Jan Smets: Monetary policy normalization - where do we stand?

Speech by Mr Jan Smets, Governor of the National Bank of Belgium, at the 7th Annual Research Conference "Around a Decade After the Crisis: Heading to the New Global Cycle and Monetary Policy Normalization", National Bank of the Republic of Macedonia, Ohrid, 13 April 2018.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Central bank speech  | 
19 April 2018
PDF version
(47kb)
 |  5 pages

Accompanying slides of this speech.

Dear ladies and gentlemen,

It's a pleasure and honour to be here with you today. Many thanks to the National Bank of the Republic of Macedonia for the invitation.

The last decade has been a challenging one for policymakers, not the least for central bankers. Confronted with a severe economic downturn but determined to deliver on our mandates, we were forced to expand our monetary policy toolkit with negative interest rates, asset purchases and forward guidance. Overall, these measures have been very effective and helped set in motion the current synchronized global upturn.

Notwithstanding the economic recovery, central bankers cannot (yet) rest at ease. First, weak inflationary pressures remain a lingering concern. Second, monetary policy normalization itself poses a new challenge which requires our attention.

My talk today will focus on these topics - low inflation and its implications for monetary policy normalisation -, primarily viewed from a euro area perspective. But let me first take a broader view, as also central banks from other advanced economies are confronted with the same challenges.

Indeed, for some time, advanced economies have been characterized by robust economic growth but puzzlingly weak or inexistent inflationary pressures.

The United States economy recovered relatively quickly from the 2008 financial crisis with real GDP catching up again with its pre-crisis peak level in 2011. The economic expansion has been ongoing since then and may be on its way to register as the longest on record. The improvement in the labour market has been impressive with the unemployment rate falling from its peak at 10 % in October 2009 to 4.1 % currently, a level below most estimates of the natural unemployment rate. Tightness in the labour market has, however, not translated into strong wage and price pressures. From a historical perspective, wage increases have been muted and core PCE inflation has overall remained below the central bank's 2 % target. Looking forward, the inflation outlook is more upbeat, though: wages have accelerated a little of late and temporary factors that have held down core inflation are disappearing. Consequently, US monetary policy makers project core inflation to pick up, running slightly above 2 % in 2019 and 2020. This suggests that the growth-inflation disconnect in the US is temporary.

For the euro area, available data call for more caution. The sovereign debt crisis meant that the cyclical recovery took much longer to manifest itself: real GDP only returned to its pre-crisis peak level in 2015 but the economic expansion has been robust and broad-based since and looks set to continue. The unemployment rate is on a downward path, although it has not yet reached its pre-crisis level. Despite these favourable developments, inflationary pressures remain weak. Underlying inflation has been basically flat at 1 % over the last 5 years - far from our objective of below but close to 2 % -  and wage growth does not yet show convincing upward momentum. Looking forward, underlying inflation is projected to only gradually rise to 1.8 % in 2020.

In contrast to the euro area and the Unites States, nominal weakness has been a permanent feature of the Japanese economy over the last 2 decades. While recently, economic growth has strengthened a bit and the unemployment rate fell to record lows, wage growth has remained weak and inflation stubbornly low. The Bank of Japan appears confident in reaching its inflation target, however, projecting CPI inflation excluding fresh food to accelerate to above 2 % in 2019.

As the three economies find themselves at different points in the business cycle and confidence in inflation getting back to target varies, the respective monetary policy stances differ as well.

On the one hand, the Federal Reserve is well underway with normalization. It ended its net asset purchases in October 2014, started raising the federal funds rate in December 2015 and has recently started to gradually reduce its balance sheet. Consequently, the federal funds rate has again become the primary tool for signalling the monetary policy stance. In March the rate's target range was hiked to 1.5 % to 1.75 %; currently markets expect 3 more rate hikes this year while Fed officials pencil in 2 more.

On the other hand, the Eurosystem and the Bank of Japan remain more cautious, keeping their monetary policy stimulus in place. In 2016, the Bank of Japan even revamped its monetary policy framework introducing negative rates, yield curve control and an inflation-overshooting commitment. In the euro area, the strengthening recovery and improved confidence in inflation converging to our aim have allowed us to gradually start recalibrating our monetary policy package over the past months. We reduced the pace of our monthly net asset purchases and removed some so-called "easing biases" in our communication. This notwithstanding, our overall stance continues to be very accommodative.

The cautious approach taken by the Eurosystem is inspired by the multitude of uncertainties that remain present in the economy and that are complicating our understanding of how the economy exactly works.

I already mentioned one such uncertainty: the link between inflation and economic growth.

Recent research has come up with a number of possible explanations accounting for the slower reaction of underlying inflation to a strengthening economy. One entails that despite the robust recovery, labour market slack might be larger than presumed.

For instance, in the euro area the "broad" unemployment rate - which includes discouraged work-seekers and employees working part-time but who would want to work full-time - rose considerably more strongly during the crisis than the conventional unemployment rate. And the gap between the two remains wider than before the crisis. This suggests that the untapped supply of labour remains possibly more significant than suggested by the simple unemployment rate.

The recent downward revisions in the natural rate of unemployment - here the European Commission's NAWRU (non-accelerating wage rate of unemployment) estimates are shown - also make us suspect that slack in the euro area is bigger than we thought. They imply that the economy could converge towards a lower level of unemployment without creating inflationary pressures.

Past and ongoing structural reforms, for one, could account for the boost to labour supply, and therefore to the economy's potential as a whole. On the one hand, this entails good news as we have more margin to grow. How much precisely remains, however, an open question as the exact magnitude of slack and the pace at which it is eroding remain uncertain. On the other hand, higher slack implies a slower convergence of inflation towards our aim and such a long period of too low inflation can also pose specific challenges, even beyond its underlying cause - the excess capacity.

Let me now try to shed some light on how important this slack story is in explaining the low inflation or whether also other factors play a role here.

A Phillips curve analysis - allowing for time-variation in the parameters - can give us an idea. It caters for many potential sources of low inflation, including idle resources but also lower trend inflation and external price pressures.

In house NBB estimates1 find that since 2013 the negative output gap and weaker import prices have been the most important factors in lowering headline inflation in the euro area. While import prices are now pulling up inflation - reflecting the impact of the recent recovery in oil prices - slack still remains a bit of a drag.

Nominal factors - related to how firms and households set prices and wages - seem to have played a more limited role. As of 2013, trend inflation seems to have fallen a bit below 2 %, but only slightly so. In addition, inflation persistence - which is captured by the ρ in the Phillips curve equation - appears to have increased somewhat. This implies that shocks have a longer-lasting impact on inflation as a result of which inflation converges more slowly to its trend. In this regard, the long period of low inflation could have led firms and unions to become more backward-looking when setting prices and wages, making inflation more persistent. Or it could also reflect that the lower bound has somewhat affected monetary policy's effectiveness in dealing with the shocks hitting the euro area economy.

As the euro area expansion proceeds, most of these disinflationary forces should disappear. As a result, inflation should recover to our target of close to but below 2 %, albeit perhaps a little more slowly given the higher inflation persistence.

Another uncertainty that warrants a cautious monetary policy approach concerns the level of the natural rate of interest. Of course, our estimates about its level are always a bit of a guess as the natural rate is unobservable. What several studies suggest, however, is that it has fallen in several advanced economies since the crisis. That puts an even higher premium on bringing inflation back towards our inflation aim - even if the absorption of slack that accompanies it is of course already a very good reason to do so.

In a low real rate world, the frequency of lower bound periods is likely higher. In such an environment, avoiding inflation to settle at too low levels is of utmost importance as it helps preserve the macroeconomic stabilization capacity of the central bank's nominal policy rate instrument. Were we to allow trend inflation to settle below pre-crisis levels and hence below our price stability definition, it would imply even lower nominal interest rates in steady state. This means that, faced with negative shocks, the central bank would have a smaller margin to reduce rates, so recessions might last longer and inflation would recover more slowly.

True, faced with the lower bound, the central bank could use non-standard measures more often. Yet, such unconventional measures - asset purchases and negative interest rates, for instance -, while effective, are typically seen to have more side effects, including on financial stability, and they also have limits.

Because of the firm commitment to bring inflation back in line with our inflation aim, I thus believe that the prudent, persistent and patient monetary policy approach followed by the Governing Council so far is fully justified.

The different environment that we are operating in since the crisis should make us more wary of past relationships guiding our decisions today and tomorrow. These will continue to be supported by a cautious monitoring of actual data, in particular wage growth and underlying inflation.

The conditionality is embedded in our reaction function in the form of the concept of "Sustained Adjustment in the Path of Inflation" (SAPI). It entails that inflation has to move in line with our aim of below but close to 2 % in a sustainable way. This requires that three criteria, namely convergence, confidence and resilience, have to be fulfilled. At present, we are increasingly confident that inflation is on a rising path and converging to our objective in the medium term. But it lacks some resilience, though, as it still relies on our monetary policy support.

Therefore our monetary policy package remains in place for now. This is fully consistent with our primary objective of realizing price stability. In doing so, we give the economy the time to tap its full potential.

This steady-hand approach of monetary policy should also pay off in the future. For one, it helps ensure that when we start normalizing we will be able to do so in a durable way, meaning in an economy where the nominal and real side are sufficiently strong. Again having mandate-consistent inflation readings will allow the central bank to perform its stabilising role at all times, including during future recessions.

This brings me to my next topic, namely: what would the ECB's monetary policy normalization look like? Even if it is early days to talk in detail about this, our forward guidance gives already some guiding principles that I would like to share with you.

The first decision we will make concerns our asset purchase programme. Its horizon is conditional on SAPI, so once we decide that sufficient progress on this front has been made, net asset purchases will gradually come to an end.

By reinvesting maturing assets for as long as necessary, we ensure that the stock of assets on our balance sheet will remain stable for a good while, and that the downward pressure on long-term interest rates will persist. In other speeches, Peter Praet has already shown that the reinvestment policy indeed has a very persistent impact on term premia2.

Likewise, our continued pledge to keep key policy rates at their current levels well past the end of net purchases has assured in the past and will assure in the future that the short-end of the curve remains firmly anchored at low levels. As we move further in time, the main instrument for shaping the stance will again become the key policy rates, including forward guidance about their likely evolution. As such, our communication will likely be further specified and calibrated to ensure that inflation continues to evolve in line with our objective.

We will continue to inform markets to the best of our ability about the state of our thinking about the way forward, making adjustments to our policy as predictable as possible.

Predictability and gradualism should also limit the spillovers to Central, Eastern and Southeastern European (CESEE) countries when we decide on normalizing our monetary policy.

Indeed, the Federal Reserve "taper tantrum" of May 2013 did lead to sizeable capital outflows from catching-up economies, including in the CESEE region. However, the monetary policy tightening alluded to in Ben Bernanke's tapering speech came as a surprise to financial markets; as said before, the ECB's monetary policy normalization aims to be gradual and predictable. Moreover, since then, CESEE countries have taken steps to reduce their vulnerabilities: most of them now have deeper local currency debt markets, smaller macro imbalances, and stronger macroeconomic frameworks.

Nevertheless, CESEE countries are still exposed to a tightening in global financial conditions, due to their generally elevated external liabilities; at 56 % of GDP on average (in 2017), most CESEE countries have a relatively high external debt ratio. Vulnerability to a rise in global rates varies greatly across the region, however, reflecting not only differences in debt levels but also in the structure of countries' external liabilities, including the role of FDI versus more volatile capital inflows in those liabilities and the shares of foreign currency and short-term debt.

Hence, CESEE countries should take advantage of the current stronger global economy to continue to build resilience to changes in external financial conditions. Structural policies to foster higher productivity and sound fiscal and financial policies remain key. Ensuring efficient corporate insolvency and restructuring frameworks would also help achieve faster and less costly resolution of repayment difficulties should they arise when global financing conditions gradually become less accommodative.

Summing up, the robust non-inflationary growth that we are experiencing now in the euro area should make us, central bankers, neither satisfied nor anxious but rather patient. Uncertainty, for instance about the underlying strength of real variables, calls for a steady-hand monetary policy, finding first confirmation from the nominal side of the economy that price pressures are sufficiently strengthening. But I am confident that inflation will sustainably converge to our aim.

I thank you for your attention.


1 See Cordemans N. and J. Wauters (2018) "What explains the disconnect between real and nominal developments?" forthcoming in the June issue of the NBB Economic Review.

2     See for instance the speech "Assessment of quantitative easing and challenges of policy normalization" given by P. Praet at The ECB and Its Watchers XIX Conference in Frankfurt am Main on 14 March 2018.