I have been invited to join a panel discussion at the annual meetings of the National Association for Business Economics (NABE) next week, on the topic of the “The Great Deleveraging” after the financial crisis. Following is a summary of the thoughts that arose from preparing for this exercise, concentrating on the household sector. (The panel as a whole will cover all sectors of the economy.)
This financial crisis, like virtually all downturns and crises that preceded it, arose in considerable measure because of excessive leverage in some sectors of the economy. Important economic actors found themselves overextended and unable to meet their obligations or to spend, and their cutting back on their spending cascaded throughout the economy, causing other actors to find themselves overextended, and resulting in a massive downturn. Resumption of normal consumption and economic growth will require that those afflicted actors get their balance sheets back in order. The “paradox of thrift,” as many households try to restore prosperity by cutting back simultaneously and thereby further reducing sales and incomes, will afflict us until this “deleveraging” process is completed.
So, where do we stand in this deleveraging process? How much progress has been made? How long until more-normal borrowing, lending and spending is underway, and growth can begin to drag us back toward full employment of the work force?
Unfortunately, the answer is hard to discern. Historical patterns derived from much milder past episodes are not helpful. It would be just as useful – or useless – to ask what the garden-variety 100-year storm looks like, because a 100-year storm destroys a lot of gardens. This financial crisis exemplifies that sad pattern.
There are at least two dimensions to the complexity of the economy’s deleveraging process. The first is that the financial crisis did a lot of deep structural damage, and so soundness will not return in a matter of mere moments. Consider the market for commercial paper – unsecured, short-term securities issued generally by fairly large corporations. In the wake of the worst of the financial crisis, when basically no one trusted anyone else to return a borrowed dollar – even in the very short term – the volume of outstanding commercial paper plunged by almost 50 percent, and has not really recovered (see the following chart). Even though the commercial paper market had grown substantially in the middle of the decade, and so the effect of that drop should not be overstated, this decline is indicative of a barrier facing the economic recovery. It is analogous to a coronary blockage, which under the best of circumstances the heart must learn, painfully, to circumvent. It will take time, and it will slow the body in all its efforts in the interim.
Commercial Paper Outstanding
An analogous hit to the economy occurred in the housing market. The flow of housing construction was cut almost in half, in very short order. This was matched by a drop in outstanding mortgage credit. The sustained drop in absolute terms since the financial crisis was unprecedented in post-War (i.e., post-Depression) history. It far exceeded the mere slowdowns of growth in other recent past downturns (see the following chart).
Mortgage Debt Outstanding
The housing market provides a convenient segue into the second dimension of assessing the deleveraging challenge, which is the potential deceptiveness of averages. The late Arthur Okun used to illustrate this issue with a fable of a six-foot tall economist who drowned in a stream that was on average three feet deep. If one is watching deleveraging today because it is a painful but necessary prelude to the recovery of aggregate consumption, then information on the reduction of average leverage across the economy is probably useful. But if the concern is instead a possible continuation of defaults and bankruptcies, then the distribution of debt across individual borrowers, relative to those borrowers’ stocks of assets and income flows, is more important.
Looking at the housing market in particular illustrates the importance of distributions. To paraphrase former Speaker of the House of Representatives Thomas P. “Tip” O’Neill, all housing is local. The responses of local housing markets during and since the financial crisis have been extraordinarily diverse. At this point, markets – on average – are beginning to recover in terms of home prices (see the first following chart); but conditions in the most extremely affected markets remain painful (see the second following chart; note cities that had or did not have a boom, and had or did not have a subsequent bust, in all conceivable combinations and degrees). Thus, it is far from clear whether the worst of the housing- and mortgage-market tremors is over, or whether perhaps further damaging economy-wide aftershocks remain in our future – if any other factor (say Europe) were to trigger a market reaction.
Home Prices Are Just Stabilizing and Beginning to Recover
So because of those two big-picture concerns – the dangers from hidden structural damage, and the possibility of remaining harm from distributional outliers – my personal bottom line is that, as much as aggregate figures might suggest that deleveraging has proceeded apace, there remains reason for concern that our economy is not yet out of the woods with respect to excess leverage. The actual situation is more dangerous than the aggregate numbers might suggest.
The chart below, taken from work by researchers (Meta Brown, Andrew Haughwout, Donghoon Lee, and Wilbert van der Klaauw, found here) at the Federal Reserve Bank of New York (FRBNY), makes essentially two points: first, very nearly all of the action in additional leveraging of the household sector in the second decade of the 2000s came from housing finance; and second, the household sector has substantially delevered since the depths of the crisis.
Growth in Household Debt
The next chart shows a kind of digest of these data, highlighting where the changes in household-sector leverage occurred from the beginning (not at all to sanctify that date) of the FRBNY data series in 1999 until the height of the financial crisis in the first quarter of 2009, and then from 2009 until the latest data point. Again, the numbers make clear that first mortgage and home equity loans were by far the primary source of the increase in household debt, and the primary source of the subsequent deleveraging.
All of this deleveraging has come in a highly damaged and dysfunctional mortgage market. Were it the result solely of households choosing to buy smaller and put more money down, it would make us rest easier. And truth be told, there has been a lot of such prudent behavior. But there also has been deleveraging in the form of foreclosures, which wipe debt off the books in the most painful way. Some foreclosures are followed by resales, which would be expected to replace some fractions of the debt just wiped out. But other foreclosures have resulted in vacant, unsold homes, and the number of vacant units remains relatively high (see the following chart).
Many Housing Units Remain Vacant
There has been an enormous swing in the amount of change in mortgage debt from year to year, and the biggest component of the swing has been the change in the amount of new first-mortgage originations plus normal (non-charge-off) payoffs. But the amount of loan charge-offs is the biggest source of actual debt reduction (see the following chart).
So, that is deleveraging, isn’t it? What’s not to like? The answers to those questions should be obvious, in about 50 years – once economic history has a chance to sift through the rubble. For right now, however, we have only suppositions. Here are some possible reasons to keep the champagne on ice for now.
One imponderable is the ultimate effect of “deleveraging” that occurs because of loan writeoffs. In one sense, the absence of debt is the absence of debt. But in another, the debt was written off because the homeowners’ finances were wounded, perhaps fundamentally. Do we really expect those persons to celebrate their foreclosures, with the concomitant relief of debt, by running out and buying cars? They may have been relieved of mortgage payments, but they also were relieved of residences, and presumably they still need residences (unless the foreclosures were on investment properties, which were admittedly numerous). So it is not clear that there was much financial relief in those foreclosures.
Furthermore, the mass of foreclosures surely severely wounded the financial sector, which we count on to facilitate the investment and economic growth that are needed to achieve the sustained recovery that alone can end this melodrama.
According to the FRBNY data depicted earlier, the ratio of household debt to personal income has declined back to about its level of 2003. That year was a better year than 2009, to be sure, but far from the best year on record. Furthermore, 2003 was not beset by all of the structural damage caused by the subsequent financial crisis. Is the 2003 level of household debt the finish line of the deleveraging process for households? Can we be sure that it is? Can we hope? Personally, I am willing to hope, but I am far from sure.
Finally, a number of home mortgages remain underwater, given the limited recovery of home prices from their declines in the depths of the financial crisis. And as is widely known, income growth at the lower and middle segments of the income continuum has been less than stellar. The dislocation from the financial crisis has been visited disproportionately on a small number of households, highlighting the concern expressed earlier that any future bad economic news could cause a renewed round of foreclosures, defaults, and bankruptcies.
In sum, the household sector has seen significant deleveraging. But it is deleveraging of an unfortunate sort, involving a painful level of foreclosures, in a still-dysfunctional housing market financed by a still-dysfunctional mortgage and broader financial system. The aggregates of debt are apparently improved, but those numbers might flatter to deceive.