Japan: Capacity Constraints vs Expectations

It has been a grim week for Japanese industrial data: industrial output fell 2.3% mom in February, and the fall in the March JMMA/Markit manufacturing to 53.9, the most modest expansion since Sept 2013 hardly suggests a rapid rebound. Trade data for the first 10 days of March tells the same story (with exports up just 3.1% yoy in yen terms). These can perhaps be dismissed as showing transient weakness, possibly anticipating a collapse in demand because of the the 3pp rise in consumption taxes which took place this week.

The point of BOJ's quarterly Tankan surveys is to capture slightly longer-term trends. But the 1Q Tankan was also disappointing, with the Large Manufacturer's outlook index falling to just 8, which is the weakest for a year and which was a quite dramatic fall from the +17 recorded for current conditions. Even worse, the all-industry capex forecasts collapsed to 0.1% yoy for the coming year.

However, hidden away in the details was more encouraging news which raises the economy's chances of surfing through the cross-currents caused by the tax rises. Specifically:

• the diffusion indexes tracking supply-side constraints continue to suggest the potential for positive cyclical pressures. Large companies' readings of excessive employment minus insufficient employment came in at minus 6 (+2 for manufacturers, minus 12 for non-manufacturers); and this is forecast to improve over the next three months to minus 4 (+4 for manufacturers, minus 12 for non-manufacturers). 
• the diffusion index for excessive production capacity minus insufficient capacity came in at +2 (+6 for manufacturers, minus 3 for non-manufacturers), and this also is forecast to tighten over the next three months to +1 (+6 for manufacturers, and minus 3 for non-manufacturers). 

It's worth putting those in context: since 1995 the tightest reading for the employment diffusion index was minus 13 at the end of 2007, whilst the average has been +10.7. For production capacity, the best reading since 1995 was minus 2 – again in 2007 – whilst the average has been +9. So among the messages of the 1Q Tankan was this: Japan still has a (very slight) labour shortage, and no longer has a significant production capacity overhang.

 

Now this assessment is strikingly similar to the views underlying Bank of Japan's calculation of Japan's output gap (the difference between actual GDP and potential GDP).  BOJ's view that Japan is likely to close its deflationary output gap in the short to medium term looks plausible: quite possibly by the end of this year, Japan's domestic economy will supply insufficient  goods and services to meet domestic demand. (Incidentally,  the emergence of a sizeable current account deficit also suggests a domestic economy which is becoming cyclically, if not yet structurally, undersupplied.) If so, this would be the first time since before the financial crisis, and potentially for only the second sustained period since 1990. 

I have stolen the chart below from Bank of Japan's Outlook report published in October 2013, partly because it so closely echoes the Tankan findings, but mainly because it shows how distant is the memory of Japan being supply-constrained. 

In practice, expectations entrenched over such a long time are unlikely to be overturned simply because the facts change. The process of changing expectations sufficiently to alter economic, commercial and financial behaviour is likely to be long and frustrating – the initial 'shock-value' of Abenomics was prized at first specifically on the hope that it might short-circuit this process of expectation-rebuilding.   That 'shock-value'  might not have dissipated entirely without result: that, perhaps, was the  underlying message from March's Shoko Chukin SME confidence index hitting the highest point since 1989.  

Nevertheless, as a guide to what we might expect, it seems realistic to take as our baseline what happened in that period between roughly 2006 and mid-2008 when large companies thought they had insufficient production capacity, and unsufficient labour. What happened to: company capital spending; labour markets; and inflation? 

Capital Spending Using the MOF's quarterly survey of private sector balance sheets & p&ls as our guide, there was a vigorous response to the threat and then emergence of a shortage of capital stock during this period.  Following years of negligible growth or actual shrinkage in spending on plant and equipment, as supply constraints began to emerge, spending rose 9.6% in 2004, 8.5% in 2005 and  14.7% in 2006 before being choked off in 2007 and subsequently.  It hasn't recovered: corporate spending on plant and equipment last year was a mere 68% of what it had been in 2007!  

On the basis of the tightening supply-side constraint on productive capacity, it seems reasonable to expect that the dire 0.1% yoy forecast of capex spending will be revised sharply higher in the coming quarters. 

Labour Markets This strength did not carry through noticeably to labour markets, however.  There was a stabilization in employment, but growth never got much above 0.6% yoy during the years of undersupply, and there was no noticeable or significant rise in wages. Rather, these were precisely the years when companies managed to push up the sales/expenses multiples of employees to highs of around 8.3x previously seen only in the immediate aftermath of the bubble years.  In short, employers reacted to a perceived shortage of labour by raising the productivity of their employees and booking the profits.  The moral is that it will take more than the perception of labour shortages to push up wages sharply. Not only is there no sign that this is taking place now, there is little reason to expect it to arrive any time soon. 

Inflation and Inflationary Expectations Finally, it is worth querying something that seems obvious – that the short bout of supply-side constraints did very little to move inflationary expectations, or, indeed to significantly mitigate deflationary tendencies in Japan. True, the CPI's downward drift was not significantly interrupted. However, there were signs that at the corporate level, there was a scaling back of deflationary expectations. Specifically, during 2006 to 2008, 15+ years of sharp corporate deleveraging, measured either by total assets/ equity, or by net debt/equity were both halted, and replaced by a short-lived bout of mild re-leveraging!  Although mild, this was also unprecedented in post-bubble corporate Japan: it almost amounts to a rush of blood to the head.

Conclusions?    Looking past the weakness of current data, and the likelihood that the tax rises will produce more such data in the short term, it does look as if for only the second time since the bubble years, Japan is hitting genuine supply-side constraints.  On the positive side, this forms a genuinely encouraging background for a capital spending cycle in a way which defies the current dire data. Quite possibly, it will also see a reversal in long-standing develeraging behaviour by corporate Japan. But the hopes that the cycle will be driven by gains in labour markets, specifically through significant wage rises, are unlikely to be realised.