The unsustainable US recovery, and why it matters

The disappointing GDP growth rate of an annual 0.7 per cent in the US during the first quarter of 2017 has raised questions about the prospect of a robust and sustained recovery. This is puzzling for several reasons. To start with, because of the adoption of an easy money regime the US has been experiencing an inflation in financial asset values.

Reflecting this tendency, the NYSE composite index has registered a trend increase from it early 2009 trough to reach levels higher than where it stood at its peak in mid-2007 (Chart 1). But this did not generate a strong wealth effect, in the form of increased private borrowing to finance consumption and investment, for two reasons: (i) consumers and banks were still unsure whether the financial turnaround would last and not be followed by a return to crisis; and (ii) wealth accumulation had occurred substantially among the already rich, increasing inequality but not spurring demand.

However, given the large infusion of liquidity and the imposition of negative interest rates on them, banks have been under considerable pressure to lend.

With the increase in liquidity swelling the volume of deposits on which some interest has to be paid and with returns from holding deposits with the central bank falling and even turning negative, bank profitability that had been restored from the low it touched in the aftermath of the 2009 crisis was under pressure.

This does seem to have increased lending by the banks to those who were willing or even desperate to borrow in the still-depressed economic environment.

According to data from the Federal Reserve of New York, total household indebtedness stood at $12.73 trillion as of March 31, 2017, which was $50 billion above its previous peak it reached in the third quarter of 2008 and 14.1 percent above the trough it touched in the second quarter of 2013.

In effect, a slow process of deleveraging that had begun in the third quarter of 2008, after the crisis, began to be reversed starting the second quarter of 2013, when total household debt was still 54 per cent above the level it was at in the first quarter of 2003 when the global liquidity surge began.

What is interesting is the composition of the increase in household credit in the US between the second quarter of 2013 and the first quarter of 2017 and the composition of the stock of household debt in the first quarter of 2013. Mortgage loans that accounted for 70 per cent of outstanding household debt in the second quarter of 2013 contributed only 50 per cent of the increment in household debt between second quarter 2013 and first quarter 2017 (Chart 2). On the other hand, auto loans and student loans that accounted for 7.3 and 8.9 of the second quarter 2013 stock, contributed 22.5 or 22.3 per cent of the increment in debt between that date and one quarter 2017.

In other words, close to 45 per cent of the increase in credit in the period when banks have been “forced” to lend, was on account of auto loans and student loans. Total outstanding auto loans at $1.17 trillion are up 66 per cent from the post crisis trough in 3q-2010 and the student loan total has more than doubled (to $1.3 trillion) relative to first quarter-2009.

There are a few implications that follow from this dramatic change at the margin in the composition of household debt. First, it appears that given the still incomplete process of deleveraging and the uncertainty associated with accepting housing assets as explicit or implicit collateral, banks are making an effort to diversify their lending away from the housing market to the extent possible.

Second, since loans for automobile purchases are small in size and short in duration and can be securitised to reduce risk of losses this is an area where lenders have chosen to move, only to find willing borrowers who are able and/or willing to take on such debt since they believe they can trade in the asset in case they are unable to service the debt.

Third, unemployment and changes in requirements in the labour market are driving the young to look for higher degrees at the cost of building up large debt. And despite the risks of default because of failure to obtain jobs that offer the required income stream, banks have decided to accommodate this demand as part of their diversification strategy.

Finally, debt financed educational spending does not directly result in material demand. On the other hand the magnitude of student debt because of high college fees and the burden of servicing that debt require deferring entering into mortgage agreements and postponing home ownership. So, the increase in this component of household debt, not only lacks the output expansion effects that borrowing for buying houses or cars would, but also reduces the demand for mortgage loans. Hence, the growth-inducing effect of this round of increased household borrowing is likely to be lower than it was in the past.

These factors notwithstanding, the ‘early’ return to borrowing on the part of households with half of the increment in debt being on account of mortgage loans has had a salutary effect on the housing market. The post-crisis decline in the housing price index, which had bottomed out in mid-2009, has been reversed since mid-2013.

As a result both financial asset prices and housing asset prices have been rising, though the latter began its rise after a considerable lag. A consequence of this asset-price inflation is that the ratio of the net worth of households and non-profits to personal disposable income has risen from its post crisis low in early 2009, with that rise gathering momentum after mid-2011.

Fortunately or unfortunately, depending on one’s perspective, this recent bubble has already run up against constraints. It is now becoming clear that auto loans were provided to many who did not have the ability to meet the debt service commitments involved, and banks did that because they diluted their standards and because they could still persuade investors to buy into securities backed by these loans. According to the Financial Times of May 30, 2017, which quotes Morgan Stanley, the share of auto securities tied to “deep subprime” loans — those given to borrowers with scores below 550 on the commonly-used FICO creditworthiness scale — rose from 5.1 per cent of total subprime deals in 2010 to 32.5 per cent last year.”

Now defaults are rising as many borrowers are unable to service the debt and unable to pay it off by selling the asset, because of a collapsing used car market that has brought prices down to levels below the value of unpaid debt. The situation with student loans is worse. The percentage of loan balances going into “serious delinquency” has been hovering around a 10 per cent annual rate over the past five years.

As a result banks are pulling back, undermining even this limited stimulus that could spur recovery.

COMMENT NOW