Vítor Constâncio: Challenges for euro area monetary policy in early 2018
Speech by Mr Vítor Constâncio, Vice-President of the European Central Bank, at the 32nd International ZinsFORUM, Frankfurt am Main, 6 December 2017.
The views expressed in this speech are those of the speaker and not the view of the BIS.
The Governing Council of the ECB at its October meeting decided to reduce the rate of monthly asset purchases. In my remarks today, I would like to explain the rationale behind that decision, and how it remains consistent with our mandate for price stability.
In short, as our asset holdings rise and the growth outlook improves, unchanged policy parameters actually imply greater monetary stimulus. Furthermore, I am confident that the apparent disconnect between growth and inflation is a temporary phenomenon and that inflation dynamics will increase.
It is therefore possible for us to scale back our net asset purchases, while keeping our policy sufficiently accommodative to support those inflation dynamics. If one extrapolates from the current brighter economic outlook, one cannot imagine that we would need to extend further our present purchases. A symmetric approach to our inflation mission would therefore be more appropriate than a one-dimensional promise to do more in case of deflation. Managing QExit will be challenging, however.
While ending the purchase programme quickly could provoke undue market reactions, we should not overlook the fact that the longer our asset purchase programme continues, the less effective the programme and the greater the risks attached to it become. Having a credible view of the exit is important to keep any such risks contained.
The recent policy decision
Let me begin with the adjustments to our monetary policy parameters made in the October Governing Council meeting. The asset purchase programme will be extended to September 2018, at the lower amount of €30 billion per month, or beyond if necessary, until the Governing Council sees a sustained adjustment in the path of inflation consistent with its definition of price stability.
This decision was made in an environment of robust demand growth, but with still modest increases in the underlying inflation rate. We have now benefited from 18 uninterrupted quarters of economic growth, with indicators pointing to continued robust growth. The unemployment gap - an established measure of labour market slack - is now closed, and employment has reached its highest ever level in the euro area.
In short, our monetary policy has successfully stimulated demand and returned slack resources into productive use. One would expect this strong demand to generate a marked increase in inflationary pressures. Yet this has not yet fully materialised in the euro area, although signs of inflationary pressures and limits to the lags have been constant for the last six month. Why then have we reduced our purchases before without there being a substantial adjustment in inflation?
The decision to reduce the monthly rate of purchases rests on two factors. The first involves understanding how the stimulus provided by our asset purchases evolves over time, and the second relates to the relationship between output and inflation.
The recessions caused by the global financial crisis and the sovereign debt crisis required significant monetary stimulus to counter the large degree of slack in the economy and ensure that inflation remained consistent with our price stability definition. Monetary policy also had to react to a number of headwinds, which reduced the impact of a given level of interest rates on economic demand.
Some, such as an ageing population and the slowdown in measured productivity growth,1 are longer term in nature. Others were more temporary. The process of deleveraging pushed up desired saving, uncertainty weighed on investment and a prolonged period of low inflation threatened to de-anchor inflation expectations, with a risk of outright deflation.
But as the ongoing recovery gains momentum, the influence of these temporary headwinds is fading. Fears of deflation have dissipated. In these circumstances, unchanged monetary policy parameters would in fact imply an unwarranted increase in monetary stimulus, and when one looks at the 10-year German government bond yield since the summer one might conclude that, despite fewer purchases, we had eased our monetary policy, a step which does not correspond to our plan.
A further reason why the Governing Council elected to scale back our monthly purchases is that, as the size of our portfolio increases, the marginal impact of a given level of purchases increases. Owing to our already large bond holdings, the free float of securities is now substantially lower than when we began our programme in 2015. This increases the relative share of our purchases in the market, and therefore the impact of those purchases per billion euro spent. In other words, a slower pace of purchases today can have an equivalent effect on yields as a higher pace yesterday.
Moreover, as the stock of bonds rises, our reinvestment needs rise in tandem and we have to increase our gross purchases - that is, our net purchases plus our reinvestments; and do that in an environment of lower new borrowing in most euro area countries. This means that, even with a slower pace of net purchases, the Eurosystem will still have a substantial presence in the market. Indeed, cumulative redemptions of around €150 billion are expected in 2018.
For all these reasons, an adjustment to our policy stance was necessary at our last meeting. By the same logic, we should be able to further adjust our policy in the future as confidence increases about the robustness of the recovery without choking off growth or inflation. And the more the recovery advances, the less the economy depends on unconventional monetary policy stimulus and benefits from balance sheets repair, structural reforms and fiscal tailwinds.
As for the current rate of inflation, it is worth noting that the euro area is by no means the only developed economy exhibiting robust output growth with inflation rates below historical experience. I do not have time today to do justice to all the various explanations put forward by economists. In truth, I'm not particularly persuaded by any of them. I do not believe that such factors will prevent us from achieving our mandate over the medium term. We may take a little longer to achieve our objective, but achieve it we will.
Let me focus on just one explanation - that the relationship between domestic slack and inflation, termed the Phillips curve, has become so flat that changes in output have little effect on inflation. Prolonged periods of low inflation can reduce wage growth, since wage-setters pay some attention to past outturns. Moreover, job insecurity, digitalisation and high unemployment have encouraged unions to prioritise employment over wages. And the deep downturn has led to broader slack in the labour market - i.e. involuntary part-time and temporary work - that needs to be reabsorbed before wage pressures rise.
Yet we are going through a transition phase out of a very deep crisis, and these downward forces should wane. Indeed, the difference between the headline unemployment rate and broader measures of labour market slack has already edged down in the past two years. The share of firms that say that labour shortages are limiting their production is at its highest level ever in manufacturing and services. This is translating into higher wage growth now, and in turn contributing to a pick-up in inflation. Various measures of underlying inflation, which is key for assessing durability, appear to have turned a corner.
I say "appear", because policymakers should be humble in their pronouncements.2 Estimates of the degree of slack in the economy vary widely and are often revised. Model outputs, however complex they may be, should be treated with due caution, and judgement based on experience remains an essential part of the policymaking process. We should avoid making unwarranted statements about points in time too far into the future over which we have little true visibility.
It also means that policy decisions not only should be based on the most likely outcome predicted by models, but should also take into account the balance of risks. There are times when the risk of inaction far outstrips the risk of action, and the prudent course is to act with vigour. Introducing asset purchases to stave off the risk of deflation is one recent example. But there are also times when the balance of risks lies in the opposite direction, and prudence dictates that policymakers should be more circumspect.
Future challenges for monetary policy
On that note, while I believe that the forces already at work should in due course bring an end to the need for asset purchases, it is a matter for the ECB's Governing Council to decide on the exact timing of such a move. Our approach should evolve in tandem with our improved expectations for developments in the real economy and ensure that our mandate for price stability is fulfilled over the medium term.
Let me instead spend a few moments considering some of the risks we face in implementing our policy over the coming years. I should also add at this point that these risks are likely to grow the longer our asset purchase programme continues.
The first risk relates to the subsiding deflationary headwinds I have already mentioned. Could it be that our current monetary setting is suddenly too accommodative? In short, because of the uncertainties and imprecisions involved in measuring slack and inflationary pressures in the economy, we might find ourselves behind the curve without realising it. Hence, through the long and variable lags of monetary policy, we will end up with inflation above the rate consistent with our price stability mandate.
This would require a sharp correction of the monetary policy settings in years to come. Yet such a correction of interest rates would pose risks to the financial sector. Banks could be hit hard as funding costs rise faster than interest income on outstanding loans.
The second risk is a related one, in that these factors could have already unwound, and inflation could turn out much higher than we expect over the course of 2018. We would find ourselves behind the curve - and realise it - but the optimal monetary policy response would involve us having to adjust our forward guidance.
This forward guidance was put in place to stabilise market expectations and to enhance the effectiveness of our asset purchases. This is because we indicate that interest rates will rise only when we are "well past" the horizon for asset purchases. So prolonging asset purchases pushes out market expectations of the first rate rise, depressing interest rate expectations across the curve. Of course, the effectiveness of that guidance also relies on market beliefs of our future actions.
In effect, forward guidance is a promise not to react to data outturns in the future so as to persuade markets to maintain low expectations for interest rates. But if we offer guidance that extends too far into the distance, beyond the point where we can form reasonable expectations about the economy, it risks unduly tying our hands unless this is backed by unanimity in the Governing Council.
This is a particular risk at present when, as I described earlier, any future increase in interest rates needs to be gradual, and hence moving late could result in policy remaining too loose for too long. Certainly as the time comes to reconsider our monetary policy stance during the course of 2018, we should reflect at length on the degree to which we wish to pre-commit ourselves.
Risks to monetary policy extend beyond the immediate outlook for inflation. Our policies have reduced the spread over risk-free interest rates paid by households and firms for their borrowing, and unblocked the flow of credit to the real economy. Such spreads are at their lowest ever level.
Yet reduced spreads may encourage investment in businesses that are only profitable at low interest rates, and such loans risk turning sour as interest rates rise. Indeed, banks and investors may be tempted to "search for yield", without being adequately recompensed for the risk they are taking on. The normal place for monetary policy is not at the long end of the yield curve, but at the short end.
The focus of monetary policy is of course on our price stability mandate, and ensuring financial stability risks are adequately mitigated is the role of microprudential supervision and macroprudential policy. While there is little evidence at present of area-wide credit-fuelled bubbles, there are some notable localised pockets, such as in commercial real estate, and evidence of "search for yield" behaviour.
The financial crisis showed how such risks can interfere with the smooth operation of monetary policy through their effects on banks, which remain a key part of the monetary transmission mechanism in the euro area. We should therefore bear in mind that these risks could potentially complicate our ability to meet our price stability mandate in the future.
There are other longer-term risks to monetary policy, namely through the allocation of credit to both productive and unproductive firms. By reducing interest rates for all firms, monetary policy may indirectly permit inefficient firms to remain in business - becoming so-called zombie firms. This blunts the productivity-enhancing function of downturns to bring about "creative destruction", whereby inefficient firms are forced out of business, freeing up resources to move to more efficient firms and boosting aggregate productivity. Indeed, there is evidence that creative destruction was weaker during the Great Recession than in previous downturns3 and that zombie firms have weighed on productivity growth in some euro area countries.4
Productivity growth plays an important role in raising aggregate living standards, but it also affects the conduct of monetary policy. Higher productivity growth spurs investment, and expectations of higher future income encourage households to spend more today. Thus slower productivity growth requires monetary policy to lower interest rates by more than would otherwise be the case to stimulate the economy. Given that interest rates are currently low, and our stock of purchases is quite large, this would restrict our ability to respond to future downward shocks.
What I have just explained are the economic risks arising from a potential misallocation between productive and unproductive businesses brought about by our asset purchases. But there are also legal risks for the ECB regarding our Treaty obligations. Article 127 requires that we "act in accordance with the principle of an open market economy - favouring an efficient allocation of resources". Thus, in respecting our price stability mandate, we should only do as much as is necessary, and be aware of potential side effects from running expansive policy for too long.
An important principle of an open market economy is price formation in markets through the interaction of private sector agents. It should be those interactions that ensure correct pricing, notably of credit risk, and not the interactions created by our asset purchases.
It is clear that, with the Eurosystem now owning public and private sector assets amounting to over €2 trillion, we have become a bigger player in the market than ever before. This means that we are now buying bonds from more price-insensitive investors, such as pension funds and insurance companies, bidding up the price at which we need to pay. Liquidity conditions are expected to become more challenging in certain market segments the longer the purchasing programme remains active.
As such, we have to be mindful not to exert an undue influence on price formation.
Another potential complication relates to our public sector purchases. A key safeguard that we have set up for these purchases is to operate so-called "blackout periods", where we do not buy around the date of a new issuance. This facilitates price formation and ensures that Article 123 of the Treaty is fully respected - the monetary financing prohibition.
Mindful of self-imposed constraints in this respect the remaining space for both net and gross purchases will largely be in newly issued bonds. This poses some potential complications that we have to monitor very closely.
Conclusion
Let me conclude.
The recovery in the euro area continues at a robust pace, employment has risen strongly. Both wages and underlying inflation appear to have turned the corner.
At its October meeting, the Governing Council chose to reduce the rate of net asset purchases, which, as I have explained today, prevented an unwarranted increase in monetary stimulus. We will continue to monitor developments in the economy and set policy in a way that is consistent with our price stability mandate.
In doing so, we must also take into account the balance of risks when setting policy.
If we withdraw our monetary policy stimulus too early and too fast, asset prices could drop and yields rise sharply, with negative spill-over effects to the economy. At the same time, we have to be mindful, as our asset purchase programme continues, that the risks attached to it may increase the longer it lasts. Nourishing a market belief that the exit might be permanently postponed could exacerbate the potential cliff effects. A credible perspective on exit is needed to keep these risks contained.
A sense of proportion will therefore be crucial in managing the QExit.
1 See, e.g., Gordon, R. (2016), The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War, Princeton University Press. It is worth noting that others contest the slowdown, arguing it relates to measurement error, e.g. Brynjolfsson, E. and A. McAfee (2011), Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy, Digital Frontier Press.
2 See Mersch, Y. (2017), "Economic Policy and the Need for Humility", speech at the Conference "Banking and Financial Regulation", Bocconi University, 9 October.
3 Bartelsman, E., López-Garcia, P. & G. Presidente, "Cyclical and structural variation in resource reallocation in Europe", mimeo.
4 See, e.g. See Gopinath et al. (2015): "Capital Allocation and Productivity in South Europe", NBER Working Paper No. 21453; Borio et al. (2016): "Labour reallocation and productivity dynamics: financial causes, real consequences," Bank for International Settlements Working Papers 534; Andrews, D. & F. Petroulakis (2017), "Breaking the shackles: Zombie firms, weak banks and depressed restructuring in Europe", OECD Economics Working Papers No.1433.