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

Friday, October 30, 2009


japanese LEI and the nikkei

as albert edwards is drawing parallels to japanese fits of frantic leading indicator expansion followed by disaster selloffs, it seems appropriate to highlight the history of the japanese LEI as maintained by the conference board through the period 1990-2005, which contains virtually the entire japanese private sector balance sheet recession.

i hope you forgive the crude overlay -- i'm sure professional web designers cringe at such things. the overlay of the nikkei has been stretched to fit exactly the same timescale, so comparison is easy.

the first thing to deduce is that the leading indicators will lag the market. this probably isn't news, as the most forward indicator in the LEI itself is the equity market. edwards may be right to the point that, once a downturn in LEI began, it often telegraphed the most severe contractions in price. but in truth it seems on this evidence that, if you wanted to preserve capital most effectively, any significant moderation in the pace of strengthening was a better signal to find the storm cellar.

the second is that the equity market in 1993, again in 1995, and again in 2002-3 was hitting new lows even as the LEI was contracting only mildly or even rising and as the economy stayed out of recession. equity is after all the ass end of the capital structure, and should in a low-growth balance sheet recession be savaged as the systemic balance sheet capacity to warehouse equity contracts severely.

third, as has been noted in reviewing richard koo's data as given in exhibit 13 that accompanied this talk on the balance of financial deleveraging in japan (and which are also reproduced in his book), the character of japan's balance sheet recession changed over time. the 1990 bust saw corporate borrowing slow to nil by 1994 and then stay near the zero line through 1996. this might be characterized as a 'first phase' in which japan's banks with the blessing of the government engaged first in pretend-and-extend lending. the 'second phase' kicked off with the advent of the only true bank liquidity crisis of the entire period, which corresponded closely with the hashimoto fiscal rebalancing of 1997. as the government attempted to withdraw spending support from the economy and improve government finances, the economy dumped into a deflationary tailspin which forced radical deterioration of credit quality. following a series of massive liquidity injections which resolved the banks' liquidity issues, profitability returned as the economy emerged from this second recession in early 1999. as can be seen on exhibit 13, however, profits were directed by the corporate sector directly into debt repayment -- and this did not materially abate until 2005, when (as can be seen in exhibit 11) japanese corporate debt fell to a 32-year low of approximately 40% of GDP.

examining the LEI-nikkei chart, it is possible to discern a change of phase around 1996 as well. (further, i suspect but cannot show one would find similar action following the distressing 1937 fiscal rebalancing attempted by the roosevelt administration.) though economic activity boosted by renewed efforts at government deficit spending thereafter improved, the prices in asset markets diverged from LEI and continued on their downward path throughout the second phase. this is i think the kind of repudiation of the cult of markets and the asset-based economy on which recently bill gross and many others have mused. japan had at long last by 2005 returned to a asset valuation regime dominated by discounted cash flows, and not leverage seeking capital gains. so i expect will the united states in time.

in any case, that change of regime still lies in the future. while material evidence of contracting credit outstanding is plentiful, a lesson of the japanese example may be that credit outstanding may merely stagnate for some time pending future developments for so long as government spending maintains employment and income levels. edward harrison has made tentative suggestions in this direction, and does again today with the PCE release. but it must also be said that the nature of the american credit bubble -- being primarily household and not corporate as was the case in japan -- i think diminishes the capacity of the banks to conspire on the kind of pretend-and-extend schemes that are easier to keep quiet at the corporate level of discussion. particularly considering the potential of household strategic default on mortgage debt, which is an option not open to companies wishing to avoid bankruptcy, individuals can free a significant amount of cash flow by simply walking away and renting for a while without terrific consequences. moreover, household cash flow may be quite a bit less reliable given the potential for growing unemployment and strategic (or simply unavoidable) wage reductions on the part of businesses. such are concerns evidenced in household surveys, as relayed by ft alphaville. should default play a larger role in the resolution of the american credit bubble, one might expect credit outstanding to households to deteriorate quite a bit faster than was the case in japan. and that would advance our transition to the 'second phase' markedly.

Labels: , ,

Good post, gm. And with respect to your last paragraph, I've had very similar thoughts regarding how a debt bubble dominated by household debt might play out differently than one dominated by corporate debt.

"2005, when (as can be seen in exhibit 11) japanese corporate debt fell to a 32-year low of approximately 40% of GDP"

I don't see this on exhibit 11. Do you mean exhibit 16? If so the credit to GDP scale is on the right and shows corporate credit closer to 52% in 2005. Also note that this is bank credit only, so it excludes non-bank forms of debt (bonds, etc). Which, while it doesn't change any of your broader points, makes your statement regarding "japanese corporate debt" mildly misleading (admittedly Koo's source material is misleading here too). As I mentioned before, I think Japan's total corporate debt (loans and debt securities) dropped from 150% of GDP to 100% of GDP from 1990 to 2005. (See the graph on this post). The deleveraging trend is clearly the same in both cases, I just thought I should point out the discrepancies.

------ ------- ------
you're right hbl -- i'm drawing the 40% figure out of memory and it isn't evident on the charts even if the delevering is. moreover we are talking bank credit and ignoring capital market debt, though iirc japanese firms are closer to the euro model of bank borrowing than the american one of corporate bond issuance when it comes to debt finance.

as a thought -- have you ever tried to contact koo with your data, hbl?

------ ------- ------
The data I found says you are correct regarding bank financing being dominant -- it shows bank loans to non-financial corporate sector at 84% of GDP in 2005, versus debt at a total of 102% of GDP when you add in bonds and other non-share securities.

It hadn't occurred to me to try to contact Koo... senior as he is at Nomura and being the expert on this stuff I assume he's tough to reach and would be aware of the source of this data already, since it's just official Japanese government stats pages... It does differ a little from the data he presents, but there is probably some good explanation.

The main surprise for me was just how much extra government debt was added relative to the reduction in private debt... which doesn't reinforce his verbal selling points very well on how to navigate through balance sheet recessions by replacing private debt with public debt, then reversing them later. So he may not be that interested!

I was more curious as to the take of some of the chartalist thought leaders (such as Bill Mitchell), but despite fishing for comments twice on his site, I got no response. Oh well, I think I know what they would say anyway (the data doesn't particularly contradict any of their arguments as best I can tell).

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

Post a Comment

Hide comments

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