Consumer credit down, but does it show deleveraging?
I have just taken a look at the consumer credit figures for September, released just yesterday by the Federal Reserve. The data do show some modest deleveraging, especially when looking at the recent increase in nominal GDP. However, it is still not clear to me that the scale of deleveraging is great enough to induce a recessionary relapse.
My baseline for deleveraging is Debt to Nominal GDP – when debt to GDP goes down, that shows deleveraging. For example, for the latest data released in September for Q2 2009, Private sector total debt to GDP (incl. financial services) in the U.S. was 292.2% of GDP. Because of the huge drop in nominal GDP, this was actually up from 283.0% when the recession began in Q4 2007. For households, the number was 96.8% in Q2 2009, up slightly from 95.9% at the end of Q4 2007. What this shows is that deleveraging has yet to begin in earnest as debt levels have remained relatively high even while GDP had collapsed.
The lack of deleveraging is probably a result of financial stress. In the Great Depression, the personal savings rate dropped from 4.5% in 1929 to 4.1, 3.9, –0.9, and finally -1.5% in 1930-1933. People had to use savings to service debt as the deflationary spiral took hold.
So, in the absence of quarterly data on debt levels, I look at data from things like consumer credit for a proxy. On a seasonally-adjusted basis, consumer credit declined to $2.471 to $2.456 trillion. That is the lowest since June 2007 and marks the ninth consecutive monthly drop.
However, looking at the non-seasonal data makes plain what is happening:
Nonrevolving credit is now increasing along with GDP. Look at the area highlighted in red; that coincides with the 3.5% real GDP print we just saw. On the other hand, revolving credit is getting crushed. Below is the reason why (click to expand):
Credit from commercial banks and savings institutions have dropped off a cliff. When you hear people saying that banks aren’t lending, this is what they are talking about. In Q3, banks are lending again (think cash for clunkers) because nonrevolving debt is up. That’s also why GDP is up. But, revolving credit lines (credit card lines) are being cut.
My conclusion is largely the same as last month, namely I had anticipated more deleveraging than we are seeing. However, consumer credit is only coming down on the nonrevolving side. And given the stabilization in house prices and increases in refinancing activity, I wouldn’t expect mortgage debt levels to be down substantially. When we see Household Debt to GDP levels from Q3, they probably will not be substantially lower than they were in Q2.
This does support recovery but only at the risk of continued high levels of debt to GDP.
G.19 data series – Federal Reserve
Wouldn’t the most appropriate way to determine whether this was supportive of a recovery from a consumer standpoint via looking at the ratio of credit to personal income?
Furthermore, since we live in a class-based society wouldn’t it be important to look at the levels of credit compared to income for various income quintiles, whose disposable income levels will vary widely even though an aggregate or average number will disguise this?
Wouldn’t the most appropriate way to determine whether this was supportive of a recovery from a consumer standpoint via looking at the ratio of credit to personal income?
Furthermore, since we live in a class-based society wouldn’t it be important to look at the levels of credit compared to income for various income quintiles, whose disposable income levels will vary widely even though an aggregate or average number will disguise this?
credit to personal income would do about the same thing, yes. But, Debt to GDP is a more generally recognized statistic.
Getting credit by class would indeed be revealing, but that data is not available.
Hours worked has declined by ~5% annually, and presumably that translates into lower income, but GDP doesn’t reflect that. in fact, GDP could still be positive while hours and income decline, ie higher productivity.
Isn’t it possible to get credit card credit balances related to income?
And if not, couldn’t we use concrete proxies? Say high-end retail store credit cards or AmEx for wealthy and gas cards
and home mortgage debt related to income?
For housing, it should be easy to get data for mortgage debt amounts for different price classes of homes, which we could infer roughly correlates to income levels.
And maybe similarly for car notes.
Shouldn’t that be “revolving credit lines (credit card lines) are being cut”?
yes. I will change that.
“…it is still not clear to me that the scale of deleveraging is great enough to induce a recessionary relapse.”
Maybe so, but how does a lack of deleveraging help sustain a healthy economic growth? I certainly don’t consider a debt-driven recovery to be “healthy” in any sense of the term, just the promise of another financial collapse down the road.
key word ‘healthy.’ I certainly don’t see continued high leverage as healthy. It does sow the seeds for more pain to come.
You could make a case for the NSA non-revolving credit increase been the result of CFC. Besides debt/gdp in the short term is going to be influenced more by gdp than debt. Individuals can’t deleverage that fast……only to the extent they realign consumption, income and debt reduction. A very slow process at first.
How does defaults and foreclosure figure into debt to GDP and deleveraging?
Home prices have fallen and a significant number of homes have been taken by the banks. How does the debt “forgiveness” through the foreclosure process get accounted for in consumer debt? Or is the consumer debt level partially offset when the home is sold by the bank to a new consumer who then takes a new mortgage?
What about mortgage modifications?
What about deed in lieu of and renting back to the consumer?