For months now we’ve been writing about the mysteriously rising subprime delinquencies afflicting auto ABS structures despite repeated confirmations from the Fed and equity markets that ‘everything is awesome’ (see “Auto Bubble Burst Begins As Subprime Delinquencies Soar To 2009 Levels” and “Signs Of An Auto Bubble: Soaring Delinquencies In These 266 Subprime ABS Deals Can’t Be Good” for a couple of recent examples).
Shockingly, as confirmed by the chart below, 2016 vintage subprime auto ABS structures are even underperforming 2007/2008 vintage securitizations.
And while most have attributed the rising delinquencies solely to deteriorating lending standards and an increasing mix of ‘deep subprime’ loans, UBS Global Macro Strategist, Matthew Mish, thinks there is a better answer, namely failed Fed policies.
As we’ve also argued over the years, while the Fed’s misguided QE and interest rate policies have done a masterful job of creating asset bubbles around the world they’ve done precious little to actually stimulate economic/wage growth, in real terms.
In our view, the root causes of the rise in delinquency rates can be traced back to US consumer income inequality and aggressive easing in lending conditions, primarily from non-bank lenders. In short, the mosaic we see is one where central bank reflation efforts, namely QE and low interest rate policies, have been more successful at fuelling higher asset prices and wealth creation for a subset of the consumer and less effective in stimulating real income growth (particularly at the median and below). Wealth creation becomes self-reinforcing in an environment of financial repression, with more cash looking for opportunities for deployment. For the financial sector that means more loan growth, and many less regulated, non-bank financial intermediaries have happily filled the void, incentivized by low interest rates that help sustain a lower cost of capital for themselves and lower funding costs for their borrowers.
However, the overall credit quality of borrowers has not kept pace with improvement in the aggregate economy. Our prior Evidence Lab work posits that about 38% of US consumers do not generate positive cash flow and roughly 25-30% of US consumers have not seen improving consumer finances (i.e. they do not own their own home or have significant wealth tied to stock markets). As of Q4’16, 18% of US consumers indicated they were likely to default on one loan payment over the next 12 months vs. 13% in Q3’16. This cohort of at-risk consumers reported being about 4x as likely to embark on a major durable goods purchase (e.g. house, car) in the next year.
This is not just a theoretical issue, but perhaps a problem already. 37% of those aged 21-34 in Q4’16 stated they were likely to default on one loan over the next 12 months, up from 27% in Q3, and outpacing other age brackets. We have only asked this specific question twice before in our Evidence Lab Survey and will be keen if these trends continue in our Q1 survey
And while the subprime auto market, on a standalone basis, may not represent the ‘systemic risk’ that subprime housing did in 2007, when combined with outstanding subprime balances on student loans and other types of debt it’s a $ 1.3 trillion issue.
Is subprime auto lending too small to matter from a financial stability point of view? In isolation, yes. According to TransUnion, subprime auto lending balances outstanding total $ 179bn, or 16% of all auto loans outstanding. And subprime balances are about 1.2x above balances as of Q3’09. However, our earlier thesis would suggest subprime auto may be too narrow a lens to view the debate. More broadly, the good news is that subprime mortgage debt outstanding totals $ 567bn, or 7% of all mortgage loans. Subprime balances are about 0.4x 2009 levels. The bad news is subprime student loans balances total $ 370bn, or 30% of all loans outstanding. And balances are 2.3x 2009 levels. Subprime credit card debt totals $ 113bn ($ 88bn bankcard, $ 25bn private label) – reflecting 12% and 20% of all loan balances, respectively, and about 0.8x 2009 levels. And subprime personal loan balances total $ 17bn, or 16% of all debt, and 1.1x levels seen in 2009 (Figure 7).
In short, we estimate subprime consumer debt outstanding totals a still significant $ 1.25tn, comprised primarily of mortgage, student and auto loans.
But, as UBS concludes, the next massive subprime debt unwind won’t be that big of a deal because this time around all of the risk has been laid off on taxpayers…
Comparatively, however, debt levels outstanding are down from 2009 peak levels near $ 1.9tn. In addition, loan loss risk is increasingly borne by the government (e.g., student, FHAbacked mortgage loans), not the banks.