The move higher in rates in 2018 has impacted bond markets directly, as is to be expected, but fixed income investors should also consider the second order effects as higher rates have the potential to impact consumer behavior and the real economy. As higher rates work their way through the economy, there is a further impact for fixed income investors to consider: a significant amount of consumer debt is securitized and becomes a part of the investable fixed income universe.
As such, trends in securitization offer a unique vantage point to the state of American consumers. The shifts that have occurred in this space since 2008 are representative of how the world has changed over that timeframe. Importantly too, as investors begin to worry that the current economic cycle is aging, the changes in securitization since the last cycle ended offer insights into how the consumer is likely to fare this time around.
Prior to the last crisis, it was common to observe the “keeping up with the Joneses” phenomenon, where individuals felt compelled to buy beyond their means using leverage to match their neighbors’ behavior (who themselves were likely borrowing heavily, but never mentioning it). In the current cycle, it appears the Joneses are being much more fiscally responsible.
Run for the hills! Oh wait, never mind
It has been fashionable on the ten-year anniversary of the 2008 financial crisis to observe that overall household debt has now exceeded levels prior to the crisis, suggesting an analog of the current moment in time to that one. Most recently, the new record was set at $13.5 trillion at the end of the third quarter, surpassing the 2008 peak of $12.7 trillion. In magnitude, this figure appears ominous, but in character, we would contend it is quite distinct. Accounting for a variety of measures such as inflation, population growth, GDP growth, household wealth, and household income, the level of debt is not nearly as portentous. The country has approximately 20 million more people than in 2008 and while not a period of significant inflation, cumulative inflation has increased approximately 17%. Therefore, using a variety of metrics to put this debt into context, on an adjusted basis the debt is not setting records the way it is on a nominal basis.
Importantly, while the nominal debt levels have exceeded the levels prior to the crisis, when viewed in the context of debt payments relative to income, the story is wholly different. Not only are current debt service to income ratios meaningfully below the pre-crisis levels, they are in fact the lowest since the Federal Reserve’s (Fed’s) collection of this data began in 1980.
Examining the changing nature of consumer debt
This low level of debt payments alone could be misleading. Indeed, the character of consumer debt today is quite different than a decade ago, suggesting any impacts on consumer decisions will be different in the prior cycle. Student loans and auto loans have driven the increase that has set this new consumer debt record, while mortgage debt and credit card debt have actually declined since 2008. This shift has multiple implications for consumers, both in the source of the borrowing and in the nature of the repayment.
The positive impact for consumers is that while credit card debt tends to be floating rate, which means consumers could face a higher burden on their existing balances as rates move up, this is less often true for student loans and auto loans, which tend to have fixed rates. This difference suggests higher rates may curtail future consumption on the margin, but consumers are less susceptible to surprise increases in their monthly budget, which could have triggered more draconian stress to their consumption patterns.
The biggest debts
The same phenomenon is true within mortgages as a sector. At the outset of 2008, the percent of adjustable rate mortgages (ARMs) was 30%, which has declined to less than 10% as of the end of the third quarter. This change is important as it increases the likelihood that the recent increase in rates will impact new marginal buyers (as has been seen recently) as opposed to triggering forced selling of existing properties as in the last cycle.
While less impactful than seeing their monthly mortgage bill increase, consumers may still lose optionality with rates moving up. If payments on a comparable home become more expensive, moving locations may become prohibitive. Consumers are then effectively “locked in” to their existing home. The result is sub-optimal for the individual consumer and decreases monetary velocity in the overall economy.
Longer-term, the challenge of this shift is that a home also serves as the primary saving vehicle for a significant percentage of the population. After thirty years, a buyer owns their home outright, which at a minimum reduces their future cash flow needs, but also offers an asset that can be monetized if needed. By contrast, while student loans should help facilitate an investment in future earnings via education, they are non-fungible as you cannot sell your college degree. Further, protections for borrowers are at opposite ends of the spectrum. Mortgage debt in the United States is non-recourse, so that if one defaults on a home loan, the bank may take the house pledged as collateral, but have no further claims on the borrower’s assets. By contrast, student loans generally survive even bankruptcy of the borrower. As the cost of higher education has gone up faster than the inflation rate, it has had a bigger drag on the disposable income of many young consumers and that impacts their future consumption, including home buying.
That lower disposable income along with shifts in the cultural value of home ownership, generally stricter borrowing standards and perhaps preference differences of millennials versus prior generations have led to the homeownership rate in the U.S. falling from nearly 70% to below 65%. This trend is observable in the lack of growth in home mortgage securitizations despite the increase in population. The other side of this coin is an increase in renting and a rapid increase in multi-family housing securitizations — another way in which consumer trends are captured by developments in the securitization market.
In addition to stricter standards from the lenders themselves, shifting approaches of regulators and rating agencies have helped enforce more stringent conditions on underlying loans and subsequent securitization. One notable exception in the general trend towards tighter standards has been auto loans, where loan terms have extended and loans to borrowers with lower credit qualities remain readily available.
And now back to investing
The impact that interest rates have on consumer behavior comes full circle as those behaviors are translated into new fixed income securities and an investable universe for investors. During the past decade, when the amount of mortgage debt was effectively unchanged, federal government debt has more than doubled from $5.7 trillion at the end of 2008 to over $15 trillion (as of Q3 2018) and corporate debt has grown by over 60% from $5.5 trillion to over $9 trillion. The result to bond investors has been that securitized products have declined precipitously as a percent of bond indices. A decade ago, securitized products comprised nearly half of the benchmark, at about forty-five percent; today they have declined to less than a third of the benchmark.
While overall mortgage debt has remained essentially flat, the proportion of mortgage debt that is eligible for broad fixed income indexes has increased relative to investable, but non-indexed securities, such as non-agency residential mortgage-backed securities (RMBS). At its peak, overall non-agency securitization was near 40%, today it is approximately 10%.
The decreasing role of mortgages in indexes is further complicated by the Fed’s large role in the mortgage-backed securities (MBS) market, which effectively removed $1.8 trillion of securities from the market at its peak. The Barclays float adjusted benchmark, which excludes these securities held by the Fed, has a less than 25% weighting to securitized products. We discussed the topic of the Fed’s unwind of its balance sheet on fixed income indices more extensively as it pertains to Treasuries in Alexa, Invest in Fixed Income.
The investable opportunity set is not only defined by index eligibility or even underlying issuance. For example, as student loans have grown from $600 billion outstanding in 2008 to approximately $1.4 trillion today, securitized student loans have fallen from $238B to $175B. This highlights the changing nature of both the consumer and the investment universe over the past decade.
Conclusions — Securitization as a window to the consumer: the consumer is doing alright
For investors, trends in securitization are important both for what they tell us about consumers, which drive two-thirds of the U.S. economy, and for the investable universe itself. While articles discussing high level consumer debt figures seem designed to trigger flashbacks to 2008, the comparison is inapt as the consumer today is appreciably healthier than ten years ago. “The Joneses” are less heavily levered relative to their income and their debt is more likely to be fixed rate as opposed to floating rate, though their children may have compromised future spending by borrowing for education today.
Securitizations as a percentage of broad fixed income benchmarks have declined to near their lowest levels since the indexes began. This does not reflect decreased importance of the sectors, but rather the relative modesty of consumer debt increases relative to their public and corporate counterparts. In this way, fixed income as an asset class provides a more holistic view into the economy than almost any other as fixed income investment touches the public, corporate and consumer sectors of the economy in a direct and investable way.
This rapid transition of the benchmark market over the past decade is also a reminder to investors that there is no such thing as a static index or investable universe in fixed income. The difference in risk factors between the 2008 benchmark of 45% securitized versus today’s 30% are stark. The differences including the non-benchmark eligible are even more pronounced as mortgage credit risk, once a significant factor for the overall market, is now confined to a narrow corner of the market, while the larger mortgage risk factors today relate to the timing of cash flows — prepayment and extension risks — and thus convexity.
These very trends make securitized assets a good counterbalance to other investable debt. This is not only because of its unique structures and risk factors (credit enhancement and collateral characteristics rather than business models), but also because the consumer may in fact be sporting the healthiest balance sheet among borrowers today.
While an overleveraged consumer and securitization were at the epicenter of the past crisis, today a more modestly leveraged consumer with a better structure of debt is in a healthier position than ten years ago. Securitizations provide a unique vantage point to examine the evolution of consumers over the past decade and to consider the investment environment moving forward. We expect both the market and the underlying consumer trends to continue to evolve, potentially leaving the investable universe quite different in the future than it is today. These changes will create opportunities for investors, but also share key information about the world’s largest economy and securitized assets will remain a key component of diversified portfolios.