Investing

Blog July 14, 2017

Can Subprime Autos Be Compared to the Housing Bubble During the Financial Crisis?

Some recent auto trends, including excess supply, declining prices and rising defaults, have led to comparisons between today’s subprime auto market and the housing market during the financial crisis. In this blog post, we examine this comparison and offer our take on risks in the subprime auto space.

The $180 billion auto asset-backed securities (ABS)1 market can be segmented by its various types of underlying collateral, including prime, subprime, lease, rental car and floor plan loans (Figure 1). Prime retail is the largest subsector of the market, and it contains the highest quality collateral. At the opposite end of the credit spectrum, subprime auto loans are those made to consumers with risk scores of 660 or less.

Subprime losses have spiked this year and are currently running above pre-crisis levels. Despite these trends, we do not view the housing market during the financial crisis as an accurate parallel to today’s subprime auto market.

Excess Supply

The auto industry faces an excess supply of vehicles coming off lease. While this trend is putting pricing pressure on the entire auto market, a surge in off-lease late-model used vehicles is particularly impactful to the subprime, floor plan and lease sectors.

However, structural differences prevent the excess supply in the auto market from being comparable to the boom in home construction during the financial crisis. While excess home supply puts pressure on mortgage-backed securities (MBS) pricing, the MBS market is much larger than the auto ABS market and has a much higher risk of contagion with other markets. Currently, the auto ABS market outstanding is over $180 billion (including over $50 billion subprime loans), while the MBS market totals over $5.8 trillion.2

Also, while off-lease supply is elevated, we think its effect on the auto market may be slightly overstated. Off-lease vehicles are expected to grow by 300,000 units in 2017, followed by an additional 400,000 units in 2018. By contrast, 17 million new cars are sold annually – not including an additional 5 million used cars sold at auction. An additional 300,000 off-lease vehicles should be absorbed easily and is a minor amount relative to total autos sold.

Declining Prices

Auto prices have declined to the lowest levels since the financial crisis. The National Automobile Dealers Association (NADA) Index is about 13 percent lower than its local high in 2014. However, the following mitigating factors should be considered:

  • The NADA covers vehicles up to eight years old, and as a result, the Index is more volatile than other indices. For example, the Manheim Used Car Index is only down 0.2 percent over the same period.3
  • Pricing declines are, of course, worrisome but the auto market does not have speculators further driving up prices as the housing market did during the crisis. These speculators, who packaged subprime mortgages into collateralized debt obligations and also sold credit protection via the derivatives market on these bonds, exacerbated the negative market impact when the prices of these securities declined.

Rising Losses

Delinquencies of 60-plus and 90-plus days have shown some volatility, and the overall trend is rising losses. Default rates are on a similar trajectory, climbing from around a 5-percent lifetime cumulative default rate in early 2013 to around 8 percent currently (Figure 2).

However, we note that the market could be starting to “self-correct” rising losses, as the origination to borrowers with sub-620 risk scores has been trending down. After reaching a local peak of 25 percent in 2Q15, origination to sub-620 borrowers is now at 19 percent – the lowest level since the first quarter of 2011.

Additionally, auto loans do not offer “teaser rates” to artificially qualify borrowers who actually cannot afford the loans they are taking out. The reset of these housing loans to market rates after the teaser period was an early catalyst to widespread defaults.

Spreads and Structure

Even though we do not liken today’s subprime auto market to the housing market during the financial crisis, we recognize the risks of going down in quality in auto ABS. We are monitoring the growing concerns about fundamental credit deterioration in the subprime and lease sectors. At this point we are cautiously optimistic that the current cycle is a correction, and will be contained to the high-risk sectors of the auto ABS market.

In terms of relative value, spreads on the lowest quality/most junior tranches have only recently started to widen out despite increased risks. Therefore, subprime auto ABS may not adequately compensate investors for the risks detailed above.

Given that we’re in the latter stages of the credit cycle, we remain cautious and focus on the prime auto sector. Currently, investors in prime auto ABS could benefit from strong credit enhancement and senior-subordinate structures that help senior tranches withstand defaults.4 Credit enhancement is the amount of over- collateralization in a deal, and can be loosely thought of as the expected losses for a given transaction. In addition, ABS transactions de-lever over time. This is a function of the excess spread (the interest rate received on the collateral less the interest rate paid on the bonds) that is built into the securities. This excess spread builds over time as sequential cash flows are allocated down the structure and debt is retired. As a result, we’re confident that our exposure in prime auto ABS is entirely consistent with Breckinridge’s overriding philosophy of preserving capital while maximizing risk-adjusted returns.

 

[1] SIFMA and Intex, as of1Q17. 

[2] SIFMA and Intex, as of1Q17

[3] www.nada.com and www.manheim.com

[4] In a 2017 report, Citi examined 2006-2008 vintages and found they could withstand default rates of 10 percent without the senior tranches taking any losses (actual default rates for these vintages were around 3 percent). Further, half of the classes had the capacity to withstand 20 percent defaults, and approximately 13 percent would lose no principal even if the entire underlying pool defaulted.  

 

DISCLAIMER: The opinions and views expressed are those of Breckinridge Capital Advisors, Inc. They are current as of the date(s) indicated but are subject to change without notice. Any estimates, targets, and projections are based on Breckinridge research, analysis and assumptions. No assurances can be made that any such estimate, target or projection will be accurate; actual results may differ substantially.

Nothing contained herein should be construed or relied upon as financial, legal or tax advice. All investments involve risks, including the loss of principal. An investor should consult with their financial professional before making any investment decisions.

Some information has been taken directly from unaffiliated third party sources. Breckinridge believes such information is reliable, but does not guarantee its accuracy or completeness.

Any specific securities mentioned are for illustrative and example only. They do not necessarily represent actual investments in any client portfolio.