Are we there yet? Glenn Stevens latest thought-piece on inflation risk - Ellerston Global Macro Fund
October 2017 | In Show All, Global Macro | BY Glenn Stevens
Are we there yet?
In their 2009 book This Time is Different, Carmen Reinhart and Ken Rogoff examined financial booms and busts and ensuing crises, dating back centuries. The title was, of course, deliberate irony. While all booms and busts have their own unique features, they also have a lot in common – not least the refrain that ‘this time is different’.
In the ways that matter most, it almost never was different on any of these occasions. What Reinhart and Rogoff found was that, with very few exceptions over eight hundred years, asset price and credit booms and busts were followed not only by a serious downturn, but almost always by a slow recovery, at least compared with the sorts of recoveries typically seen from garden-variety business cycle downturns.
So it has proved since the great crisis of 2007-2008. The United States, the country that was at the epicentre, responded most aggressively and most quickly to the crisis. The Fed slashed rates, was very inventive in supplying liquidity and adopted large scale ‘QE’ policies. The US authorities pushed ahead very quickly with repair of balance sheets of key institutions and regulatory reform. It is hard to see what more could have been done in the monetary and financial spheres (though on the fiscal side we could perhaps debate whether large-scale infrastructure projects, funded at the lowest interest rates in human history, might have been worth a try). Even with all this, US real GDP did not regain its 2007 peak until 2011. And it still has not regained the level of GDP that a continuation of the 1980-2007 trend would have predicted. In fact, the US economy is still diverging from that trend, which shows that potential growth is lower these days. That may be in part because of resource misallocation prior to the crisis, and in part due to other things like shifting demographics and poor underlying productivity performance, which had been concealed in the pre-crisis boom.
In Europe, for reasons very much embedded in the way the ‘European project’ works, fixing the banks was a much slower process. The banking crisis became a sovereign debt crisis for weaker players in the Euro area and at one stage threatened the euro itself. The ECB responded but even with that, Euro area GDP didn’t regain its pre-crisis peak until the middle of 2015.
But another message of Reinhart and Rogoff was that, eventually, usually after quite a lengthy period, economies and financial systems move on from crises, and things improve. So today, a decade after the failure of Northern Rock showed how serious the crisis was becoming, and nine year after the failure of Lehman Brothers saw it erupt into a full scale financial meltdown with potentially catastrophic consequences, the question perhaps is: are we finally there? Are we, at long last, emerging from the shadow of the worst financial crisis since the 1930s? And if so, can we look forward to a resumption of more ‘normal’ times, including more normal settings of monetary policy and more normal long-term interest rates?
There is certainly some encouraging evidence. The US appears now to be more or less fully employed, as measured by most of the standard labour market metrics – even if that is occurring at a level of GDP considerably below what would have been expected based on experience up to 2007. Apart from a brief period in 2000, the standard unemployment rate has not been lower since 1970. The initial unemployment claims data have been at their lowest level since the early 1970s – and that is in absolute terms, when the US economy has almost double the number of people in its workforce that it did back then.
Economic performance in the Euro area has improved considerably over the past year or so. As usual, there are differences between countries. Germany is booming, with unemployment the lowest it has been since the fall of the Berlin Wall, while others are less advanced in recovery. Still, even France and Italy have seen some decline. The euro area’s unemployment rate has re-traced about half the rise seen from 2008 to 2013, which owes a good deal to Germany’s strength and the sizeable fall in Spain’s unemployment from very high levels. Spanish unemployment, though still high, is, believe it or not, almost back down to its average for the 20 years prior to the crisis.
Over 2017, moreover, there has been an increasing sense of a more synchronised and robust global upswing. This is most easily observed in high frequency business survey data which have generally remained positive in the US and have moved up noticeably in Europe. China’s economy has, if anything, gained a little speed this year – defying the pessimists yet again. Even in Japan, recent Tankan readings are the highest since the ‘bubble economy’ period in the early 1990s. The IMF upgraded its global forecasts at the Spring meetings in April – the first such upgrade for a few years and seems likely to do so again at the Annual meetings this month.
How likely is it that this will continue?
Forecasters have never been very good at predicting recessions, or turning points in general. Historically, it has usually been some time after a turning point has actually occurred before it is well recognised. Admittedly, the army of private analysts and staffs supporting policy-makers have, over the past twenty years, become a lot better at ‘now-casting’: that is, recognising where the economy has been in the very recent past and guessing where it might be right now. But they are still not very good at predicting a major turning points far enough ahead to help policy makers avert major problems, or to help investors adjust their portfolios.
With that disclaimer, what can we say?
At one level, the synchronous nature of the expansion gives it a self-reinforcing characteristic, all other things equal. If everyone is growing now, it makes it easier for them all to keep growing, as incomes expand and create further demand, absent some force which outweighs that dynamic. Both at home and across borders, the circular flow of income, multipliers and so on work to keep an expansion going.
The fact that the expansion has been moderate in pace thus far also means that inflation has, to date, remained quite low, unusually so for an upswing that, in calendar time, is now getting on. (The US upswing, for example, is now 100 months old or more, one of the longest since business cycle records have been kept.) That has meant that monetary policies everywhere have remained remarkably accommodative which, in turn, has ruled out, so far, the expansion ending because of abrupt policy tightening to counter inflation in the fashion seen from the 1960s to the late 1990s.
Of course it’s not just inflation that can help to precipitate the end of an upswing. Over the past couple of decades, the usual story has more often one of financial problems: some sort of ‘imbalance’ or ‘vulnerability’ that arose in an economy during the expansion and which put the economy at heightened risk of a reversal. A build up in debt, associated with rising asset values, took place. Then at some point an event caused people to re-evaluate the likely returns on assets, and hence their prices, and what degree of borrowing seemed sustainable. Sometimes, such re-evaluations turned out not to be that big a deal for the real economy. A case in point was the dot-com bust of the early 2000s. That reversal was associated with a recession in the US, but quite a mild one compared with those that had been associated with the fight against inflation. It seems that the relatively moderate extent of leverage behind the asset holdings at the time was a key ameliorating factor. On other occasions, the asset price-debt-credit quality nexus started a contractionary process that was devastating – as in the great financial crisis in 2007-2008.
The scarring experience of those events spawned a large industry of detailed analysis of every conceivable indicator of debt and financial stress, watching for financial vulnerability of any kind. No one ever wants to be caught out again failing to warn of the risks of debt. There have been plenty of such warnings over recent years. How concerned should we be?
One problem in assessing this is issue that no one can be sure what ratio of debt to GDP or income is too high, because what is sustainable depends on interest rates, among other things. Since interest rates have been remarkable low for the past decade – some suggest the lowest in recorded human history in nominal terms – debt servicing has been fairly easy. Indeed, that is part of how expansionary monetary policy works. Of course there is the question of how borrowers will fare once rates rise – an important consideration and one which, all other things equal, will surely see central banks proceed very carefully in their tightening programs over the years ahead.
But this whole discussion begs the question of why interest rates have been so low. To be sure, central banks in major jurisdictions have been undertaking a remarkable experiment, in an effort to do something – anything – to restore growth in aggregate demand and head off deflationary forces. We could debate how effective this was ever likely to be, if they were ‘pushing on a string’, but they had little choice. That other arms of policy seemed unable and/or unwilling to help much only left the central banks in the major jurisdictions in an even more difficult position.
The expansion of central bank balance sheets undoubtedly bid down yields and alleviated financing constraints that governments and private borrowers would surely otherwise have faced. (By the same token, unless central banks are intending to dump all that debt back into the market very quickly, and they clearly are not, measures of vulnerability due to debt build-up need to be suitably nuanced for the fact that it is the institutions that print money that hold a portion of the debt.)
Still, the question is: was this all the result of central banks? Real yields have also been very low – below zero since late 2008 for the US policy rate, and at the long end the lowest since the short-lived inflation outbreaks in the mid and late 1970s. Can the actions of central banks have held down real interest rates for so long? After all, everything we learned as students said that central banks can’t affect real magnitudes other than in the short term. Monetary policy is about nominal magnitudes, we were taught; that is why central banks ultimately target prices: policy is ‘neutral’ with respect to real things in the long run. A decade or more is getting to be beyond what we normally think of as short run.
Some observers are suggesting that we have to question that ‘neutrality’ proposition: maybe central banks can have effects on some real things for longer than had earlier been assumed. If that were true, it would call into question some key intellectual underpinnings of current policy frameworks.
Others have pointed out that there are periods in history where real interest rates have been very low for extended periods, even as low as today, under very different policy regimes to those in place today. In fact there were periods of apparently very low real rates when there were no central banks in the modern sense of that term.
So maybe the very low rates have not been entirely due to central banks. Maybe potential real growth rates of economies have declined. Maybe animal spirits are chronically subdued. These are variations on the ‘secular stagnation’ idea. Maybe risk preferences of the increasingly aged population in western countries, coupled with the peculiar demographics and particular social setting of the increasingly prominent Chinese population, have led to savings being allocated increasingly to bond-type assets. Maybe rates are low just because they are, and will be until they are not.
It is very difficult to know but looking ahead, the factors behind the very low rates matter a great deal. If the reason for very low rates in fact owes a lot more to deeper real forces than central banks, then the central banks won’t be able to raise rates very far.
If very low rates are mainly due to central bank actions, then as central banks change course we must expect the whole structure of interest rates to rise quite a bit.
A reasonable guess would be a combination of the two explanations: structural forces, ranging from the persistent to the more or less permanent, have lowered real long-term interest rates; and central banks, in their efforts to provide stimulus, have lowered them some more. So some lift in rates will follow if the expansion continues. Even the IMF, which had warned against premature tightening for years, now seems to be endorsing some gradual adjustment.
Central banks will, of course, make this as gradual as possible. Gradualism is in their nature and in addition they are well aware of the posited ‘vulnerabilities’ of higher debt. On the other hand, we know that in past tightening episodes, upsets and volatility in markets tended to be more the rule than the exception. There is always the possibility that some borrowers somewhere, or some set of investment strategies, will get into trouble as financing costs rise, even if gradually.
But one of the other things that could conceivably cause an upset would be a noticeable lift in inflation. Not a break out as in the 1970s, but just enough of a lift to signal that the risk of inflation is something that, once again, needs to be thought about.
Central bankers, for their part, would welcome a bit more inflation – that has been their avowed aim. Thus far, however, market pricing seems to embody doubts that they will be successful.
To some extent that is understandable, given that inflation has in recent years been more remarkable by its absence. Plenty of observers have been pointing out that there are structural reasons for prices to be subdued. Globalisation of the labour market makes it harder for workers in any one country to push up their wages. There is evidence that global capacity pressures have become more important relative to local ones in a range of developed economies. Technological change in the form of disruption to supply chains – the ‘Amazon effect’ – takes out layers of costs. And technology also informs consumers much more fully about prices, effectively increasing competitive pressure, at least for ‘commoditised’ goods and services.
These are all important points, but it surely must also be true that a moderate upswing after a deep downturn has meant that good old fashioned cyclical spare capacity has been at work too, for an unusually long period. While it is fashionable to talk about flat or even broken Phillips Curve relationships, that fact is that those relationships were never that tight. Moreover, unemployment rates have only recently reached the neighbourhood of earlier estimates of NAIRUs in the US, and it can take some time for pressure to build.
Tim Toohey has recently written about why US CPI prints might be noticeably higher in the near term. Below I make a different, though complimentary point about long-run inflation, using some very long run data – back to about 1680 – very conveniently published by the Bank of England. It is UK data, but the point I think has more general applicability than just the UK.
The charts show some interesting things. Note that for about two hundred years or more, periods of high inflation were usually followed by deflation. The inflation line fluctuates around zero. The price level (in the second chart) had very little upward drift.
Readers may find this surprising but in many countries, the idea that we would always focus on the change in the CPI, and always expect it to be positive, is historically fairly recent. It’s not a big exaggeration to say that most of the inflation in human history (in anglophone countries at least) has occurred in the period since World War 2. And it was only once that had been under way for a while that people cottoned on and began systematically to have what these days we call ‘inflation expectations’.
That shows up in another feature. For the first 250 years of the chart, there wasn’t actually much relationship between bond yields and inflation. Presumably there didn’t need to be. After all, if the long run expected rate of inflation was roughly zero, bond holders didn’t need to worry about the value of their wealth being systematically eroded by inflation – they presumably worried about other forms of loss.
All the relationships between inflation and bond yields is post 1950, when inflation was persistently positive.
Now one interpretation of these facts would be that the period from 1950 or so until say 2010 was historically unusual – an outlier – and that we are now back in a world in which there is little inflation to speak of. To accept that view with confidence, though, we would want a good explanation of why that 60-year period was so inflationary, and why it has passed.
So what was it that, post-World War 2, meant that people had to think about expecting a persistently positive rate of inflation, and that this had to show up as a positive relationship between long term interest rates and inflation? And has that factor really passed?
The most obvious answer is policy regimes were different, as were the political circumstances in which they operated. Governments and central banks took a much bigger role in managing economies. It is inconceivable in the modern world that, following a period of high inflation, a country could decide, as Britain did following World War I, that the price level had to return to its previous value. Deflation is simply too painful for an economy, especially one in which everyone gets to vote. Today, the objective would be to simply return the rate of change of prices to where it had previously been, allowing the price level to remain permanently higher.
It is not that this is a bad thing. The point rather is that policy regimes matter. Even if central banks find it harder to generate inflation than they expected, even if there are some structural forces which put downward pressure on price levels – Amazon etc – we haven’t gone back to a world of commodity money; we are still in a fiat money world, and in a world of democratic politics that would find it excruciatingly difficult to deal with a situation in which nominal incomes persistently fell. If one thinks that institutional arrangements and so on matter, then it is hard to believe that we have seen the death of inflation.
None of this is to suggest that inflation is about to return to the ‘bad old days’ of the 1970s. It is just that it is worth thinking about how much – or rather how little – compensation for the risk of inflation investors have been offered in recent years.
For now, it is sufficient to observe that there is reasonable chance that a pick up in inflation could see markets flip from their current scepticism about inflation targets being approached from below, to worries about them being exceeded. This, if it were to occur, would put central banks in a rather difficult position – wondering whether to tighten faster to ‘catch up’ and retain anti-inflation credibility that was so costly to acquire 25 years ago, or to give more emphasis to a market led tightening in financial conditions slowing growth.
There are of course other risks to the global expansion. ‘Geo-strategic’ ones loom large – perhaps the most unsettling of which is the possibility of conflict on the Korean Peninsula. People will have thought through various scenarios but it’s not clear we have much experience to go on in a conflict where both sets of combatants have nuclear weapons. There is not much economists can say.
It is still possible that financial excesses will ultimately cause the demise of the current expansion – and that is the direction in which most official radars are trained.
But maybe, if we really are now moving out of the shadow of the crisis, some of the more conventional cyclical factors will start to come into play once again. That would be more likely, in the US, should the Trump Administration actually manage to persuade Congress to enact an expansionary tax package. Expansionary US fiscal policy, tightening US monetary policy: that would be something we have seen before, with considerable implications for interest rates and currencies. But that is a topic for another day.
 Carmen M. Reinhart and Kenneth S. Rogoff This Time is Different, Princeton University Press, 2009.
Gertjan Vlieghe Real Interest Rates and Risk, at http://www.bankofengland.co.uk/publications/Pages/speeches/2017/995.aspx
 Twenty years ago, it was de rigueur for central banks to be told they should pre-empt economic developments. Since the financial crisis the advocacy of pre-emption has tended to give way to the fear of being premature.