Data on past consumer deleveraging during recessions
I found the recent consumer credit data unsatisfying because the data seemed to point in two directions. The seasonally-adjusted data showed a large $12 billion decrease in consumer credit which received headlines. Meanwhile, the non-seasonally adjusted data showed a large $7 billion increase in consumer credit. I suspect this divergence has a lot to do with cash-for-clunkers (post here), but we can’t be 100% sure until we get another month or two of data.
Let’s delve into this issue a bit more because we are at a critical stage in the economy right now. A lot of pundits are talking about a V-shaped recovery, while others are pointing to a secular move toward consumer deleveraging which basically spells no recovery.
I am somewhere in between, believing we will get a weak recovery that could or could not last more than a couple of years depending on the path of consumer deleveraging. While I do believe the U.S. economy is fast approaching ‘terminal debt,’ I am not convinced we are there in this particular cycle, largely due to the massive stimulus and easing. The posts which best outline this view are two recent ones, “The recession is over but the depression has just begun” and “A brief look at the Asset-Based Economy at economic turns.”
So, here are my questions on consumer credit:
- What happened to consumer credit in past downturns?
- Specifically, how much did it decline from peak to trough?
- Was the decline mirrored by a similar fall in output as measured by nominal GDP?
- Is the pattern this time around any different?
- And what does that say about the thesis of secular deleveraging?
The way I am going to attack these questions is by looking at past recessions and peak-to-trough declines in both nominal GDP and consumer credit (note: the credit series is monthly, but the GDP series is quarterly – and I am leaving out the recessions in 1945 and 1949 because of wartime anomalies). Here goes.
- GDP declined 1.8% peak to trough. Consumer Credit declined 1.6%. It took 8 months to hit the previous credit peak.
- GDP declined 2.7% while credit declined 1.6% from January to June of 1968.It took 11 months to hit the previous credit peak.
- GDP declined 1.0% and credit also declined 1.0%.It took 8 months to hit the previous credit peak
- Nominal GDP did not decline due to inflation. Credit did decline from October to November 1970 by 0.2% however.It took 2 months to hit the previous credit peak.
- Again inflation meant nominal GDP did not decline. Meanwhile credit declined 1.8% from September 1974 to June 1975.It took 13 months to hit the previous credit peak.
- Again inflation meant nominal GDP did not decline. Meanwhile credit declined 1.8% from February to June 1980.It took 12 months to hit the previous credit peak
- GDP declined 0.3% and credit also declined 0.1%.It took 2 months to hit the previous credit peak.
- Inflation meant nominal GDP nearly did not decline (down –0.08%). Meanwhile credit declined 2.0% from November 1990 to well past the recession’s end in June 1992. It took an enormous 27 months to hit the previous credit peak. This was the worst credit cycle to date.
- We cruised through this recession with neither a decline in nominal GDP or credit.
- GDP declined 2.7% and credit has already declined a massive 4.6%. This is the largest decline in consumer credit to date.
The table of data looks like this:
I draw the following conclusions :
- In an economic downturn, consumer credit has generally declined more than nominal GDP.
- The 1990-91 recession saw a very drawn out drop in credit and easily generated the longest time before credit regained its previous peak.
- The 2001 recession was unusual mild as it was the only time in the last 56 years that the U.S. has experienced a recession without a drop in consumer credit.
- The 2007-2009 recession has been unusually severe. The drop in nominal GDP is the largest in the last 56 years. However, the drop in consumer credit has been even worse at more than twice the largest previous decline.
None of this necessarily helps me decipher whether we are in a secular deleveraging cycle. But, the severity of the fall-off in consumer credit relative to output may point in this direction.