Is the consumer really deleveraging? « naked capitalism

Thursday, October 8, 2009

Is the consumer really deleveraging?

Submitted by Edward Harrison of Credit Writedowns

Why is everyone saying consumer credit is falling? It’s not. But, everywhere I look, everybody is saying it is.

I would like to be true to the data and not just take the government’s seasonally-adjusted numbers at face value.

Judge for yourself. Here’s the data:

This is what everyone is focused on – the seasonally-adjusted data. The part in red shows consumer credit down $12 billion.

But, what about the actual unadjusted data?

What do you know, it’s up $7 billion. It is indeed down $4 billion for revolving credit as banks are cutting credit card limits. But, non-revolving credit is up over $11 billion. It was decreasing and is down 4.4% year-on-year (see the section highlighted in green above), but that ended this month.

Yes, I too believed that consumers were poised to begin deleveraging, but with stocks up 60%, interest rates at record lows, and house price declines stalled, why would you do that?

Conclusion: consumer credit is increasing, not decreasing. I wish people would actually look at the data.

The question you should be asking is not whether consumer credit is increasing, but whether it will continue to do so after August and cash for clunkers.

And I did a full review of the asset-based economy during economic turns yesterday. All indications are that the consumer is not deleveraging as I would have anticipated (see post here).

Sources

G.19 Current – Federal Reserve

G.19 Historical – Federal Reserve

53 Comments:

bob says:

October 8, 2009 at 4:54 pm

My rough math on the cash for clunkers program worked out to about 12 billion.

Ethan says:

October 8, 2009 at 4:57 pm

My rough math on the cash for clunkers program worked out to be about 53 gajillion dollars.

Anon says:

October 8, 2009 at 5:05 pm

Good topic, but I am not persuaded. Isn’t the year-over-year data, which shows a decline in credit, a whole lot more probative than a shift over one month, particularly when there has been, recently, so much stimulus (home buyer tax credit, cash for clunkers)in the system?

Edward Harrison says:

October 8, 2009 at 5:21 pm

Exactly. That’s why my post reporting the credit data was titled “Consumer credit falls 4.4% from year ago levels.”

It’s purely speculative at this point whether these declines will continue given the fact that the unadjusted data showed an uptick in August. Reports you read to the contrary show bias.

The numbers I crunched yesterday suggest that the drop in household debt mirrors the drop in nominal GDP and that consumers are not deleveraging. Moreover, the latest data shows an uptick in mortgage demand as 5-year ARMs and 5-year fixed rates are at record lows.

So, while I had anticipated deleveraging before I did the analysis, it is far from clear that this deleveraging is going to happen. Now, I lean more toward thinking there will be no deleveraging.

We will have to wait and see.
- Jake says:
  
  October 8, 2009 at 7:59 pm
  
  As I posted over at your blog…
  
  Why disclude a seasonal adjustment? I am not positive as to why it would naturally jump in August, but back to school shopping seems like one obvious answer. I guess my question would be, would you want to ignore seasonal adjustments for toy sales in December too?
  
  If you want to ignore those trends, then take the unadjusted data and look at year over year figures. August ’09 was down 4.4% compared to down 4.1% in July ’09 vs. July ’08. In other words… consumer credit is falling.
  - Edward Harrison says:
    
    October 8, 2009 at 8:08 pm
    
    Jake, you can look at the y-o-y data for SA or NSA data and they say the same thing. We are down.
    
    When I reported the data I said:
    
    The Federal Reserve has just released the most recent data on consumer credit. The data show outstanding consumer credit falling to $2.47 trillion in August from a December 2008 peak of $2.59 trillion – on a non-seasonally adjusted (NSA) basis. That is down 4.4% from the year ago period, continuing the acceleration of the year-on-year change that has been in place for 15 straight months. The seasonally-adjusted data tells an even worse story.
    
    http://www.creditwritedowns.com/2009/10/consumer-credit-falls-4-4-from-year-ago-levels.html
    
    You would find the same for the SA data.
    
    But looking at SA month-to-month data in isolation and saying the sky is falling completely ignores the fact that consumer credit just ticked up for the first time since December. One could explain this via cash for clunkers, but to overlook it doesn’t make sense to me.
  - Edward Harrison says:
    
    October 8, 2009 at 8:15 pm
    
    I should also add that my reporting of the data was not bullish by any stretch – and I focused on the year-over-year comparisons which are accelerating to the downside.
    
    But, nonetheless I did point out the increase and, given other research I have done, it is far from certain we are witnessing consumer deleveraging.
    - rootless cosmopolitan says:
      
      October 8, 2009 at 10:23 pm
      
      Saying now it wasn’t certain whether the consumer is deleveraging sounds to me like back-pedaling a little bit from your rather definitive statement in your original post that the consumer isn’t deleveraging based on the m-o-m increase in the not seasonally consumer debt. Your first post sounded like that you are just dismissing the seasonally adjusted data and take the nominal m-o-m data to be showing the real picture, indeed.
      
      Anyway, the trend looks like deleveraging to me and like a significant change to the previous increase in debt, not seasonally adjusted:
      
      http://research.stlouisfed.org/fred2/series/TOTALNS?cid=101
      
      or seasonally adjusted:
      
      http://research.stlouisfed.org/fred2/series/TOTALSL?cid=101
      
      The m-o-m changes of the not seasonally adjusted data are extremely noisy, as you can see, if you click on the respective link below the graph. Thus, I don’t really know why you even would try to draw any conclusion from those. The m-o-m change of the seasonally adjusted data are noisy as well, although the noise seems to be reduced somewhat compared to the not seasonally adjusted data. However, the general trend shows a much clearer picture.
      
      rc
    - Edward Harrison says:
      
      October 8, 2009 at 10:37 pm
      
      rootless, please read my post before you respond. I clearly point to y-o-y data:
      
      http://www.creditwritedowns.com/2009/10/consumer-credit-falls-4-4-from-year-ago-levels.html
      
      And the post headline is for a 4.4% decline.
      
      As for what I actually said regarding deleveraging and credit decreasing/increasing, these are the exact quotes:
      
      From when I posted the data yesterday:
      
      “The Federal Reserve has just released the most recent data on consumer credit. The data show outstanding consumer credit falling to $2.47 trillion in August from a December 2008 peak of $2.59 trillion – on a non-seasonally adjusted (NSA) basis. That is down 4.4% from the year ago period, continuing the acceleration of the year-on-year change that has been in place for 15 straight months. The seasonally-adjusted data tells an even worse story.”
      
      From this post:
      “consumer credit is increasing, not decreasing. I wish people would actually look at the data.
      
      The question you should be asking is not whether consumer credit is increasing, but whether it will continue to do so after August and cash for clunkers.”
      
      I am clearly stating that the trend in credit has been down. However, this months NSA change needs to be watched as a contrary indicator.
      
      “It is still not clear to me that debt levels, when measured as a percentage of GDP are decreasing significantly. I am, therefore, much less sold on the prospect of a secular deleveraging in the household sector than I was yesterday.”
anon says:

October 8, 2009 at 5:22 pm

Right.

Biased data mining.
- Anon says:
  
  October 8, 2009 at 6:32 pm
  
  Hmmm, I was the first Anon here…the second Anon is someone else. The hazards of being Anon. Anyway, if we’re not deleveraging, it’s only because of federal actions. Which aren’t sustainable indefinitely, but might be sustainable “longer than the bears can stay solvent” to paraphrase John Maynard Keynes. So I think what we’re really trying to evaluate is how long fixed-income markets will tolerate the Fed and Treasury being this aggressive.

albrt says:

October 8, 2009 at 5:41 pm

I think the deleveraging is going to happen. It just isn’t going to be voluntary. So if anybody was lending last month, somebody was borrowing. No potential source of credit will be left on the table. Including all those off-budget government guarantees.

Edward Harrison says:

October 8, 2009 at 5:53 pm

For the aggregate deleveraging to be involuntary, we need a source of credit stress: 1. high interest rates; or 2. declining income through job loss; reducing credit demand or 3. capital-constrained banks limiting credit supply.

The point of maximum rate of stress on all of these variables has passed. So unless we see a voluntary deleveraging (i.e. balance sheet recession) I suspect aggregate deleveraging won’t be involuntary.

Vinny G. says:

October 8, 2009 at 5:46 pm

My rough math on the cash for clunkers program worked out to about twenty bucks…

Vinny G.
PS — the only way consumers are deleveraging in this economy is by default

Ben says:

October 8, 2009 at 5:53 pm

If they aren’t deleveraging, then the pain hasn’t even started yet for the broader economy – and China.

ScottB says:

October 8, 2009 at 6:40 pm

I’m not sure that contrary views show bias–they just take a different view of seasonal adjustment.

A different question, though: is the y/o/y decline due to a) voluntary consumer belt-tightening; b) involuntary belt-tightening due to loss of job/income; and/or c) banks tightening up their lending?

Thomas Barton, JD says:

October 8, 2009 at 7:05 pm

Monsieur Harrison, Please pick a mall in New jersey or Missouri or California or Arizona and go there and ask the retail store owners and their employees about their sales over the last 2 months. It is appalling that a Depression can be hidden in plain sight and then rationalized by Fed-backed paper stats. Benjamin Disraeli was never more right than today in this country, there are lies, damn lies and statistics.

Edward Harrison says:

October 8, 2009 at 7:10 pm

Have you seen same store sales for September?

http://online.wsj.com/article/SB10001424052748703746604574460740944888098.html

They are up. The point is not that we are about to vault into a V-shaped recovery (the data brings us back to 2005 levels). I too believe this is a depression. But, it just isn’t true that consumers are deleveraging and not spending.

Dave Rosenberg took the SA data and made it sound like the sky was falling when in fact it was all due to seasonal adjustments. To me, that is bias.

When the September and October data confirm the downward pattern, we can say something.
jack in dallas says:

October 9, 2009 at 9:49 am

Pick a mall in Houston, Dallas, Atlanta, etc. and one wonders “What recession?” New Jersey is probably the worst example the could have been chosen since your politicians have regulated and taxed your state to its knees.

Skippy says:

October 8, 2009 at 7:09 pm

Reminds me of slot/pokie machine math ie: bet 5 dollars, machine reports win of 2 dollars, not loss of 3, players mental conditioning via light, sound, win, loss, input precludes them from realizing true sums. Now to further the conditioning, from time to time, random chance awards of 20 dollars occur (see home buyer tax, clunkers, HTM, etc) to further inhibit sound mental processes aka critical thinking.

Disclaimer: Have friend from Boulder Co who made his fortune in slot machine programing and design (hes kicking it down by the lake now, out side of Vegas). Machines are depending on day and time, to pay anywhere from 1 in 7 to 1 in 20 with rolling pay out values…umm mm sounds a bit like Wall St. these day…eh.

Skippy…Some where in Afgainistan…Sir permission to use HTM theory with regards to next combat patrol mission planing…good thinking troop we can use our high tech pretend and extend strategy to win this war…Victory is close!…if we believe in it! Now klick you heals together 3 times and say we will win…we will win , we will win.

Sir radio report coming in from combat patrol using HTM…they report that HTM is a total failure as the enemy is using MTM and their LEAD is real!…pray son…pray…that is all we can do for them now.

ScottB says:

October 8, 2009 at 7:19 pm

Nonfarm employment rose by 300,000 jobs in September. Is it bias to focus on the seasonally-adjusted numbers instead of the nominal? I don’t agree with your approach here at all (which is the exception–I usually take away a lot from your posts).

Edward Harrison says:

October 8, 2009 at 7:27 pm

It would be if the household survey were saying something different. For example, in the last Employment Situation Survey, there was an enormous difference between the number of unemployed in the household survey (700K+) and in the NFP (~260k). Rosenberg and many others pointed out this discrepancy, saying, in effect “the employment situation is much worse than it appears.”

I agreed that when there is a discrepancy, you HAVE to as questions. You cannot just blindly accept one data series in isolation as fact.

Another example. The Halifax just came out saying that y-o-y UK house prices are 7% lower than at this time last year. Meanwhile Nationwide was singing a different tune:

http://www.creditwritedowns.com/2009/10/uk-house-prices-up-again-in-september.html

The discrepancy was huge. It makes you sit up and take notice. Not asking why the data is pointing in different directions means you miss part of the puzzle.

ScottB says:

October 8, 2009 at 8:17 pm

Ed, thanks, bear with me here– what’t the discrepancy in the consumer credit data that makes you discount the seasonally-adjusted series?

Anon says:

October 8, 2009 at 8:24 pm

Well put, ScottB! Ed, that’s the question…there’s normal fluctuation in the data from things like back-to-school spending in August. What makes you think that the seasonal adjustment is wrong or misleading in this case?
- Edward Harrison says:
  
  October 8, 2009 at 8:49 pm
  
  You have to question ALL seasonal adjustments (employment, jobless claims, consumer credit, etc) as we are potentially at a turn in the business cycle. The adjustments are made in advance and they are always a swag.
  
  Tell me this? Should I say:
  
  “The big story yesterday was the further massive $12 billion decline in outstanding consumer debt in August — the consensus was looking for an $8 billion contraction. This was the seventh month of debt retrenchment in a row. In other words, the tidal wave of the credit collapse continues unabated, and this is the primary reason why bond yields are still in a fundamental downtrend. Keep in mind that the $12 billion (5.8% annualized decline) reduction took hold during the peak for cash-for-clunker auto sales in August when they soared temporarily to a 14 million annual rate (and subsequently to 9.2mln in September).”
  
  That’s what David Rosenberg is saying and it’s not accurate. Consumer credit increased significantly ($11 billion on non-revolving credit – precisely because of cash for clunkers.)
  
  Do you see that as an accurate depiction?
  - eric says:
    
    October 9, 2009 at 1:02 am
    
    “That’s what David Rosenberg is saying and it’s not accurate. Consumer credit increased significantly ($11 billion on non-revolving credit – precisely because of cash for clunkers.)”
    
    People who could buy cars under the clunkers program were making a rational decision to do so, I guess, if they have any kind of income security. But that program is over with. Maybe consumer credit declines again immediately, maybe it doesn’t. With the economy continuing to bleed jobs (at a lesser rate, but still bleeding), obviously this is going to have an effect on credit. Maybe not this month, but the next, or the next.
    
    I appreciate your pointing out the raw numbers, though. We must accept what is, not rationalize it away or “adjust it” to what we expected.
    
    With state budgets coming under pressure (my state Iowa, far from the hardest hit state, is talking about 8% across the board budget cuts) and more gov’t employees being furloughed or laid off, or salaries/benefits scaled back, this is also going to have an increasing impact on the economy. The “fake recovery” is not likely to last long. Who is hiring? Who is getting salary increases? Maybe a few smart folk will sell their stock now and have a fling. Otherwise, it’s only a pause before more pain.
Edward Harrison says:

October 8, 2009 at 8:43 pm

I don’t discount the series, I look at it in concert with the NSA data. But since they are telling me different stories, I prefer the NSA data, using the y-o-y data.

I looked at the seasonal pattern in the historical data set here (also linked above):

http://www.federalreserve.gov/releases/g19/hist/cc_hist_mh.txt

And I ran through the numbers from May-August to see if there were aberrations like what we are seeing now (in which we saw NSA up $7 billion and then falling again). I couldn’t find any examples.

What I did find was 1991 data showing an uneven pattern of back and forth – but this was after the recession was over and during the recovery.

The only conclusion I can draw is what I said before:

“It is still not clear to me that debt levels, when measured as a percentage of GDP are decreasing significantly. I am, therefore, much less sold on the prospect of a secular deleveraging in the household sector than I was yesterday.”

My biggest beef with the reporting of these numbers is that, while everyone said essentially what I did – that y-o-y data was falling – a lot of people (David Rosenberg, for example) looked at the m-o-m SA data and made it seem like August was worse than ever. It was not.

Mark says:

October 8, 2009 at 9:06 pm

Edward, first of all, I really appreciate what you have to say, which includes some brilliant commentary (at least in my view).

You seem perplexed by the fact that consumer credit is not shrinking, but isn’t this entirely consistent with the “fake recovyer” thesis? Wouldn’t you expect the fake recovery to provide an artificial boost to consumer credit? Once the fake recovery ends, consumer credit should again begin to shrink.

Edward Harrison says:

October 8, 2009 at 9:16 pm

Mark. That’s exactly right – the fake recovery SHOULD be goosing the debt numbers. Honestly, I hadn’t expected it to goose them so much. I would have expected some deleveraging given the financial distress.

But, ALL of the debt reduction we have seen to date is matched by falls in nominal GDP. Put another way, debt to GDP levels are HIGHER now than when the recession began. That – I did not expect.

So, if we are indeed in a technical recovery, we are entering it with debt levels that are even higher than in 2007. It’s absolutely crazy.
- Mark says:
  
  October 9, 2009 at 12:04 am
  
  Edward, thanks for the response. It is bizarre to think that a recovery could take hold while consumer debt levels increase relative to GDP. This would make the recovery seem all the more fake. However, we are in truly uncharted territory, so who knows?
- SidFinster says:
  
  October 9, 2009 at 3:47 am
  
  I don’t understand why some here love to hate Ed so much, but to me this behavior isn’t exactly crazy. It’s more like junkie logic (or maybe even a rational bet on a bailout.)
  
  To the extent deleveraging (which, along with saving money, is communism!) has not occurred: American consumers will deleverage only when and to the extent to which noone will continue to loan them money.
  
  Until then, they are something like junkies, hooked on easy money. And money is one hell of a drug (almost like catnip).

Rahul Deodhar says:

October 8, 2009 at 9:47 pm

I am not sure why the consumer should not spend away, particularly when everything else is “back to glory days”.

Why should consumer simply sit back and take up the tax burden while everyone is having a party? It is like asking a kid to do his chores while we are celebrating kids party.

Those who can lever up may still lever up – mostly through housing (though there may not be sub-prime loans anymore), cars and such larger assets. This does have very small positive effects on corporate earnings in short term.

Same-store-sale, in limited cases, benefits in short term from store closings. We cannot interpret SSS same way on upside and downside.

So net-net I think pain is a further away in future. Possible trigger could be dollar movement.

Edward Harrison says:

October 8, 2009 at 9:49 pm

Well said.
Skippy says:

October 9, 2009 at 4:14 am

How about setting standards for kids or ethical reasons.

I noticed the *we are celebrating* bit too.

When the ruling class/enabling priestly class BBQ starts will you be on the menu (see historical parallels).

Skippy…BTW I make a mean BBQ sause from scratch and well versed in the gutting and cleaning of pigs, monkeys and other beasts.

Ingolf says:

October 8, 2009 at 10:06 pm

Edward, I’m with ScottB; I also tend to really enjoy your posts, but I think you’re wrong on this one.

Out of curiosity, I looked back through the July/August data (SA and NSA) from 2001 to 2009. The average adjustment factor was $14.23 billion, ranging from a high of $21.2 billion in 2008 below of $9.8 billion in 2001. This year, as you’ve shown in your figures, it was $19.3 billion.

So, the NSA Consumer Credit data is consistently adjusted downwards by a substantial amount. As others here have suggested, presumably there’s a good reason. What’s so special about this year’s adjustment?

Edward Harrison says:

October 8, 2009 at 10:30 pm

There isn’t anything special. But, explain to me why you would say this:

“Keep in mind that the $12 billion (5.8% annualized decline) reduction took hold during the peak for cash-for-clunker auto sales in August when they soared temporarily to a 14 million annual rate (and subsequently to 9.2mln in September).”

When actual credit increased – probably because of cash for clunkers. I don’t know why people don’t get the logic

1. There was a large increase in actual credit – the first in 8 months.
2. It is unclear whether this increase was due to cash for clunkers or seasonal patterns.
3. It is legitimate to say credit is still declining (especially if one uses y-o-y data).
4. However it is illegitimate to say “Keep in mind that the $12 billion (5.8% annualized decline) reduction took hold during the peak for cash-for-clunker auto sales in August when they soared temporarily to a 14 million annual rate (and subsequently to 9.2mln in September)”

It should be “the NSA $11 billion increase took hold during the peak for cash-for-clunker auto sales in August when they soared temporarily to a 14 million annual rate. Expect a fall to coincide with 9.2mln in September)”

Look at those two statements. Are you following the logic here?

Let’s have this conversation in a month’s time. What would you say if credit doesn’t show the plunge that even I expect due to C4C?
- Ingolf says:
  
  October 9, 2009 at 7:41 am
  
  Edward, what consumer credit does (or doesn’t do) in September isn’t relevant to the procedural issue I was trying to highlight, which is whether the seasonal adjustments applied to the raw data are reasonable.
  
  If they are, I don’t see anything inappropriate in Rosenberg’s comment; surely he’s simply saying that (given the cash for clunkers program) the NSA figure “should” have risen by much more than $7.3 billion in August. Last year, for example, from July to August it rose by $14.9 billion while the seasonally adjusted figure fell by $6.3 billion. Was this adjustment also questionable in your view?
  
  I guess it’s just not clear to me what you’re trying to argue when it comes to this data point. If you’re suggesting the whole seasonal adjustment process for consumer credit is fundamentally flawed, then I think that case needs to be far better argued. Perhaps it is (although I confess I doubt it), but simply asserting that it’s so won’t do.
  - Edward Harrison says:
    
    October 9, 2009 at 7:49 am
    
    That’s not my read of what he is saying at all.
  - Edward Harrison says:
    
    October 9, 2009 at 7:52 am
    
    He’s making the logically inconsistent argument that credit declined despite an increase in auto sales via cash for clunkers by using seasonally adjusted data. I imagine he didn’t realize that was what he was doing and simply made a mistake.
    
    And yes I am saying I won’t trust the seasonal adjustments until they are confirmed by additional data.
    - Ingolf says:
      
      October 9, 2009 at 8:59 am
      
      I don’t see the logical inconsistency you do, Edward, but we seem for the most part to be talking past each other so I’m happy to just let it go.
    - Edward Harrison says:
      
      October 9, 2009 at 9:09 am
      
      Ingolf, I’ll give it one last go though. I believe this is what happened. The number as reported is seasonally-adjusted. Reporters ran with this number as the always do, reporting it as is.
      
      David Rosenberg, who is great and with whom I generally agree on the macro picture, did as well. I don’t think he noticed the fact that the decline was seasonally adjusted and that the actual numbers showed an increase. Instead of noticing that the numbers were seasonally-adjusted and that actual credit was up, he made the argument that it was incredibly bearish that credit was down despite cash for clunkers.
      
      Everyone ran with this interpretation, which is wrong. And I think it does us all a disservice that they did.
      
      A better (but still Rosenberg-friendly interpretation) is the one I gave, namely that credit was up in August but it was probably due to cash for clunkers. Let’s wait until the next two months’ data comes out to see. If the data do show a return to the bearish side, fine. If not, we need to re-think whether the great deleveraging is upon us.
      
      By the way, I also did a write-up of how consumer credit has fared in past cycles.
      
      http://www.creditwritedowns.com/2009/10/data-on-past-consumer-deleveraging-during-recessions.html
      
      This analysis does support Rosenberg.

Edward Harrison says:

October 8, 2009 at 10:54 pm

I will make this general statement. You have to look at the data as they are without any preconceived notions of what they mean.

When the actual underlying data points in a new direction, you SHOULD stand up and notice. When the trend in two data series that were moving in concert diverge, you SHOULD ask questions.

Here are the questions I ask:
1. Is the decline in output responsible for the reduction in household debt we have seen to date? or is it a secular change in debt?
2. Was the increase in actual consumer credit in August a cash-for-clunkers induced aberration? or was it a switch toward more consumer spending as reflected in upside surprises in same- store sales. And is this why we are hearing the word recovery everywhere?
3. Even if consumers do increase debt levels, will economic weakness put the kibosh on this?

I simply do not accept that you can look at this without asking those questions or at a minimum waiting for more data to judge the trend.

Consolidated Aubergine says:

October 8, 2009 at 11:22 pm

From Capital Economics below, note while this refers to seasonally adjusted figures it provides a convincing explanation why consumer credit on the non-revolving side ticked up in August. Presumably, once C4C wears off in the Sept data we should see some return to trend in the non-revolving credit figures… all things begin equal.

“The smaller decline in US consumer credit in August is not a sign that households’ financial situation is improving.

Instead, it is entirely due to the temporary surge in auto loans associated with the now defunct cash for clunkers scheme.

Consumer credit fell by $12.0bn in August compared with a $19.0bn drop in July (revised from a $21.6bn fall).

Non-revolving credit, which includes auto loans, fell by $2.1bn, much less than the $16.6bn decline in July. This presumably reflects an increase in auto loans as households borrowed more in order to take advantage of the cash for clunkers scheme.

The downward trend in non-revolving credit is likely to continue in September. In contrast, revolving credit (e.g. credit cards), fell by $9.9bn, a much larger decline than July’s $2.4bn fall. In the last six months, households have reduced their total consumer credit by roughly $90bn. If maintained, this would decrease total borrowing by $190bn per year. That’s not bad, but there is a long way to go if households are to decrease their total debts by $3,000bn, which we think is necessary to bring the household debt to income ratio back to a more sustainable level of 100%.

Falling employment, declining incomes and a desire to pay back debt will mean that consumer credit will continue to fall for the next few years, undermining consumption growth and the sustainability of the economic recovery.”

ScottB says:

October 8, 2009 at 11:36 pm

Ed, thanks for explaining your reasoning, it’s given me something more to ponder. I also charted out the NSA for SA and it gave me a good visual for what you’re talking about. We shall see!

It is rather startling to look at the entire history–it’s basically an exponential curve, with the only pause in monotonic growth being a sideways movement in the early 1990s–until a year ago. The decline is noteworthy in the post-WWII era.

In terms of your questions: in the first question, why decline in output, which seems to be somewhat indirect, as opposed to job/income loss? Or am I missing your point?

On number 2, as a third option, why not a sub-normal seasonal uptick (thus the negative SA)?

Thanks again.

Edward Harrison says:

October 8, 2009 at 11:52 pm

Scott, I could just as easily use income. It generally moves in line with GDP.

And, yes there could be something askew in the seasonal adjustments, which is why I would rather use y-o-y data to describe the events.
- rootless cosmopolitan says:
  
  October 9, 2009 at 1:10 am
  
  Ed, how can a decline in GDP/income reduce household debt, in another way than by purging debt from the system due to defaulting borrowers who can’t pay off debt and interest anymore? What is the causal relation?
  
  rc

FT.com | Money Supply | Economic news digest says:

October 9, 2009 at 8:20 am

[...] of the signs of long-term good news for the American economy had been consumers reducing debt. But is the consumer really deleveraging? Two datapoints suggest not. First, while balances of revolving credit (e.g. credit cards) are [...]

FT.com | Money Supply | Economics headlines says:

October 9, 2009 at 8:33 am

[...] Consumers deleveraging (2) Non-revolving credit is up – naked capitalism [...]

Myron says:

October 9, 2009 at 10:15 am

I wonder how long it will take for the market to realize that we’re ALREADY in the next bubble.

The only reason the economy hasn’t completely collapsed into a depression is that the Fed inflated the largest bubble they possibly could, but all anyone can focus on is that we’re X percent off of the previous bubble’s highs.

How, for example, did net worth drop 20%, house prices drop 30%, wages drop in real terms, and business credit dry up completely… yet consumer credit remains near/at all-time highs?

It’s a bubble meant to sustain the real “new normal” of just above zero cash flow living for the common American as nearly all income is used to satisfy debts. It is a really twisted world we live in where defaulting on debt impacts the debtor and not the creditor. Risk should be applied both ways in a credit economy, yet insolvent institutions like AIG can continue to get better (no risk) loan terms from the government for their part in the default frenzy while an American citizen with two delinquencies or a few credit inquiries sees their credit pricing worsen up to 50%.

rootless cosmopolitan says:

October 9, 2009 at 12:10 pm

Myron,

“How, for example, did net worth drop 20%, house prices drop 30%, wages drop in real terms, and business credit dry up completely… yet consumer credit remains near/at all-time highs?”

Because a debt load just doesn’t vanish when the inflated book value of the assets like houses that made the debt load in the previous boom possible, crashes or the income available to pay of debt and interest falls. As for businesses. Although business debt has declined somewhat since it peaked (for now), it’s still very high as well. See the Fed’s Flow of Fund Accounts, Table D.3:

http://www.federalreserve.gov/releases/z1/Current/z1.pdf

rc
- Myron says:
  
  October 9, 2009 at 8:53 pm
  
  “Because a debt load just doesn’t vanish when the inflated book value of the assets like houses that made the debt load in the previous boom possible”
  
  Actually, it can and should. If/when businesses and households are actually allowed to fail when they make poor decisions, then that loss in book value or house value results in bankruptcy. Bankruptcy makes debt “vanish” since assets are used to pay the debt.
  
  Instead of the above scenario, we see an environment where the government has subsidized the entire economy in order to save the vicious debt cycle of booms and busts and reduce the number of bankruptcies for insolvent and zombie companies and families.
  
  There’s no true deleveraging for the consumer in a scenario where the government has used public debt growth to maintain the level of private debt.

Ingolf says:

October 9, 2009 at 11:42 am

Edward, I wanted to tag this onto our little part of the conversation, but the “reply” button was missing.

I appreciate you taking the time to reply in some depth. In one sense, the specific matter we’re debating is minor, but at the same time it’s linked into rather important issues so I’ll respond in kind. I suppose at worst we’ll both just end up looking like a pair of obsessives!

Short of getting a more detailed explanation from Rosenberg, to a degree we’re all just picking through the runes. Still, I think the evidence strongly suggests Rosenberg was well aware the $12 billion drop was a seasonally adjusted number.

Consider that quote of his you posted earlier in this thread: “The big story yesterday was the further massive $12 billion decline in outstanding consumer debt in August — the consensus was looking for an $8 billion contraction.” Now, according to Econoday’s Economic Calendar, the consensus expectation was in fact minus $8.5 billion for the seasonally adjusted figure, so this fits.

Also, immediately after the segment you quoted, Rosenberg went on to say: “This held the contraction in non-revolving credit to $2.1 billion – based on what the pace had been running in recent months, it seems as though the total decline outside of the auto subsidy would have exceeded $20 billion during the month.” This also seems to support the view that (absent the C4C) he would have expected consumer credit to have plunged by a great deal more than it did.

As for the larger issue of consumer deleveraging, he said: “Over the past year, consumers have run down their debt by a record $113 billion (and this does not include mortgages). This is an absolutely epic shift in household attitudes towards credit and discretionary spending”.

Doesn’t mean he’s right, of course (although I think he probably is), but all of this taken together does suggest he was neither misled nor careless in his comments.

rootless cosmopolitan says:

October 9, 2009 at 12:24 pm

The more general question not just focused on consumer credit, but on the total private debt load would be, why should we likely expect a significant debt deflation in the system going from here?

A possible answer: Because the total private debt load of 41 trillion US-dollars or 300% of US GDP has become as high that any GDP growth to be reasonably expected won’t be sufficient to generate sufficient income for the net debtors to even pay off the aggregate interest on this debt, so that we will see increasing defaulting on this debt.

rc

Subscribe to: Post Comments