Ladies and gentlemen[i],
It is a pleasure to be here with you for this lunch-debate of the Centre des Professions Financières. I would especially like to thank your chairman Michel Pébereau for his kind invitation. Today, I am delighted to be able to speak to you on a highly topical subject: monetary policy in Europe in the context of economic recovery. The prospect of a very gradual normalisation of our monetary policy gives rise to concerns and expectations. I will begin by focusing on what our accommodative monetary policy has already achieved today, before talking about prospects for the future.
Economic recovery in the euro area is here [slide]. It is robust and broadly based across countries and sectors. Our latest ECB forecasts[ii] confirm this: growth should stand at 2.2% in the euro area in 2017, i.e. a significant upward revision of 0.3 point compared with June’s forecast of 1.9%. This means that for the second consecutive year, euro area growth will be comparable to that of the United States. And that, according to Eurosystem calculations, the output gap will return to close to zero by the end of the year (‑0.6%), even though France’s situation is less favourable due to its growth lag (output gap of around 1.4% according to our calculations).
The non-standard monetary policy that we have been conducting since 2014 is contributing to this acceleration of growth in the euro area, by supporting domestic demand and by fostering very favourable financing conditions. In concrete terms, this acceleration results in more jobs – over 6 million jobs created in the euro area since the start of 2013 – and in a pick-up in investment. Corporate investment in particular, decisive for supply capacities, is clearly recovering: up 4.7% in 2015, 6% in 2016, and is expected to be up 4% in 2017 and 2018). [slide] This improvement in investment is underpinned notably by ongoing growth in lending to the private sector, at 2.5% year-on-year in August 2017 for firms and at 2.7% for households, with interest rates moreover remaining very low. This growth in lending is over twice as high in France, which incidentally leads us to be very vigilant at the High Council for Financial Stability.
We are also observing favourable developments in inflation, which is gradually moving back towards our target of 2% over the medium term [slide]. After peaking at the start of 2017, the inflation rate stood at 1.5% year-on-year in August. Over 2017 as a whole, euro area inflation should reach 1.5%, against just 0.2% in 2016; for the coming years, we foresee inflation at 1.2% in 2018 and at 1.5% in 2019 – with a change in composition: less “energy” inflation, more underlying inflation. This represents a first success: our monetary policy measures have succeeded in warding off the risk of deflation which was still threatening the euro area last year – we mustn’t forget that inflation was declining, standing at ‑0.2% in April 2016.
That said, despite the progress made towards our target, it is still too slow. The underlying trend in inflation is still not strong enough to support prices in the medium run. Underlying inflation, excluding energy and food, has admittedly increased – rising from 0.8% in the first quarter of 2017 to 1.2% in July and August 2017 – but it still remains at subdued levels, despite the strong recovery in activity and employment.
A significant factor is the relatively low nominal wage growth. Among the cyclical factors, I would like to mention two in particular: first, persistent slack in the labour market – the unemployment rate has fallen sharply in the euro area, to stand at 9.1% in July 2017, but it is still too high; [slide] second, backward‑looking wage bargaining, which means that low past inflation is feeding through to wages today. The influence of these two causes should nevertheless diminish with the recovery in economic growth and inflation.
Structurally, many studies[iii] suggest that the Phillips curve, i.e. the relationship between inflation and fluctuations in economic activity or unemployment, appears to have flattened since the 1980s [slide]. Inflation seems less responsive to changes in economic activity in advanced countries, notably because of globalisation,[iv] which appears to exert downward pressure on inflation via the decline in imported goods prices and competition from low-wage countries. But a flatter Phillips curve does not call into question this relationship: we are in no doubt about the way we are heading: the recovery and job creations will lead to higher wages and, ultimately, more inflation. But we nevertheless remain less certain as to the speed of this adjustment. We are both confident about the effectiveness of our monetary policy and willing to be patient regarding the time it will take.
Let me now turn to the consequences for our monetary policy. You are well aware that we at the ECB’s Governing Council will decide this autumn on the re-calibration of our policy instruments beyond the end of the year – and, I quote Mario Draghi, “probably the bulk of these decisions will be taken in October”. We are now faced with a simple requirement, in line with our mandate to maintain price stability, and the progress towards our inflation target: we have to reduce the intensity of our net asset purchases, while maintaining overall a substantially accommodative monetary policy.
As regards our asset purchases, we have to reduce their intensity in a pragmatic manner, as we already decided successfully in December 2016. I insist on the word “pragmatic” because, while keeping the current rules, we should on the one hand exploit the margins of flexibility of the programme and on the other hand hold in reserve an additional purchasing capacity – if needed.
That said, monetary policy is not limited to QE, and QE is not limited to net asset purchases: we can keep a substantial degree of monetary accommodation even if we reduce our monthly flow of net purchases.
1/ On QE, the academic literature provides consistent evidence that the impact of asset purchase programmes on the yield curve and asset prices is primarily driven by the total stock of assets held by the central bank (the so called “stock effect”), rather than by the flow of transactions conducted over a given period (“flow effect”). In-house studies of the effects of the PSPP find that holding a stock of bonds equivalent to 10% of GDP lowers the 10-year yield by about 45 bps in the euro-area.[v] By contrast, flows of purchases have no significant additional impact. These results imply that our current programme lowers the 10-year euro-area government bond yield by about 100 bps. These effects are of similar magnitude than results obtained for the programmes conducted in the United States and the United Kingdom.[vi] According to these studies, what matters is the extent to which and for how long central banks’ holdings reduce the availability of assets by removing them from the market. This implies that we can continue to exert downward pressure on the yield curve by keeping our stock of asset holdings at a high level for a prolonged period.
Let me first state the obvious: even if we reduce the monthly flow of net purchases, we will continue to increase the size of our balance sheet. And let me add as a second consideration for later: in any case, the Eurosystem will remain in the coming years a major buyer of euro-denominated bonds thanks to our reinvestment commitment, which we took in December 2015 and which was perhaps not widely enough noticed: the repayments that we will obtain when bonds reach their maturity date will be fully reinvested, thus keeping the size of our asset holdings unchanged.
Most importantly, our monetary policy is based on a set of instruments [slide]: it is not a solo, but a quartet. On top of QE, we play with our policy interest rates, our forward guidance and the provision of liquidity to financial institutions.
2/ Perhaps the most unconventional policy is the use of negative interest rates. I insist on the fact that this measure is the most unconventional because for most people, nominal interest rates are positive. It is true that negative interest rates have their limitations; I have already stressed them in the past.[vii] And at our Governing Council’s meeting in Tallinn last June, we clarified these limitations in our forward guidance stating that we do not intend to lower our deposit facility rate below its current level of -0.40%. But slightly negative interest rates have their virtues, even if they are not very popular among bankers. First, by removing the zero lower bound constraint on expectations of future short-term interest rates, they contribute to easing financial conditions by lowering rates along the curve, with a positive impact on the demand for credit by firms. Second, the negative deposit facility rate (DFR) positively interacts with asset purchases. Indeed, the portfolio rebalancing effect of QE is reinforced by the negative DFR, as banks with excess liquidity holdings have a stronger incentive to reduce these holdings and put their reserves to work. The truth of the matter is that we have not so far seen significantly detrimental consequences of negative rates on the profitability of banks, notably thanks to the favourable fact that the slope of the yield curve has remained positive.
3/ Our forward guidance provides further indications as to the future path of our monetary policy; and by steering expectations, it has an impact on the yield curve. Our forward guidance is clear as to the sequencing: “we expect [the key ECB interest rates] to remain at their present levels for an extended period of time, and well past the horizon of our net asset purchases.” There is no doubt within the Governing Council about this sequencing. In all events, we should not consider raising interest rates until we have reached a sustained adjustment in the path of inflation.
4/ The fourth instrument is the provision of liquidity to financial institutions. In October 2008, we introduced the fixed-rate full allotment policy in all our regular refinancing operations. It was designed in order to guarantee banks’ access to liquidity – and this policy is still active today, for both 1-week or 3-month operations. The maturity of this liquidity provision was then extended with the launch of 3-year long-term refinancing operations (LTROs) and, later, targeted longer-term refinancing operations (TLTROs). The latest series of TLTROs provided banks with up to 4-year funding at attractive rates – as low as -0.4% – in order to further ease financing conditions. In September 2017, a total of EUR 760 billion is still lent to euro area banks through TLTROs, which is substantial, over 7% of euro area GDP.
Before I conclude, let me add a few words on the recent evolution of the euro. A number of factors play a role; monetary policy is only one of them. The sharp appreciation of the euro has notably been stemming from the significant reappraisal of growth prospects for the euro area, as well as from the geopolitical context. The impact of the strong euro on growth is currently offset by the strength of domestic demand in the euro area; but the recent appreciation of the euro exchange rate has led us to revise down slightly the outlook for inflation. Let me recall that the exchange rate of the euro is not a policy target of the ECB. But the exchange rate remains an element that we closely monitor owing to its possible implications for the medium-term outlook for inflation.
I would like to finish by insisting on one word: confidence, but confidence without complacency. Confidence, because our monetary policy is effective and the economic recovery is robust. But confidence without complacency, because monetary policy is not the only game in town. It must be supplemented by national reforms, particularly in France where the current acceleration of reforms is positive. It must also be supplemented at the European level by strengthening the euro area: in order to lastingly boost growth, we must complete Monetary Union by invigorating Economic Union. All those who legitimately reject overloading monetary policy - on both sides of the Rhine - must now firmly support a stronger economic union. Thank you for your attention.
[i] I would like to thank P. Andrade, P. Antipa, G. Cette, M. Dujardin, O. Garnier, A. Lojschova and B. Mojon for their help in preparing this speech.
[ii] Eurosystem staff macroeconomic projections for the euro area, September 2017.
[iii] See e.g. Blanchard (O.), Cerutti (E.), and Summers (L.) (2015) “Inflation and Activity - Two Explorations and their Monetary Policy Implications,” National Bureau of Economic Research Working Paper No. 21726, November.
[iv] Guilloux-Nefussi (S.) (2016) "Globalization, Market Structure and Inflation Dynamics," Banque de France Working Papers 610.
[vii] See speeches of August 2016 (Handelsblatt conference), September 2016 (annual CERS conference) and April 2017 (conference at Columbia University).