ES -- DX/CL -- isee -- cboe put/call -- specialist/public short ratio -- trinq -- trin -- aaii bull ratio -- abx -- cmbx -- cdx -- vxo p&f -- SPX volatility curve -- VIX:VXO skew -- commodity screen -- cot -- conference board

Wednesday, October 14, 2009


approaching the double dip

pragmatic capitalist links to this letter by hoisington capital, which kicks off with a view of the most important chart in the world -- made additionally interesting by the inclusion of data reaching back to the long depression of 1873-79. i disagree with much of what hoisington has to say on the multiplier of government spending -- while they view fiscal stimulus as pointless, in truth as stated by richard koo within the concept of the balance sheet recession and highlighted by marshall auerback it is the only realistic means of avoiding an overwhelmingly intense deflationary spiral should the private sector set about debt reduction. koo estimates a multiplier of fiscal stimulus in japan during their balance sheet recession between 1990 and 2005 of close to seven -- which is to say ¥2000tn of private sector debt was retired income was created (see comments) with a stimulus of ¥300tn -- and this appears yet better in light of the alternative, which is a collapse of incomes, deposits and wealth to perhaps half their former level.

absent the political will to massive new fiscal stimulus, however, thomas palley in the financial times via mark thoma lights the pathway of deleveraging from this monster debt precipice.

The future is fundamentally uncertain, which always makes prediction a rash enterprise. That said there is a good chance the new consensus is wrong. Instead, there are solid grounds for believing the US economy will experience a second dip followed by extended stagnation that will qualify as the second Great Depression. Some indications to this effect are already rolling in with unexpectedly large US job losses in September and the crash in US automobile sales following the end of the “cash-for-clunkers” programme.

... The economic crisis represents the implosion of the economic paradigm that has ruled US and global growth for the past thirty years. That paradigm was based on consumption fueled by indebtedness and asset price inflation, and it is done.

palley analogizes a two-step credit crisis outlined by mckinsey earlier this year. with perhaps a thousand american banks set to fail as step two unfolds over the next 24 months, this is about the near future.

mckinsey ... is describing a second-wave credit crisis -- one related not to mark-to-market but to mark-to-maturity, which constitutes 90% of the assets of the american banking system -- which would see banks largely unable to earn their way to solvency over the next few years as loan losses accelerate and net interest margin (NIM) deteriorates even in an environment of stable or increasing loan demand. this dynamic will be particularly effective in killing off smaller banks, as they are both concentrated in commercial/industrial loans with exactly the delayed-fuse features mckinsey describes and suffering already from a deteriorating in NIM.

and this presuming stable loan demand. in an environment of contracting loan demand and insufficient stimulus, this figures to be a lethal banking (and economic) environment, with collapsing incomes driving asset prices lower and losses higher -- notwithstanding this morning's earnings announcement from JPM, which is heavily reliant on asset writeups resulting from remarking risk assets which, though still impaired and now wildly overpriced, have benefited from the liquidity-fueled rally since march. notably, via calculated risk, JPM highlighted continuing increases in loan loss reserving as they see weakness in their prime conforming mortgage portfolio -- a trend mckinsey and others obvious expect to continue.

the $64 question is whether sincere private sector deleveraging gets underway in spite of the efforts of the government. could the government blow another bubble? edward harrison has been playing angel's advocate with regard to this by examining unadjusted figures, and appropriately so. but this interesting takeaway from the JPM call underscores what i think will, with so much damage to the valuations of pledged collateral already having been done and likely to continue as valuations return from ponzi finance models to discounted cash flow, become the dominant dynamic:

Analyst: Loans were down about 5% linked quarter 16% year-over-year. Is that supply or demand, what are some of the ins and outs there?

JPM: Consumer portfolios, you have run off portfolios from Washington mutual and in retail, some tightening of underwriting standards in those businesses generally. So expect that at the origination levels, that for a period of time here, we are going to have downward pressure on those balances. We're in the business of making loans against our underwriting standards today. So it is active supply, meeting demand on that score. On the commercial side, you have seen it a little further down this quarter, and that is you know more, it is a little bit of everything but it is more demand clearly because we see extended credit lines utilized at the lowest levels of all time. You can see a swing in those numbers as soon as confidence returns in our commercial clients and they have some use for that money.

they may be waiting a very long time for that swing. a combination of continuing increased bank prudence (it would after all be hard to be less prudential than banks were in 2006) and more importantly net negative demand for loans as households and businesses retrench into balance sheet repair make, in combination with a populist political impulse to control sovereign debt growth, palley's and hoisington's darker view all too plausible. total loans and leases figure to be the predominant face of deleveraging -- until or unless there's another run on wholesale bank funding. retail sales -- via calculated risk -- will be driven in large part by both how (or if) households attack their balance sheet problems and what the government can muster in terms of deficit spending to accommodate their repairs.

recall that peak stimulus is right now. as CR noted yesterday and albert edwards among others have highlighted, all subsequent quarters will see diminishing effects, within a few months yielding to a tightening effect on GDP.

as this effect takes hold, with no further stimulus now in the pipe, with the prospect of accelerating bank failures, accelerating savings, contracting loan demand and continuing tighter lending standards, palley's prospect of a double dip figures to materialize in 2010 with frightening intensity. as was true in japan, this may suddenly forge the political will for greater fiscal stimulus globally -- but i expect only in a much deeper economic crisis can parliamentary will coalesce.

Labels: ,

Thanks for the post, GM. The irony of all this is that the government simply cannot (in the absence of a major war) ramp up spending fast enough to make a major difference even if it wanted to. There is no consensus absent an national emergency which will allow spending to increase quickly--defense spending will increase, but only by about 5%--ditto discretionary non-defense spending in the best case. SS payments are constrained by the lack of increase in the CPI, and Medicare to some extent in the same way. Those are the four basic pieces of the federal budget. The stimulus can increase spending, but slowly (since moneys need to work their way through the approval labyrinth) or else by simply filling holes in the state budgets.

Mind you, the inability of the government to significantly ramp up spending is not necessarily a bad thing, but it is ironic that Washington is incapable of even spending money in a way that will help. The WWI and WWII spending ramps were more significant because they increased from a vanishingly small base (in present terms), and even the small amount of money spent by the New Deal had an impact because it represented a significant increase in percentage terms.

The one body capable of ramping up quickly and almost infinitely is the Federal Reserve--for example, last fall it increased its balance sheet by about $1 trillion in one day. However, the cost is the end of any pretense of central bank independence. Previously, the Fed had only monetized debt in wartime emergencies; now it has played its last trump to bail out bank bondholders. . .

------ ------- ------
whether it can be done quickly or slowly, bwdik, i think it has to be done. the alternatives are bleak.

one of course is a deflationary debt deleveraging. the other would be a successful reinflation that somehow sparks some combination of households and businesses to borrow madly once again.

there's a third avenue -- currency devaluation -- but (with apologies to eric janszen, whose views i'm reading up on) i think this something of a phantom. it's easy to forget that the US engineered a severe devaluation in 1933 by exiting the gold standard and nevertheless experienced a balance sheet recession that lasted until 1954. and hoisington has this much right in my opinion:

The inflation outlook from the monetary and fiscal standpoint looks truly deflationary, yet some believe that dollar weakness will reverse this circumstance and create inflation. This is unlikely. First, our imports are about 13% of GDP, and even if the dollar were to halve in value, the price of imported goods would not only have to compete with U.S. producers, but also their price adjustment would have to offset the other 87% of factors included in the pricing indices. Second, unlike the 1930's a 50% decline in the dollar would be difficult to engineer. Fisher recommended to Roosevelt that the U.S. should exit the gold standard, which he did in April of 1933. That was a fixed exchange rate system, and within three months the dollar lost more than 30% against the gold block countries and fell to 60% of its former value within the next five months. This spurred our exports and provided some price inflation (2.9% per year, GDP deflator) for the next four years. Then, in 1937 the tax increases (the next policy mistake) reversed the positive growth rate of the economy and drove price levels and economic activity downward again. However, even with that small period of price increases the overall price level never recovered from the 25% decline that occurred from 1929 to 1933, and thus deflation reigned. Today the declining dollar is a good thing in terms of our trade balance, but the modest change will be insufficient to offset the negative forces of insufficient domestic demand.

moreover, the yen strengthened remarkably, from 144 in january 1990 to a high of 83 in 1995 -- and yet japanese deflation remained controllable and their GDP did not collapse. that too would seem to indicate that the more important determinant by far is fiscal policy.

------ ------- ------
the economist also notes the trend.

------ ------- ------
Hi gm,

"koo estimates a multiplier of fiscal stimulus in japan during their balance sheet recession between 1990 and 2005 of close to seven -- which is to say ¥2000tn of private sector debt was retired with a stimulus of ¥300tn"

What is your source for this? I don't remember it in either the October or March CSIS talk. Is it in Koo's book?

I recently dug up Japanese debt data including the private sector and graphed it -- see this post. My derivation of the debt data seems sharply at odds with Koo's statement above (see especially this graph of nominal debt by sector which though unlabeled is in billions of yen).

By this data, private sector debt was well under 2000 trillion yen in 1990 (so couldn't have shrunk that much in any scenario), and really didn't contract much through 2005 since it kept growing until 1997 in nominal terms (though not as a ratio to GDP) before shrinking. Yet the public sector debt expanded significantly during the same period. I don't contest Koo's claim that fiscal policy prevented a depression, but I am puzzled by the debt data here.

Any suggestions on how to reconcile these facts? Even looking at total balance sheet liabilities in Japan (not just those categorized as debt), the official Japan Cabinet Office data shows them expanding from 4405 trillion yen on 1990 to 5890 trillion yen in 2005, so that again is inconsistent with an overall contraction. jck of alea blog kindly looked at my sourcing and seemed to think it valid, plus it matches some published data points I've found elsewhere. Hmmmmm.

------ ------- ------
i thought someone would ask hbl! just after 28 minutes in the march talk, though i expect you could also do the math out of the data in his books (i haven't).

i doubt koo would disagree with you re: private sector debt -- his exhibit 13 slide shows debt growth halting in 1990 and earnest repayment (>4% GDP) kicking off only in 1997 following the only true credit crunch of the delevering.

i expect his 2000tn yen figure also includes 1500tn in capital losses in land and shares (ex 12). elsewhere in the talk he points out that this actually underestimates the losses by exclusing other assets, such as golf club memberships. in ex 26 he shows the change in japanese systemic bank balance sheets from 1999 to 2007 as being down about 100tn yen.

ex 13 also shows how he arrives at the 300tn yen fiscal stimulus debt figure.

------ ------- ------
sorry -- in ex 26 he shows the change in japanese systemic bank balance sheets accounted as lending to the private sector from 1999 to 2007 as being down about 100tn yen.

------ ------- ------
gm, thanks so much for your very helpful thoughts on this!

In terms of amount of government stimulus, the cabinet office data show a 640tn increase in government debt... but now that I look at the raw numbers again I see that the change in net worth of the government was closer to 300tn yen, i.e., the government also expanded its assets. So looked at in that way, the "cost" half seems to agree with Koo's claim.

I looked at the data and still couldn't figure out where the 2000tn reduction in private debt would come from (or why you would add in the loss of asset value to that). Then I managed to watch the part of the video again that you mentioned (though the time seek doesn't work right and I can't get to it again!) and I think he said what was avoided was a potential 2000tn yen loss in GDP (not a reduction in private debt). Which is baffling given that GDP was only 442tn yen in 1990. Perhaps he meant a potential cumulative 2000tn yen GDP loss over 15 years... I bet that's it.

------ ------- ------
hbl -- of course you're right. in calling it the most successful fiscal stimulus in human history, he notes that spending 300tn yen prevented a cumulative loss of GDP of 2000tn yen -- and then says that the multiplier is therefore about seven, and not one or less than one.

this is a different statement of the fiscal multiplier than i've usually heard, but it's no less effective -- koo is essentially saying, is he not, that 300tn in govt spending translated into something like 2000tn in income that would otherwise very likely not have existed. koo derides much of the criticism of japan's fiscal stimulus which assumes (consequent of equilibrium modeling of this kind) that had the govt done nothing there would have been no change to GDP -- this makes it seem that fiscal stimulus produced only the difference between no growth and very slow growth, whereas in reality is produced the difference between very slow growth and total collapse.

but of course that extra income was not exclusively directed to debt retirement -- it seems very much that the govt spending replaced the portion of the income stream which was redirected from consumption to debt retirement on a one-to-one basis. so, for example, if businesses directed 6% of GDP to debt retirement and households further saved 4% of GDP more than before, the shortfall of 10% of GDP is what the govt had to borrow beyond its initial condition deficit to maintain GDP. ex 16 indicates lending to corporate sector declined from 400tn to 250tn, a reduction of 150tn, and in so doing declined from 85% GDP in 1990 to 55% in 2005.

------ ------- ------

Yeah, I basically agree with you on what Koo is saying (thanks for elaborating). However I'm unclear on where how he derives the 2000tn yen estimate and how subject to debate that figure is. In the audio I got the sense maybe it was something basic like assuming a return to 1985 GDP levels (but then what level of growth?) but I could be wrong and it wasn't obvious to me which slide he referred people to.

So whatever the multiplier really was (since we don't know what would have happened under a different policy response), I do think your last paragraph is basically right on with respect to fiscal deficit spending needing to offset the increased private sector desire to save (where saving includes debt repayment) in order to prevent a contraction in GDP (and thus the desired savings level being thwarted). This is one of the core tenets of Modern Monetary Theory (Chartalism) as I understand it, though I still have more reading to do to finish assessing it. You might be interested in this Bill Mitchell post that mostly agrees with Koo's policy approach but comes from a different understanding of the monetary mechanics.

------ ------- ------
Great comments, GM and HBL. I have always assumed that Mr. Koo's numbers were a little nebulous and stretched to suit his case. But I have a bigger bone to pick with his argument. One of his unspoken assumptions was that because massive government borrowing and ZIRP "worked" (i.e., prevented a disastrous deflation) in Japan in the last fifteen years, that the same strategy will "work" in the US now.

I have to think that the yen carry trade from 1995 (the advent of ZIRP) had some effect on world asset prices through the increased leverage it allowed. A second major economy going to ZIRP will have a much smaller effect (those who wanted to borrow at 0% have for the most part already done so. This assumes that the world's major financial players were able to borrow at near 0% in Japan.) But the rise in worldwide asset prices surely led to the continued increasing level of Japanese exports, which no doubt helped reduce the private debt levels. Absent this boost, would things have been the same?

I don't know if this theory is relevant, but it is only one example of how results could be different the second time around. Once people have been burned once or twice with the excess leverage that "free" money encourages they generally learn to leave it alone.

------ ------- ------
Macro-man leans toward inflation by highlighting the argument that the "domestic" output gap may not be as large as believed.

------ ------- ------
bwdik, your point -- we can't export our way out in a global contraction -- is well taken. of further particular distress to me in considering koo's argument is the primary difference between japan and the US -- which is the nature of systemic funding.

as john hempton really drove home to me, japan's banks were always surfeit with cash -- never did loans exceed deposits, and excess cash was always a feature.

the US is not that way; a large chunk of american loans are backed by wholesale funding based on borrowed foreign deposits. as a result we're highly unlikely to see ZIRP beyond policy rates -- we must compete for global funding as japan never did, and will suffer interest rate volatility as a result. more likely than not (presuming the fed fears and will not engage in a hyperinflation) that loans decline over time to better fit the deposit funding base as our current account imbalance normalizes.

stubbornly high interest rates limiting borrowing capacity, general decline in bank credit outstanding, permanently smaller securitization complex, working down wholesale funding hopefully without a run -- not a pretty picture, and not nearly so clean as was japan's.

------ ------- ------
That was an interesting link to the Janszen presentation in your Twitter feed. BTW, the Mitchell link was by hbl above.

------ ------- ------
Einhorn is betting on a currency crisis up to four or five years out.

------ ------- ------
rb -- here's the speech in toto. i'm still waiting for reinhart/rogoff to come to fruition, which is to say something like 50% of sovereigns will default.

------ ------- ------
hbl -- sorry for the twitter slight!

------ ------- ------
check gregor's october 5 postings re: argentina -- interesting to note the powerful deflation that held for a year before the default and hyperinflation. anyone remember this? i doubt it.

------ ------- ------

Post a Comment

Hide comments

This page is powered by Blogger. Isn't yours?