Thursday, July 8, 2010

How Long and How Fast Will We Deleverage?

The other day, I posted this graph of the various components of US debt:

Obviously, deleveraging in the household and financial sector has begun. The question is, how long will it go on, and how low must the various debt-to-gdp ratios go?  In this post, I look at various ways of guesstimating this.

One approach is to assume that the level at some point in the past was "normal" and figure roughly how long it will take to get back to that at the current rate of deleveraging.  Eg, maybe the household debt to GDP  ratio of 60% in 1990 was ok.  Since in the last three years we are doing about 1.3% per year of household deleveraging (percentage points of GDP), it will take about 25 years to get there.  Clearly this is extremely imprecise - if you thought we only had to get back to 2000 levels, it would take less time - a little over 15 years - whereas if you thought you had to get back to 1975 levels, it would take even longer.   And we don't have any basis for picking what is an appropriate level that the economy is going to decide it needs to get back to. Also, the pace of deleveraging might change in the future.

Another approach is to look at other historical deleveraging episodes.  The most famous is of course the Great Depression, where we have this graph (from John Mauldin) of total US credit market debt to GDP:

You can see that the main deleveraging episode went from 1929 to 1942 (13 years), and total debt fell from 300% of GDP to about 160%.  So deleveraging was happening at about 10% of GDP/year.  Presumably, that was too fast, since output fell massively, unemployment went through the roof, and misery spread all over.

If you compare the current episode to date, on the same basis of total credit market debt, it looks like this:

(There's a small discrepancy between the absolute level of Mauldin's figures and mine - computed from the Federal Reserve Flow of Funds data and BEA GDP - that I don't know the reason for.  Could be due to revisions in the series).

On a total credit market basis, we are deleveraging at only a few percent per year, with increases in government debt partially offsetting the private sector deleveraging.  Arguably, it's much too soon to tell how fast it's really going to go, however.

Another case of interest, is Japan post 1990, where I found this graph (at this interesting post):

On a total basis, Japan has so far deleveraged from 345% of GDP in 1998 to 283% at the end of 2008.  The process is presumably not over, and the rate is about 6%/year.  So slower than the US in the great depression.  And Japan has had stagnant growth, but some growth (of course, Japan was running a trade surplus the entire time, which is not likely to be the US condition).  The private sector in Japan was deleveraging at about 13%/year, but half of this was offset by a dramatic expansion of the government's liabilities.  Japanese private debt is now less than half of what it was at the peak.

A similar picture for recent decades in the US would look like this:

So the US entered the crisis with similar ratios to Japan.

An important difference, however, is this time it's global.  This interesting McKinsey report shows debt to GDP for a number of important economies:

So all the advanced economies need to deleverage at the same time...

Finally, yesterday morning, long time reader and friend Southsider1 stopped by my blogging cafe for breakfast, and we hashed out another method of guesstimating how long deleveraging needs to run.  The idea is this: deleveraging will be required, and the economy will not be normal, until there are no more than a few percent of US households in negative equity.  In a normal healthy housing market, people are required to put down 10-20% on their house, and house prices are flat or rising in nominal terms, so few households have negative equity.  Mortgage debt is the bulk of household debt.  Also, as long as households are deleveraging, the asset side of bank's balance sheets will be shrinking, and so they will need to shrink their liabilities also.  (And to the extent the government requires them to maintain higher capital ratios, this will be so a-fortiori).  Thus financial sector deleveraging and household sector deleveraging will be coupled.

I found this research report from CoreLogic which estimates when negative equity will be eliminated in a sample of US geographical areas (you have to give them your contact info to get the report).  Their core scenario makes the following assumptions over coming decades: 1) inflation is about 2% annually, 2) real house prices rise by 1% annually, and 3) the main mechanism of eliminating negative equity is people paying off their mortgages while house prices slowly rise.

I have a number of problems with this: right now, core inflation is below 2% and dropping, I think it's implausible that houses will appreciate in real terms over the coming decade, and I think a lot of negative equity situations will be resolved by foreclosure, short-sale, etc, which they appear not to be modeling.  That said, the core scenario produces this:

Ie the negative equity is mostly gone after about ten years.

They also briefly mention an alternative scenario with a 1.5% nominal appreciation in house prices which would make it 2017 rather than 2015 when the first markets elimate their negative equity.  Presumably, then, the overall process would spin out to about 2025.

All of these methods have problems, and you can pick your poison.  What they all have in common though is this: deleveraging will take at least a decade, and it could easily take 15-20 years.  In the meantime, the economy is likely to be pretty choppy at best.


Burk said...

That is very interesting.. but seems to mainly pertain to the housing finance world.

It would be nice if policy makers could otherwise keep employment, income, and inflation at positive settings so that this process (a real estate slump) isn't more painful than it has to be, insulating main street as far as possible.

As you illustrate government debt isn't the issue at all, and indeed the savings vehicle it provides is what is soaking up much of the deleveraged resources.

rks said...

The stuff that is produced is consumed more or less immediately [particularly if you regard housing as infrastructure that produces accomoadtion, etc]. Partly it is consumed by people who earned money in the past, and partly by people who plan to earn money in the future. Over the long term these must be equal?? I suspect that there is a lot of money out there whose owners have no intention of ever using it to consume. This has to be a huge distortion. I'd love to understand this stuff better.

Adam Schuetzler said...

To say that "a lot of negative equity situations will be resolved by foreclosure, short-sale, etc." is almost to understate the case. In worst hit areas like my area, foreclosures and short-sales make up 50% of residential real estate sales. That's on top of very high inventory during a recession. Prices continue to fall. It makes sense, you would be insane to pay a mortgage on a house that is suddenly worth 1/2 as much as what you paid - the only one profiting is the bank. That would be like debt slavery.

The only thing holding the house of cards together is government intervention. In real terms, however, I'm not sure what it means. I know that houses go empty and grow nasty lawns, I know that people stop moving into distant suburbs. But other than that, the money ain't coming back. The government can do what Japan did and cover up this fact by taking on debt, but it won't save the economy. Japan's economy is still not all that great.

I'm not sure if there is a good solution. Looking at the history of the US debt ratio, it seems that while deleveraging itself is painful, it creates the conditions for massive growth (like in the 1950s and 1920s). Perhaps we are just deleveraging in the wrong way? Maybe we should just call a jubilee year, get rid of all debts, and start over? Debt is not a real thing, it's a social construct. We just don't like to remember this fact.

The closest to a sensible solution I've ever seen is Salman Khan's idea of just getting rid of the banks and using the stimulus money to set up new ones with no liabilities (it's on the khan academy website). Basically, rather than supporting a dead financial system, create a new one from scratch. Here it is on youtube:

permabear said...

I find the comparison between the U.S. in the 30s and Japan since 1990 interesting. The author suggests that the Japanese experience is far superior to the American experience because obviously Japan has experienced growth while the U.S. experienced a Great Depression. But with Japan's public debt still spiraling higher, I'm not sure the end game for Japan will be more positive than the end game experienced by the U.S. Out of the rapid deleveraging (and of course WWII) the U.S. became a lean and mean economy that thrived for decades following the Great Depression. By contrast after 20 years Japan remains mired in endless malaise with increasing concerns about how their public debt will be dealt with. My conclusion is thus perhaps a rapid deleveraging, even if it means an economic depression, is superior to an endlessly slow deleveraging with increasing public debt, for the longterm health of an economy. Extrapolating it to today, all that intervention we have experienced may only be delaying the deleveraging that is needed and may only be putting off even worse pain for a future day. In short, maybe we need an occasional depression to cleanse the economy of all the crap, namely debt.

Unknown said...


I'm wondering what you or anyone here thinks of the perspective over at the financial blog The Automatic Earth ( ). Are you familiar with that site? I believe they think that the deleveraging will be much vaster and faster than the possibilities you describe here in your post. While I can not summarize their reasoning (the whole topic being way over my head), I can say a that they point to the huge derivatives market which, even if only 3% of it unwound would be enormous. They say positive feedback loops will get going. They talk about there being multiple paper claims on real underlying assets. They talk about herd psychology. They have several primers on their page.

It is a dire vision. I don't know enough about high finance to know if it is too extreme.

Thanks in advance for any thoughts!

rks said...

The Automatic Earth and Mish are fun to read because they collect all the doom and gloom in one place. TAE used to say "The government won't print money and there'll be massive deflation". When the GFC hit and the government (the FED in America) did print $1.5 Trillion, then the line changed to "they won't print enough money to prevent deflation". Actually I think we know that Bernanke is very keen to print enough to prevent deflation, and the question is whether Congress will be able to stop him. Remember that there is a real economy with people making and consuming. The financial sector is a servant to that real economy. The feedback loops that TAE worries about will only have traction when there is a shortage of liquidity.

Stuart Staniford said...


Yes, I'm somewhat familiar with The Automatic Earth, since Stoneleigh and Ilargi were "The Oil Drum: Canada" during the period I was blogging at TOD. They spun off on their own after a while. I have had debates with Stoneleigh in the past, though I haven't read their site regularly any time recently.

My general sense is that they are biassed over-pessimistic. I don't doubt that we are going to go through some difficult times as a result of the ongoing problems in the financial system, but they tend to think it's going to be the end of civilization, and I don't think that's very likely. I think they have tended to underestimate the degree to which the political system can respond, and will respond once there is a sufficiently severe crisis to motivate action. (In particular, they severely over-estimated the consequences of the 2008 crisis).

On the huge quantities of derivatives specifically - I certainly don't understand the derivative world well enough to say everything is ok with it, and maybe nobody does. But let's try gaming out the worst case - suppose something goes very wrong in derivative-land such that a bunch of critical firms are about to fold, and the derivative situation is so complex that no-one can untangle the mess in a timely way. So in that scenario, what might governments do with their backs to the wall? Well, how about declare derivative contracts null-and-void en-masse. Then any critical financial firms that are at risk as a result, the government takes them over, identifies the vanilla consumer and business loan units, and makes sure those are recapitalized and continue to operate.

Hell of a mess, for sure. Stock-market through the floor while everyone figured out the winners and losers - sure. But the end of the ability to grow wheat, mine iron, or deliver bread to supermarkets? I very much doubt it.

Still, this kind of possibility is the sort of thing that I have in mind when I say I don't want to see the federal government lever up to Japanese levels. I think we should keep some balance sheet in reserve for the unknown unknowns.

permabear said...

I'm personally not familiar with Automatic Earth but I'll check it out after posting here. My username indicates my bias. I believe we are in for some very hard times for all the reasons the author of the article above indicates. I just don't see an easy way out of this debt problem short of a nasty deleveraging. I think a best case scenario is an endless Japanese malaise. I personally don't subscribe to the end of the world camp either. I see a real possibility of either a Great Depression scenario or an extremely high inflation scenario. There are many developments that could produce such a scenario.

The only reason the U.S. economy is growing today is a combination of a lot of government intervention and historically low interest rates. If the U.S. chooses to take the Paul Krugman, endless stimulus, approach, there is always the possibility that the world will wake up one day to a loss of confidence in the dollar, a spike in interest rates, and a complete change in the economy in a very short period of time. In such a scenario the Fed will be backed into a corner, having to choose whether to monetize the debt on a massive scale. This is the scenario that could lead to the very high if not outright hyperinflation.

Japan is different from the U.S. in that they are able to fund their public debt almost exclusively domestically. The U.S. depends on foreigners, especially nations like China, to purchase massive amounts of debt. If the Chinese and others for whatever reason, decided they no longer wanted to play this game, or they all of a sudden decided to let their currency really rise or even float, it could be the end of the subsidization of the American economy.

My view is that we are in a massive ponzi scheme. The trade and budget imbalances are unsustainable. I think that something has to give at some point. I don't feel we can keep on running up these massive deficits without a major consequence. And it could be one that appears overnight.

aangel said...

Thanks for this useful post, Stuart.

The deleveraging will go through giant spurts, in my view, each time oil increases in price as we follow the downslope of oil production. A giant spurt happened in 2007/08 and it was so severe that the only thing that got the pump going again was the central banks acting concertedly to prime it again. This may work one or two more times but are you willing to bet many more times than that?

Deleveraging in general has to happen because as oil is removed from the system the system cannot do the work that would create the profits to pay back debt. So, in this case, TAE may be incorrect with their timing but the endgame is absolutely already determined.

Using Greer's model, we are moving from Abundance Industrialism to Scarcity Industrialism. Thereafter comes the Salvage Society in which it is less expensive in energy terms to reuse steel, for example, from skyscrapers rather than make it from iron ore. We are some decades away from that but I doubt it will occur beyond 2100. We are seeing glimpses of the Salvage Society now as people steal copper pipes and aluminum window frames from empty homes. This will increase and anything not nailed down will be stolen for hard cash as unemployment continues to rise.

There is more to this line of thinking but the important point to note is that Scarcity Industrialism is a whole new ballgame, economically and politically. Where anyone could purchase a barrel of oil if one had the money, we are moving into a situation where if I purchase one barrel extra you must purchase one less. That is a recipe for geopolitical tension if there ever were one.

This switch in how we operate on the planet has far-reaching impacts, one of which is the forced and faster develeraging process that I foresee. The rate of the Great Depressions seems about right to me as a lower bound.

Batten down the hatches.


rks said...

"...the system cannot do the work that would create the profits to pay back debt."
The assumption here is that the people with money won't lend it to the people who want to do useful things because they have a better option. It is the crucial role of government to ensure that they don't have a better option. In particular the government must ensure that the inflation rate is 3% higher than the decline rate. Then the nominal economy will chug along as we all get poorer in various ways. It is wrong to think that there is no way to run a declining economy.
The other question that doom folk ignore is the possibility of running economies on cheap electricity, which can only be Nuclear. The discussion on this is happening now at "The Oil Drum" and at "Next Big Future". But even if that works there will be a grim decade or two of transition. That's what I expect.

Alexander Ac said...

Hi Stuart,

thanks, very interesting and I posted it on TAE :-)

Now, Ilargi and Stoneleigh (and Gail Tverberg too) would say that you are too optimistic in general and I would say that you are too optimistic about the rate of climate change, see this huge overnight collapse of Jacobshavn Isbrae on Greenland (h/t DD):-)

Regarding the pre and post war deleveraging - Paul Krugman argues that that debt was NEVER paid down, only that GDP (or GDP/Debt) increased, thats why he supports increased spending, not austerity.

If one is deleveraging (irrespective of the speed), would not we have deflation, thus lower oil prices, and thus lower oil production and thus peak oil?

Ilargi and Stoneleigh would certainly argue about your comment that they over-estimated the 2008 crises consequences, bacause basically they claim that the crisis is NOT YET over (and was never "solved"), and that we are still in the midst of the biggest bubble ever created (yes, with the help of derivatives and fossil fuels...)

Thanks aangel for the comment - I too think that the endgame is already determined... lets see.

Al said...

I agree deleveraging seems to be mainly undertaken by default and write offs of debt rather than consumer repayment. Something everyone seems to want to take advantage of.

Another way to look at this issue is that the Fed has engineered a very steep yield curve that is extremely favourable for bank profits. They are looking to the banks to accelerate this debt forgiveness process rather than paying it out to themselves.

The losers are obviously savers and investors who are suffering low rates.

But what I wanted to say was that an alternative method of assessing when the deleveraging is over, is to look at the rate at which the banks have the capacity to write off debts using their excess profits.

Alexander Ac said...

Hi Stuart, here is Stoneleigh note on Japanese deleveraging:

"As for Japan being an example of 'sustainable deleveraging', I would disagree. They simply had a lot of money to burn though, and a functioning export market to lean on, so that they were able to put off the day of reckoning for years. The rest of us will not be so lucky. We have an enormous debt hole and global trade will be hit very hard as credit dries up.

Deleveraging does not play out as a slow squeeze. It happens in rapid spurts, with intervening rallies. I think we will see an enormous amount of deleveraging ver the next few years, and the world will be almost unrecognizable by the end of it."

rks said...

I'm not an economist and nobody should take any notice of me, and I'm sure they don't. Still it is helpful to post to clarify my own thoughts.
The deleveraging story is portrayed as one way. In fact it is a 2-way thing. Things work well if there is a balance between people lending, who intend to consume later and people borrowing who intend to produce stuff later. The problem we are in come from people who earned money and have no intention to do matching consumption later. This started with the oil producing countries in the 70s, and we're still trying to invent ways to recycle petrodollars. Then came China. Deleveraging and delending can proceed in 2 ways (with points in between). The borrowers can end up with nothing, with their assets ending up at bargain prices in the hands of the lenders. That's what the rich want, and their proxies the world's right-wing parties. However we should feel no sympathy for the lenders who had no intention of consuming later. The alternative way to reduce the debt in real terms is to inflate it all away. This is what the left should want, even though it will also benefit many leveraged rich people. To give a flavour of how right-wing style deleveraging will play out: yesterday we saw Neil Ferguson on CNBC saying that America should reduce its debts by selling off the farm, mentioning the New Jersey Turnpike. As long as the lenders keep insisting on lending, America could do that till its all gone.

Basil said...

Very interesting post, but I believe you miss one vital point: There are two ways to decrease debt as a % of GDP. One is to deleverage (decrease your ratio of debt to equity); the second - which I believe you miss - is to increase GDP. Increasing GDP is increasing the denominator, which decreases debt as a % of GDP.

Unless that's taken into account when you say the US is deleveraging by a few % points each year?

suicide kitten said...

I stopped reading TAE when I realized Ilargi and Stoneleigh are not interested in checking themselves against the unfolding reality. Both writers have made predictions that so far haven't materialised (China faces hard landing in 2009 (Illargi); Obama will be the shortest-term-ever US president (Illargi); stock markets to start crashing again as of October 2009 (stoneleigh)). Now getting the future wrong in itself is fine; no one is a prophet. But from a serious person I would expect to go back to their predictions, and explain why it didn't happen this way. I never saw any of it on TAE, just a general warning "it's still coming....".

Not to mention the fact that whenever challenged by someone who could make good arguments, they either ignored the challenge or banned the challenger. Again, it's their blog, but I'm far less inclined to listen to their message.

In substance terms, I have also serious problems with Stoneleigh's analyisis, which I believe is far too mechanistic (and based on zoological understanding of humanity); draws too much on the parallels to the Great Depression, without acknowledging the differences and the much more powerfull policy tools available today to decision makers.