Monday 27 October 2014

China's 3Q GDP - Nominal Conclusions

One popular reaction to China’s quarterly GDP announcements is simple disbelief: no statistical agency can hope to estimate quarterly GDP so quickly with any accuracy, so they must just be making it up. My approach is to acknowledge that China’s statisticians face a task which is not only logistically impossible, but also conceptually very difficult (because China is changing so quickly). Nevertheless, through herculean efforts and heroic assumptions, they produce a set of figures which need not simply be written off, but rather interrogated for what they might tell us about China’s trajectory.

The least-useful series is the one that generates the headlines: I have never managed to find any combination of reported monthly economic data -  demand or supply - that bears any statistically interesting relationship to the series of real GDP growth which in 3Q produced a growth rate of 7.3% yoy.  All one can say is that theheadline 7.3% yoy growth in real GDP continued the sequential slowdown vs trend which has been uninterrupted (except in 4Q) since 2010.  Officials seem resigned to a further slowdown in yoy terms in 4Q: if this happens it will represent a further genuine loss of momentum slowdown. If momentum stabilizes, however, we can expect a modest rebound in the yoy rate in 4Q - to around 7.5%.

Nominal Conclusions. However, the nominal GDP series is altogether more interesting, and has much to tell us about how China's economy is functioning.  In summary, it tells us is that the headline nominal GDP was boosted considerably by a doubling of the trade surplus and rather less by a modest fiscal relaxation. Private sector nominal domestic GDP growth (ie, excluding the impact of trade and fiscal balances) slowed sharply in yoy terms, but stabilized sequentially. The private sector’s caution, however, resulted in a rise in the private sector savings surplus to around 4.9% of GDP, which suggests the financial system’s underlying cashflows are slightly improved.   Overall the efficiency of finance within China’s economy remains miserable, but has stabilized, and possibly improved slightly.


If the aim of reforms is genuinely to remodel the economy away from investment and net exports, and towards domestic consumption, those reforms are emphatically not yet producing results. If the short-term aim of reforms is to improve capital allocation and drag China away a debt-dependent growth model, it is probably fair to say that 3Q represents a stabilizing first step on a long and difficult road.

Nominal growth slowed to 8.6% yoy in 3Q from 9.1% in 2Q, but in sequential terms this was actually a modest outperformance of historic quarterly patterns. If maintained, it means that the slowdown in yoy rates may well have hit bottom in 3Q, and be in recovery from 4Q.

The impact of trade. However, nominal GDP was propped up substantially by the trade surplus, which rose 108% yoy in Rmb terms during 3Q, and was equivalent to 5.2% of GDP, and added 2.9 percentage points to nominal GDP growth. Excluding the improvement in the trade surplus, domestic nominal GDP grew only 5.8% yoy,  the slowest since 1Q09. If China’s strategy partly involves weaning itself off the export sector, it failed in 3Q.

The impact of fiscal policy. Secondly, nominal GDP growth was also slightly improved by a slight relaxation in fiscal position, with a Rmb276bn deficit in 3Q up 36.7% yoy from 3Q 13, and equivalent to 1.8% of GDP (vs 1.5% of GDP in 3Q).  On a 12m basis, the deficit is running at 2.4% of GDP, up from 1.5% during the 12m to 3Q13, but the 12m total reflects, as it always does in China, the dominating fourth quarter deficit.   Still, subtracting the impact of the fiscal deficit from nominal domestic GDP leaves a private sector domestic demand growth rate of only 5.3% yoy, down from 5.9% in 2Q, and cutting the 12m to just 7.3%.   However, even after all these subtractions, it remains the case that sequential growth in this figure was slightly faster than historic seasonal trends, and that the fall in the 3Q yoy reflects the sequential slowdowns of 4Q13 to 2Q14, not the underlying dynamic of 3Q.

The impact of private savings/investment decisions.  Meanwhile, the private sector’s underlying economic caution continues to grow, with the estimated private sector savings surplus more than doubling in 3Q to around Rmb 960bn (vs Rmb 469bn in 3Q13), equivalent to 6.3% of GDP. This highly seasonal number is up from 3.4% in 3Q13, and pushes the 12m to 4.9% of GDP (vs 3.5% in the same period last year).  Whilst the rise in private sector savings surplus depresses growth, as it shows the private sector consuming and investing a smaller proportion of its income than at any time since 2010, it has the benefit that it alleviates underlying cashflow pressure in the financial system: the private sector savings surplus, after all, measure the net flow of cash from the private sector to the financial system.  


The impact of finance. Third, with M2 growth slowing to 12.6% yoy in 3Q, whilst nominal GDP growth slowed only to 8.6%, it is just possible that the collapse of monetary velocity (GDP / M2) seen in the immediate aftermath of the 2009 credit binge, and subsequently in the post-2011 slowdown, is stabilizing. This observation is tentative, provisional and vulnerable: however, the slight rise in monetary velocity in the 12m to 3Q reverses the (seasonally expected) falls of 1Q14 and 2Q14, and is slightly better than the stabilization normally encountered in 3Q.  Just possibly, the tighter credit environment is resulting in very slightly better credit allocation. However, please note the caveats. 



But if the efficiency of finance is improving, it is doing so only very marginally, and from historically very low levels. One can also track trends in the efficiency of China’s financing by looking at the relationship between marginal additions to bank lending and/or aggregate financing, to marginal GDP gains.  The chart does that, and it shows that over the last 12 months, one extra yuan of bank lending has been associated with 53 fen of nominal GDP, whilst one extra yuan of aggregate financing has been associated with 31 fen of extra nominal GDP.  In the case of bank lending, this relationship has been effectively unchanged since mid-2013, and whilst it shows an improvement from the credit-binge of early 2009, it remains substantially lower than the levels achieved in 2011-2012, and approximately only half pre-crisis levels.  The crackdown in aggregate financing has resulted in a slightly better story: the 31 fen level is up from 25 fen at the beginning of 2012, but, as with bank debt, if remains at a painfully low level, and less than half the normal pre-crisis levels.  



Thursday 16 October 2014

US Expansion - Still Perversely Non-Cyclical

Here's the problem: the 0.2% mom sa fall in US retail sales ex-autos reminds us that the current US expansion is almost entirely devoid of the cyclical accelerators we always expect from business cycles.

Right now this as a source of woe, since the  expansion simply refuses to develop in the way we are used to tracking. Or one can view it as potentially a source of strength: in recent history it is virtually unheard-of for developed Western economies to generate fundamentally disinflationary supply-led expansions, but that is what seems to be happening in both the US and the UK.  On this 'optimistic' reading, the fact that the economic news repeatedly subverts attempts by central bankers on both sides of the Atlantic to reassert a 'normality' which plainly isn't there, is the good news.

However, it obviously delivers a different set of problems: serial disappointment. Let's concentrate on that.

First, notice that September's retail disappointment probably shouldn't be viewed in isolation. If one looks at the trend growth since 2009, what stands out is that retail sales have never really recouped the losses inflicted by the harsh winter.  The rebound in 2Q never managed to restore current spending levels to the post-2009 growth trendline. September's fall widened that underperformance vs trend to around winter 2014 levels.


That disappointment didn't happen in isolation: average hourly wages were static in September as they had been also in July. Looked at before seasonal adjustments, average weekly wages fell 0.3% mom in September, which was 1.5SDs below historic seasonal patterns. Auto sales have risen more slowly than historic seasonal trends consistently between June and September, with sales rising just 1.9% yoy in the 3m to September.

Constructing a domestic demand momentum indicator which reflects deviations against seasonal trends for employment, wages, retail sales, auto-sales and construction orders reveals the pattern. It's not disastrous: September is revealed as a pretty ordinarily disappointing month, bad enough to depress the 6m trendline back to winter levels, but containing no threat of recurrent recession.


But here's the problem:  this domestic demand momentum indicator has been a reasonably useful guide to US GDP growth ex-inventory movements.  And what it suggests is that we should expect 3Q GDP growth, ex-inventories, to slow to around 1.9% annualized, with 95% boundaries at 1.2% to 2.5%. Right now, the Bloomberg consensus shows an expectation of annualized growth of 3% in 3Q14 and 4Q14.  Frankly, unless there are major inventory surprises ahead (and yesterday's total business inventories growth of 0.2% mom sa in August makes that slightly less likely), those forecasts are too optimistic by far.


Of course, it's only a model, and a fairly crude one at that: it is victim to war, chance, revisions, and the thousand natural shocks that data is are heir to. But the upside surprises  and revisions would have to be fairly dramatic to justify the gap between the data and the consensus.

Monday 13 October 2014

China Fiscal & Monetary Volatility - A Short Guide to the Recent Past

The sheer extremity of the slowdown in China's monetary numbers in July remains startling, even three months later. July was the month aggregate financing dropped to just 273bn yuan, the lowest since October 2008 (which was the absolute climax of the last anti-inflation overkill, and which provoked such a dramatic response) and from 1.975tr yuan in June.  And, of course, all monetary aggregates slumped in sympathy.  


It raises the question: what the heck happened?  Was this financial suffocation actually intended, or was it just tolerated as it emerged? Or did it happen by accident, perhaps the result of some miscommunication between the various arms of government,  the actions of which each make an impact on financial conditions?

Since it's important, it's worth taking a short tour around the recent data, as a sort of financial battle-field trip. The crucial thing to understand is that the PBOC's open market operations, although closely watched, are not necessarily the largest factor determining monthly shifts in market liquidity. That honour quite often belongs to movements in the government's deposits in PBOC. These deposits currently stand at 3.914tr yuan, and whilst over the last year to August, they have risen by a relatively negligible 17bn yuan per month, they have huge volatility, with a standard deviation of 551bn yuan in their monthly movement.   Obviously, if these deposits are rising, that reflects not merely a net fiscal tightening (a fiscal surplus) but is also, ceteris paribus, something that will tighten monetary conditions.

In July, the monthly fiscal surplus was nothing exceptional: 240.6bn yuan, vs 249bn in July 2013 (with revenues up 6.9% yoy and spending up 9.6%).  But this probably understates the impact of government activity on private sector cashflows, because govt deposits in PBOC rose by 566bn yuan in July, up from 446bn in July 2013, and 1.2SDs higher than normal seasonal trends.  In other words, the fiscal tightening in July was greater than advertised.

Moreover, this reversed significant de facto fiscal stimuluses:

  • in June, when government deposits fell 275bn yuan during a month when historically they tend to be near-unchanged, and 
  • in May, when the 105bn yuan rise in deposits was less than a third of the c350bn yuan rise normally expected. 

What is more, this de facto fiscal tightening was not offset by PBOC market operations, which added on average only 74bn yuan per week, down from 204bn yuan in June and from 96bn yuan in July 2013.

At this point, it is tempting to believe that July's tightening had an accidental component, in which the monetary impact of Ministry of Finance's determination to retrieve a fiscal situation which had threatened to deteriorate beyond expectations was not fully appreciated by PBOC.

Moving to August, there are signs of a modest reversal: government revenues are still weak, rising 6.1% yoy on a monthly movement slightly below trend, whilst government spending also slowed to 6.2% yoy, which although below trend, represents a slightly recovery from July's stringency. The  monthly deficit of 109.5bn yuan is virtually unchanged since August 2013. More importantly, the marginal 20bn yuan rise in government deposits is 0.3SDs below the 84bn yuan normally expected in August and is slightly lower than in August 2013.

Meanwhile, there is no sign of relaxed accommodation from PBOC, with open market operations averaging an expansion of just 50bn yuan per week in August, falling to approximately zero in September.

Conclusions time: There is no sign that PBOC has significantly loosened policy in response to July's distress-signal. But both the overt and de facto fiscal squeeze which bit in July was not repeated in August - rather fiscal conditions were normalized.  My interpretation is that the July's financial squeeze was exacerbated, though not solely caused, by the unexpected diligence of the finance ministry in retrieving a deteriorating fiscal position.  






Monday 6 October 2014

BOJ, ECB, Fed: Three Ways to Lose Credibility

The collapse of ex-US confidence beyond anything justified by current economic data is just one early ramification of the dollar's strength. Nevertheless, it focusses attention back on central banks, and in particular on their perceived ability to exercise sufficient influence or even control on financial conditions to head off trouble.  In other words, it puts central bank credibility at a premium, at the same time as undermining it.

And in each major economy, the challenges to central bank credibility have evolved differently. But in each case, the risks of lost central-bank credibility are suddenly more thinkable, and more visible, than they were a few months ago. In fact, the threats to central bank credibility may themselves form a new category of risk to the global economy.

BOJ – Classic currency/bond market/financial system meltdown

The credibility of the Bank of Japan seems the most immediately fragile: it is horribly easy to envisage a situation in which a run on the yen forces up interest rates (in the mild version, to head off inflation; in the alarming version, simply to underpin the currency temporarily), which in turn destroys the value of JGBs held by financial institutions.
Just to put some figures on those holdings:
I) public sector debt makes up 22% of the total asset base of Japan's commercial banks, and are equivalent to just under 30%  of their deposit base.
II) Public sector debt account for 48% of insurance and pension assets, and are equivalent to 61% of total insurance and pension reserves.
III) Public sector debt accounts for 79% of Bank of Japan's assets, and equivalent to 141% of the deposits made with the bank, and are 2.36x the currency issue.

More generally, with public sector debt as a percentage of GDP having sailed past 200% approximately a decade ago, any significant and sustained rise in interest rates would threaten to overwhelm almost any imaginable plan for fiscal consolidation.

So what can be said for Bank of Japan.  Well first, this: that the underlying government debt situation is so extreme, and the ticking of the bond-yield bomb has been so audible for so long, that worries about the Bank of Japan's credibility have themselves developed a credibility problem. The reluctance to believe in the doomsday scenario in which Bank of Japan 'loses control of the situation' isn't simply a matter of recoiling from the horror – it is also testimony to the regularity with which the anticipation of Japan's doomsday has cost investors money.

The best chance of defusing this bomb demands: a) an economy which is growing in nominal terms  but also b) inflation which is sufficiently positive to help foster nominal GDP growth, by changing ingrained deflationary expectations), but sufficiently suppressed in both absolute terms and in volatility, to keep the bond market calmly accepting of a life of modest but sustained loss of value. Tricky, very tricky.

Before the latest devaluation of the yen, however, the trajectory for inflation suggested that BOJ's attempt to rise the inflation rate to around a steady 2%, excluding tax rises, was on course.


So the question now is: what impact is the devaluation of the yen likely to have on CPI? The most obvious impact is on the price of imported fuels: overall, energy has a 7.7% weighting in Japan's national CPI, falling to 4.6% if electricity is excluded, with the largest components being gasoline (2.3%) followed by gas (1.8%). In other words, every 1 percentage point fall in the yen vs the dollar would be expected to raise the CPI by 0.05 percentage points.

Now, in September, the yen declined on average by just under 4% against the dollar, and has fallen a further 2.5% in the earl days of October. Were this to be passed through entirely to gasoline and gas prices, the combined fall would be expected to raise the CPI index by 0.3% during those two months.
But this, of course, isn't what will happen, since a) oil prices have been falling as the dollar has risen; and b) the bulk of Japan's oil contracts will not be priced in spot terms either for the commodity or for the fx rate.  In both cases, in the short term this is likely to mute any inflationary impact from rising yen oil prices.

More importantly, oil remains quite a small part of the overall index weighting: the things which weigh most heavily on Japan's CPI are food (25.3%), housing (21.2%), transport (14.2%, includes oil) and culture/recreation (11.5%) - and for most of these, the pass-through from oil prices is likely to be very small.


ECB – The Discovery of Impotence 

Meanwhile, over at the ECB, Mario Draghi increasingly looks like an honest man getting used to public deception. Too personal? His problem is that ever since his July 2012 willingness to to 'whatever it takes' to save the Euro,  his ability to change expectations, and thus savings/investment behaviour is linked to whether he can make good on this claim.  There are two major problems which would seem to restrict the scope of 'whatever it takes'. The first is simply political: German opposition to quantitative easing seems entrenched, and to have resulted last week in Mr Draghi being unwilling or unable to quantify the size of his earlier stated plans for the ECB to start buying private sector assets in the aftermath of the ECB's latest policy meeting. 

 The second is legal: Article 21 of the ECB's constitution forbids 'overdrafts or any other type of credit facility . . . in favour of Community institutions or bodies, central governments, regional, local or other public authorities' and bans 'the purchase directly from them by the ECB . . . of debt instruments'.  On October 14th, the European Court of Justice will be hearing arguments from Germany academics on the legality of ECB's current bond-buying plans.  

But does it matter just now if confidence in Mr Draghi's ability to deliver the ECB to full-blown quantitative easing gradually ebbs away? Arguably, it matters less that confidence in his promises of ECB largesse is now waning, than that it was present first in July 2012 ('whatever it takes') and again 2Q2014, since then it helped rally bond and equity markets. In the first instance, confidence in Mr Draghi's intentions helped cut the premium of 10yr BBB bonds from roughly 200bps to around 120bps; in the second, it helped push it down further, to around 95bps.   More dramatically, it was part of the circumstances which allowed the premium on 10yr Spanish bonds to fall from c500bps in July 2012 to  slightly more than 100bps now. Arguably, Mr Draghi's ability to buy time then was more important than the possibility that some of the currency in which he bought it may yet turn out to be lacking. 


In turn, this provokes the wider question of what quantitative easing can be expected to achieve in the first place. Whilst it seems likely that quantitative easing, and indeed the prospect of quantitative easing, can and does move asset prices in a way which can be very useful to extremely-stressed financial systems, it is altogether more uncertain that it can effectively change saving/investment decisions in a way which makes a clear impact on the economy. Studies by the US Fed found that the required internal rate of return demanded of corporate investment decisions were extremely 'sticky' and so likely to be surprisingly little encouraged by falls in short-term rates or even long-term bond yields. The results of this are there for all to see: years of quantitative easing in both the US and UK have brought forth expansions which are surprising mainly for their almost complete lack of normal cyclical 'accelerators', or indeed, of any noticeably cyclical structure.  


Fed – The Nostalgia for Normality

And this brings us back to the Fed, which now faces a challenge to its longer term credibility which reflects the curiously a-cyclical nature of the expansion currently underway: a nostalgia for 'normality' which seeks to re-instate a policy-making structure which simply is no longer available. Almost all commentary on the Fed's policymaking in one way or other amounts to attempts to re-interpret, re-state, or (most ambitiously) re-calculate the Taylor Rule conditions. When John Taylor first suggested the 'rule', it was simply put forward as a way of interpreting what had actually been the revealed policy of the Fed – that rate changes had been made to reflect deviations in both the actual rate from the targeted rate, and changes in the output gap (or, more loosely, how far actual output was deviating from potential output). In practice, this output gap was estimated by measuring the deviation over a long-term growth rate which had been stable enough to measure and extrapolate with confidence. 

The problem is that the financial crisis has left the economics profession profoundly uncertain as to the size of the output gap, and similarly unsure as to the potential growth rate of the US.   This ignorance and uncertainty was neatly illustrated by Friday's labour markets release: the 248k mom rise in non-farm payrolls suggested relatively buoyant growth, but the  shocking renewed fall in the labour participation rate suggested this growth was not drawing people back into the workforce as had been expected. If that is the case, the potential supply of labour must be smaller than appreciated (and so the output gap smaller than expected). But, finally, the market doesn't seem to be tightening, since since hourly wages were unchanged on the month.  As far as wages are concerned, it seems that even the mild upward pressure seen earlier in the year is abating. The paradoxes of the report point directly to the impossibility of using Taylor Rule metrics to structure monetary policy in current conditions. 


In these circumstances, the credibility issue the Fed faces is twofold: the the ability to make the right decision; and the ability to persuade the rest of the world that the FOMC knows why it is making it in the absence of a credible rationale. The worry is that the 'nostalgia for the normal' will tempt them to grab for a  solution which looks 'normal' – in this case, by raising interest rates – before any sort of cyclical normality is in fact re-established.

Thursday 2 October 2014

Northeast Asia Gets Ready to Cut Prices

Before the striking weakness of August's industrial data the region seemed to be avoiding the usual fate it meets when the dollar strengthens: export prices were not falling sharply, international terms of trade were holding up, underpinning margins, profitability and cashflow.  It was different this time.  But this week's industrial news from Japan and South Korea changes the picture for the worse: not only did industrial output slump in China (output rising 6.9% yoy only), Japan (output down 3.1% yoy) and South Korea (output down 2.8% yoy), but this slowdown was insufficient to stop inventory ratios blowing out.

The desire to cut inventory ratios, coupled with the accelerated depreciation in the yen, sets the stage for a renewed period of deflation coming primarily from Japan and South Korea. That pricing pressure is likely to radiate out to China and to Southeast Asia, and we're likely to feel its effects very soon.   

First, let's remember that the last few years (since when?) have been a period of exceptional pricing and margins stability for most of Northeast Asia. Deteriorating terms of trade had been a way of industrial and trading life for most of Northeast Asia for as long as most managers can remember: between 1994 and 2011, terms of trade fell almost uninterruptedly for both Japan and S Korea; Taiwan held out a little longer and a little better, but was unable to escape, with a sharp and seemingly unstoppable slide between 1998 and 2011. However, since 2011, the pattern has been near-stability: between 2Q11 and the 3m to August, Japan's terms of trade fell only 5%, whilst they improved 1.7% for S Korea and rose 2.7% for Taiwan.

This new stability wasn't just the result of commodity price movements favouring Northeast Asia's industrial commodity-consumers – though that certainly helped.  But in addition, there had been no repeat of the bouts of savage deflation in dollar export prices experienced in 1997-2002, again in 2006 and once again, briefly in 2010. In its place has been a period of unusual pricing stability for NE Asian exporters. 

It is this stability which August's shockingly weak industrial data puts under threat. To recap:
  • China's industrial output growth slowed to 6.9% yoy, on a monthly slip that was 0.7SDs below trend
  • Japan's output fell 3.1% yoy, and was 1.3Ds below trend
  • S Korea's output fell 2.8% yoy, and was 1.1SD below trend.

Now, this could perhaps be dismissed as a shared blip – and what's more, a blip in the month of the year which matters least as far as industry is concerned.  And it remains the case that momentum trends in G3 imports and Northeast Asian exports remain sufficiently robust to allow a yoy acceleration through the rest of the year (although perhaps slightly less than expected six months ago). It also remains the case that global domestic demand remains in reasonable shape.

However, the problem is that at least for Japan and South Korea, August's slowdown has left  both with an inventory problem which managements will want to address quickly (particularly in Japan).  Japan's inventory/shipment ratio has spiked up to the sort of levels seen in the immediate aftermath of 2011's triple disasters, and again in the angst-ridden period which brought PM Abe to office. Getting rid of these inventories will demand either further cuts in production, or significant price-cuts to get the inventories off the books. Since the end of August, the yen has fallen  more than 5% against the dollar, so the temptation simply to slash dollar prices will be hard to resist.

That then throws the pressure back onto Korea, where the inventory turnover ratio  (total inventories/sales) has been climbing since 2011, and in August reached record heights. 

And, of course, that pressure will also be felt in China.  China's August industrial data was awful too: not only did output growth slow to 6.9% yoy, but the slowdown in the topline crushed profits, which fell 0.6% yoy in August (if the data is to be believed). Once again, if China's data is to be believed, mainland companies have spent the last 18 months or so successfully protecting their margins, and this was still the case in August's slowdown.  However, as the chart also shows, China's margins are highly cyclical and tilted sharply towards the year-end. It is the fattening of these year-end margins which would be directly threatened by Japan and S Korea's efforts to offload inventories by cutting prices. 

Conclusion? It's price-cutting time for Northeast Asia: there will be bargins this Christmas.