or has a full grasp of the full magnitude of the potential subprime losses, and certainly not the banks, as Bank of America recently found out in the bankruptcy of tiny nonbank lender, Summit Financial, which BofA accused of misreporting losses from soured loans. The good news: Summit was tiny, and losses will be a rounding error for BofA.

But now that rates are rising, and defaults on auto subprime loans exploding, BofA suddenly "finds" itself with $30 billion, and Wells Fargo is facing a further $81 billion in exposure to subprime loans that the banks are, if only on paper, not exposed to: numbers which are clearly material and would have a dire impact on the banks' capitalization and market cap. Which begs the question: will subprime be the proverbial lightning that strikes twice?