According to the latest quarterly report on Household Debt and Credit compiled by the New York Federal Reserve, US Household debt increased at a fairly robust pace in the final quarter of 2016, rising by 1.8% q/q or USD 226bn to USD 12.58trn. Despite the sharp increase in 2016, the total amount of household debt outstanding still remains some 0.8% below the prior cyclical peak of USD 12.68trn registered in Q3 2008.
As such, growth in overall GDP and nominal disposable income has easily outstripped the growth in overall household indebtedness. When coupled with still very low financing rates, this has ensured that the overall household financial obligations ratio of household debt service payments as % of Disposable Personal Income remains at or near a multi-decade low.
Not surprisingly, given this economic backdrop, overall delinquency rates in the US have also declined sharply over the past five years. The latest data from commercial banks shows that the delinquency rate on all loans declined to 2.06% which is close to the cyclical trough for this ratio reported between 2004 and 2007. The New York Federal Reserve reports that this ratio including securitized loans is somewhat higher at 4.8% which is also stable at or near a cyclical low.
As we have highlighted many times, it is not the overall level of debt that is important but the ability to service that debt. Given the apparent commitment by the Federal Reserve to continue raising its key policy rate only gradually, as well as continued employment and wage growth, it would appear that overall delinquencies are likely to remain quite low for some time to come.
Nevertheless, the recent report compiled by the New York Federal Reserve does also present several interesting features of the current credit cycle. The most striking feature dating from late 2009 is the almost stagnant growth in outstanding mortgage debt. Outstanding mortgage balances stood at USD 8.48trn at the end of 2016, which is up USD 130bn or 1.5% q/q, but still considerably below their 2008 peak.
This is in contrast to consumer debt such as auto loans, credit card and student debt.
The chart above shows how mortgage debt now only accounts for some 67% of all outstanding household debt compared to around 72% in 2008. In contrast, the relative proportions for consumer credit as a % of all household debt have increased. The aftermath of the 2008 housing bubble and associated reforms, along with more stringent regulation, have likely acted as a powerful headwind for credit growth in this segment of the market.
It is therefore not surprising that the mortgage debt obligation ratio also reflects a very sharp decline since 2008.
Given the large proportion of overall outstanding household debt by value accounted for by mortgages, is it possible that the deleveraging observed in overall debt since 2008 is being obscured by the relative stagnation in overall mortgage debt?
The chart below illustrates how fairly rapid growth in outstanding auto loans has meant that overall auto debt as a % of disposable income has increased notably since reaching a cyclical trough in 2010.
On the positive side, the current ratio remains below the cyclical peak recorded in 2004/5. With financing rates remaining near record lows, the overall debt service payment obligation ratio also remains well below the two cyclical peaks recorded over the past three decades.
The overall consumer debt service payment ratio including credit card and student debt reflects a similar picture.
If we take a look at total outstanding consumer credit including auto, credit and student debt as a % of disposable income, we can see that this ratio has moved to a new cycle and record high in recent years.
Why? This dynamic is likely explained in large part by the growth in outstanding student debt. Coupled with fairly robust growth in auto and credit card debt, total consumer debt has grown quite significantly since the 2009 recession. Nevertheless, while financing rates remain low, the overall debt service payment ratio does not presently exceed prior cyclical peaks. This would suggest that overall delinquency rates will remain at levels consistent with prior expansions.
Furthermore, consumers with existing mortgages are typically older and are thus unlikely to have any student debt. Given that mortgage rates in the US are typically fixed for long-periods of 15 to 30 years and coupled with the ongoing rise in house prices, these same consumers appear likely to remain in fairly healthy shape, provided the US housing market remains well supported.
However, the same cannot be said for younger and lower income consumers, especially those that currently do not own their own home. Many of these younger and lower-income consumers currently opt to rent rather than own their homes. Should interest rates further over the next 12 months, this particular segment of the consumer credit market could potentially become one area of concern.
This is particularly relevant given the likelihood that home affordability will deteriorate going forward. This would imply that the marginal mortgage debt service payment obligation ratio for younger first-time home buyers will also be much higher than the average ratio.
What does all this mean?
At a high level, the deleveraging that has taken place in the US economy since 2008/09, particularly with reference to mortgage debt in the context of still historically low financing rates, would suggest that overall delinquency rates will remain low. Interest rates could also still rise quite significantly from current levels before triggering any real general slowdown in consumer demand and economic growth.
However, consumer loans tend to be tied to variable rates which mean that as rates go higher the debt service payments associated with these types of auto and credit card loans will follow suite. As such, rising shorter-term rates will push auto and credit card repayment obligations higher. This will disproportionately negatively impact younger and lower income consumers, particularly those that continue to rent and will in all likelihood be facing rental price growth above that of inflation. This demographic segment could face a further erosion in real disposable incomes should energy and food price inflation accelerate and remain elevated over the next few years.
Given that most commercial banks have diversified lending books and income streams would a pullback in auto and credit lending have a significant impact on their profits? If less stringent regulations and lending growth lead to a renewed acceleration in mortgage credit growth, then the impact is likely to be quite muted on balance. On the negative side, specialized non-bank consumer and auto credit providers would possibly be more negatively impacted and could face substantial bad debt losses in specific cases.
What about the impact on the broader economy? Such a dynamic would negatively impact on retailers catering for millennials and/or low-income consumers, specifically sellers of vehicles and industries associated with these sectors at the margin. However, a significant pullback in demand in these areas leading to another financial crisis or even recession still appears unlikely.
Is there a scenario where rising rates in the US could eventually lead to a recession? At a high enough level, interest rate levels will eventually trigger a recession and the only question is that exact level or ‘tipping point’. We have argued on many occasions that the interest level the US can accommodate is probably much higher than many market participants in the past have suggested.
It is more likely that a further appreciation in the US Dollar, coupled with a policy error on trade (tariff imposition) combined with rising interest rates that would lead to another recession as opposed to a further substantial increase in the Federal Fund Rate in isolation. Nevertheless, there is also one further area that we highlight which could become a cause for concern over the next three years.
The ongoing disruption in the retail sector as more and more sales take place online has undermined many traditional ‘brick and mortar’ retailers. Many of these operations have announced the closure of 100s of physical retail locations. This dynamic has taken place with only some 10% of all retail sales having transitioned to the ecommerce environment. The question we need to ask ourselves as investors is what will the landscape look like if e-commerce constitutes 90% of all retail sales and not 10%? How many malls or strip malls will be left vacant and what does this all imply for commercial and industrial loan growth?
The chart below outlines how the delinquency rate for commercial and industrial loans has risen over the past year, but remains close to the cyclical trough lows registered over the last three decades. Some of the recent increase has also been driven by souring loans to the US energy sector, while many commercial property companies will also issue their own debt as opposed to tapping bank loans directly. This will nevertheless be another potentially vulnerable segment of the economy that should be closely monitored going forward.
A sharp rise in commercial property vacancies and delinquencies in this segment, coupled with a simultaneous increase in financial stress among younger and/or low-income consumers, would likely be sufficient to see overall lending standards tighten and create a sufficiently powerful headwind to induce a more generalized slowdown in consumer spending and the overall economy. On the positive side, such a dynamic, although a risk, does not appear an imminent threat to current expansion at present.