This morning I again wanted to challenge my somewhat bullish medium-term outlook but bearish longer-term view on the US economy – this time by looking at the data on debt. What follows is going to be a very numbers-heavy post. So, I apologize in advance if you are not a numbers jockey like me. But, do bear with me; I think you will find the analysis useful.
Now, as I write this paragraph, I have compiled the data, but have not yet dissected it. So I approach this without any definitive conclusions at the outset (although I must admit the last time I saw the data last year, they supported my thesis). Let’s see if the data still support my beliefs .
The Asset Based Economy View of America
My pre-conceived thesis is as follows:
- The U.S. has been living beyond its means for a generation as reflected in the increase of debt to GDP across a wide-spectrum of sectors of the economy.
- This increase has not been worrying to policymakers because they have only been watching debt service burdens, to the degree they have been tracking debt.
- Because of “the Great Moderation,” interest rates have fallen, permitting a secular increase in debt to GDP levels without increasing debt service burdens.
- The Federal Reserve has a dual mandate to support economic growth (through full employment) while maintaining low consumer price inflation (through price stability). Cognizant that debt services burdens were not acute and consumer price inflation was low, the Federal Reserve was able to target asset prices through lowering the Fed Funds Rate as a mechanism for reviving the economy when cyclical downturns occurred.
- As a result, the Federal Reserve under Sir Alan Greenspan followed an asymmetric monetary policy of only increasing interest rates slowly in the face of large levels of asset price inflation but reducing those rates very quickly to stem asset price declines.
- The result has been a belief that the Fed would save the economy when it ran into trouble, the so-called Greenspan Put. This has increased the appetite for risk in the financial sector and, most crucially, has meant that debt levels always increased after a brief downturn. The heroic actions of the Bernanke Fed have only increased this belief in the Fed as economic savior, sowing the seeds of the next asset bubble.
- This Asset-Based Economic Model can last through several business cycles – but will eventually collapse when debt service burdens become too large.
So, in sum, I believe that we are now poised to either a) collapse under the weight of debt only if debt service burdens are too much to bear or b) continue apace in the Asset-Based Economy until these burdens do eventually become crushing. I see b) [this originally said a) erroneously. A reader caught the mistake. Freudian slip?] as the more likely outcome during this cycle. Whether that crushing level of debt eventually comes as the result of a decline in incomes not matched by a decline in debt burdens (deflation) or via an increase in interest burdens not matched by an increase in incomes (inflation) is a question for another day.
What I want to look at here is the narrow issue of how debt burdens have moved at turns in the economic cycle. Specifically, I am about to examine the debt to GDP levels of specific sectors of the economy as presented by the Fed Flow of Funds right around recessions. And then I will compare these levels to the growth in nominal GDP and draw conclusions based on the data (debt cannot grow more than nominal GDP for long or it is a clear sign that growth is predicated not on sound investment and productivity but on leverage).
So, this is not an exercise in crunching the numbers to fit a conclusion, but rather a look-see at whether the data supports my thesis.
The Numbers: Federal Reserve Z1 Data Series
The Federal reserve releases a data series called Z1 every quarter. This is the basis of my analysis (link at the bottom). The Z1 series shows debt from the following sectors of the economy:
- domestic nonfinancial sectors credit market instruments, excluding corporate equities liability
- households and nonprofit organizations credit and equity market instruments liability
- households and nonprofit organizations home mortgages liability
- households and nonprofit organizations consumer credit liability
- nonfinancial business credit market instruments, excluding corporate equities liability
- nonfarm nonfinancial corporate business credit market instruments liability
- state and local governments, excluding employee retirement funds credit market instruments liability
- federal government credit market instruments liability
- total finance credit market instruments, excluding corporate equities liability
- rest of the world credit market instruments, excluding corporate equities liability
My pre-conception is that the sectors I want to key in on are the mortgage market (3), the household sector (4) and financial services sector (9)
What should be abundantly clear from the two charts below is that the U.S. has been growing in an unsustainable way since the recession of 1982. There was a brief period during the 1990-91 recession when the change in nominal GDP outstripped increases in debt levels, but that’s it (note, I use year-over-year change levels throughout). This is completely at odds with the preceding period in which every recession induced declines in debt to nominal GDP comparisons.
Debt levels at the end of Q2 2009 are 357% of GDP, a massive increase from the 160% that prevailed in 1982. The data clearly demonstrate that since 1982 the U.S. has relied on an increase in debt, even during recession, to avoid downturns. My thesis of policy asymmetry is, therefore, confirmed.
This chart is fairly benign when you look at aggregate levels as a percentage of GDP. Pundits forecasting an imminent increase in U.S. interest rates because of too much government debt have obviously not looked at these data. However, what is striking is the huge and unprecedented surge in debt as a percentage of GDP since the latest downturn hit. This discrepancy to nominal GDP cannot go on indefinitely. My general conclusion is that deficit spending can indeed continue for a long time without stoking an increase in interest rates given the low level of government debt as a percentage of GDP. This is bullish for U.S Treasuries in a muddle through economic scenario.
There is less here than I anticipated. One thing is clear: the household sector has breezed through the recessions in 1990-91 and 2001 without decreasing debt significantly. As a result, the increase in debt levels in the household sector are pretty astonishing. In 1952, it began at 24% of GDP, rising to around 40% by 1960, where it remained through the Ford presidency. Afterwards, it shot up again to its present 97%, four times the level a half-century ago.
This pattern is largely the same as the previous one.
Consumer Credit Debt
Consumer Credit seems to be much more volatile than mortgage credit. You can see the fluctuations in comparison to nominal GDP are greater. And the absolute amounts are much less than in the mortgage market. The conclusion I draw from this is that,to the degree household debt levels have increased unsustainably, it is mortgage debt which is to blame.
Non-Financial Business Debt
There is a lot more volatility in capital spending as reflected in non-financial business debt levels as well. Nevertheless, there has been a secular increase in debt levels of the business sector, from 30% in 1952 to the present 78%. The one thing to notice on the chart on the right is how short business cycles were pre-1982. It is more striking in that chart because business debt levels always adjust during recession.
State and Local Government Debt
Since the 1960s, state and local government debt levels have been basically flat as a percentage of GDP. There is not much to say here except to note the huge spike in the mid-1980s relative to nominal GDP. Can someone explain this for me? I think that area circled in red is quite intriguing.
Federal Government Debt
This chart looks basically the same with the total government debt charts as Federal Government debt dominates. What you should notice is that debt levels are lower now than they were in the 1950s and have just passed the post 1950’s high-water mark in 1993 of 49%. Again, this does suggest there is ample room for deficit spending without an increase in interest rates. The data are more favorable for treasuries than I had anticipated.
Financial Services Debt
This is probably the key damning piece of data confirming the asset-based economy thesis. The data are much worse than I expected. Not only do Financial Sector debt levels rise from negligible to percentages well over 100% of GDP, but the entire post-1982 period sees zero decline compared to nominal GDP until last quarter.
What conclusions can one draw here?
- The financial services sector is six times more important than in 1982 when its debt is measured as a percentage of GDP.
- The financial sector protected the American economy since 1982 by increasing its debt burden relative to nominal GDP even during recession.
- The financial services sector contracted in Q2 relative to GDP for the first time since 1982. If this is a rear-view mirror view, that means recovery could continue. However, if this is a canary in the coalmine, that is negative for the U.S. economy. This number bears watching.
There was an absolutely massive decrease in foreign debt relative to GDP when the economy was falling. Q4 2008 saw a gap of -12.4% between the change in foreign debt and the change in nominal GDP. The year-over-year differential has diminished as Q1 and Q2 2009 saw foreign debt increase, albeit to a level much lower than in Q3 2008.
Most of my basic beliefs regarding the asset-based economy are still intact. What now seems clear to me is the degree to which the post-1982 period is a departure from the one which preceded it. Moreover, the data on the financial services sector was surprisingly stark. I would go as far as to say that the US economy depends on leverage in the financial services sector to continue growing. I come out of this thinking it is the financial services sector more than the household sector dictating the course of events. And as the financial sector just began to really deleverage in Q2, it bears watching how this proceeds.
As for the household sector, aggregate debt levels are not decreasing substantially – not in mortgages or consumer credit. This may have changed in Q3 but given the fact that the worst of the recession was in Q4 2008 and Q1 2009, the data suggest that the consumer will not deleverage. If deleveraging doesn’t occur in the mortgage market, the household sector will not be the cause of a double dip. So, not having looked at debt service levels yet, I am not anticipating an imminent downturn in consumption demand or a further increase in savings – although this could change based on the data.
In the end, my somewhat more bullish medium-term outlook is justified by the data. Here, I am not talking about longer-term sustainability but the prospects of a multi-year recovery. Watch debt levels in the financial sector as a contrary indicator.
Z1 Data Series – Federal Reserve