The paths of inflation in advanced countries have been puzzling over recent years. While inflation rates fell sharply during the financial crisis and thus behaved as expected, it is much harder to explain their behaviour since then. In fact, rather than rebound as might have been expected as economic activity stabilised and then recovered, the global inflation rate fell from 4.2% in 2012 to 3.5% in 2014 and, in OECD countries, prices dropped from 2.3% to 1.7%. The drop in inflation is lasting despite the fact that oil prices have stabilised and a recovery is under way in the US and UK and burgeoning in the euro area. In the US for example, core CPI inflation has fallen from 2.3% in January 2012 to 1.8% in March 2015. In the UK inflation rate fell during the same period from 2.6% to reach 1%, below the targeted 2%.
This persistence of low inflation has consequences for the economic outlook and monetary policy. When inflation is low, companies, households and governments have a harder time cutting their debt loads, a problem that is still acute in a number of countries in the Eurozone. Moreover, with low inflation expectations, the private sector may postpone expenditures, thereby weakening economic activity.
In fact, inflation is now below target in most advanced economies. As a result central bankers are asking themselves whether this decline is caused by common shocks, by a coincidence of domestic factors or by structural changes in the reaction of prices to the output gap.
While recent price developments in the US and in the UK are largely explained by global disinflationary forces, oil prices, currency appreciation and the sensitivity of inflation to unemployment in a context of ongoing economic recovery1, my overall assessment regarding the euro area situation is that a sizeable domestic component is at play. This has motivated the bold steps taken by the Eurosystem to counteract the risk of inflation remaining too low for too long a period of time.
Part A – What happened to Euro-area inflation
Let me now describe recent price developments in the euro area in more detail. HICP inflation has fallen consistently below our target since January 2013, declining from 2% to −0.1% in March 2015. Forecasters have repeatedly been surprised by this evolution. Three main reasons have been put forward to explain this.
A first explanation is that forecasting has been complicated by the fall in global crude oil prices and raw materials and its transmission to other sectors. In this regard, recent research conducted at the Banque de France supports this view by showing that forecast errors for oil prices explain about half of the errors for headline inflation. The oil price futures used in our forecasts turned out to be poor predictors in volatile oil markets2.
But downside surprises also occurred for core inflation suggesting the need for complementary explanations. One candidate is the continuous appreciation of the euro between mid-2012 and the beginning of 2014, which has lowered prices of many imported non-energy industrial goods. One can expect this impact to be transitory as the euro has subsequently depreciated since the middle of 2014. This depreciation reflects the differences in the growth cycle between the euro area and its main partners and accordingly the differences in monetary policy.
Those explanations do not account for the slowdown in domestic prices, which must then be attributed to an increase in domestic slack. For the European Commission the output gap widened from −2% in 2012 to −3% in 2013. This came as a surprise because realized GDP growth rates in 2013 and 2014 were about 0.5 percentage points lower than forecasts made in 2012.3 Our analysis suggests that this economic slack contributed substantially to low inflation. For example, in the fourth quarter of 2014, if we assume no economic slack, HICP inflation would have been higher by around 0.8 percentage points. Such a counterfactual exercise is however surrounded by a large uncertainty. First, potential output is hard to estimate. That said, if potential output and domestic slack were lower than assumed, the weakness of inflation would be all the more puzzling. Second, some observers argue that the Phillips curve of the euro area is steeper than before,4 partly explaining the low level of inflation. Although we do not share this diagnosis for the recent past,5 a stronger sensitivity of prices to slack might be observed in the future as a result of labor market reforms and lower unemployment in several countries. Third, downward nominal wage rigidity may blur the relationship between inflation and domestic slack. Nominal wage inflation, which remained persistent during the crisis, is likely to stay subdued during the recovery, thus leading to a lower inflation.
One caveat specific to the euro area is the asymmetric adjustment process currently at work among jurisdictions. Member countries that experienced large capital inflows in the years preceding the financial crisis experienced significant reductions in price competitiveness. The subsequent fiscal and financial adjustments have put a strong downward pressure on prices and wages (the HICP fell in 2014 in Greece, Spain, Portugal, Ireland and Cyprus). To achieve relative price adjustment while keeping aggregate euro area inflation close to 2% would require higher inflation in the so-called core economies. In principle, this could occur through a rebalancing of financial flows from Core to peripheral countries within the euro area and higher domestic spending in core economies. In practice, net savings rates have remained high in several core economies with capital instead flowing outside of the euro area, mirroring an increasing current account surplus. Downward price pressures are therefore occurring in peripheral economies without offsetting upward price pressures in the core economies. For example, the year-on year inflation rose in 2014 in Germany by 0.8% and in France by 0.6%.
In summary, in the Euro area half of the decline in inflation can be attributed to a fall in energy prices and another quarter to the fall in food prices. The remaining quarter is explained by the slowdown in core inflation, which has declined by 0.7 percentage points, reflecting a lower inflation in services and in non-energy industrial goods.
Part B – Monetary policy in the Euro Area
Let me turn now to monetary policy. The Eurosystem’s decisions are based on its assessment of the path of inflation over the medium term. Yet, measuring inflation expectation is a difficult task. One reason is that monetary policy is transmitted to the price level with a varying lag. Survey measures of inflation expectations can also be an unreliable guide. For example, research conducted at the Banque de France has shown that survey-based measures are biased in a systematic and predictable way.6 It has been shown that professional forecasters surveyed by the ECB fail to systematically update their 1-year or 2-year ahead inflation forecasts when new information is released, and the distribution of their forecasts is wider when they update them. This explains our preference for using a range of indicators to gauge inflation expectations. Nevertheless, in the last couple of years, most measures of inflation expectations, be they survey-based or market-based, have pointed in the same direction of a prolonged period of too low inflation, well below our target of being close to but below 2% over the medium term.
As a consequence, monetary policy has become progressively more accommodative. First, in a context of fixed rate full allotment introduced on the 15th of October in 2008 to offset liquidity risk in the market by ensuring banks’ continued access to liquidity, we have introduced forward guidance to avoid an unwarranted tightening of the effective monetary policy stance. Initially implemented in July 2013, this guidance was “firmly reiterated” in early 2014 and remains in place today. Second, the Eurosystem introduced targeted longer-term refinancing operations (TLTROs) and the asset purchase programs for covered bonds and asset-backed securities in June 2014, with the goal of improving the transmission of our monetary policy. Third, the main refinancing rate was cut to 5 basis points and the deposit rate entered into negative territory in two steps (-20 basis points).
However, by January this year it was clear that economic recovery as well as monetary and credit indicators were too weak to bring medium term inflation close to but below 2%. Market-based measures of inflation expectations, even if prone to liquidity shocks, had been falling further below the target and at longer term horizons. The Governing Council was determined to act boldly to counteract the risk of a too low for too long inflation.
In this context, an expanded asset purchase programme was then launched. On March 9th, the Eurosystem began purchasing public sector securities in the secondary market along the curve from 2 to 30 years, being mindful of being market neutral. Purchases of securities at negative yield are permissible, as long as the yield is above the deposit facility rate. Marketable securities purchased under this program will be made available for securities lending.
This programme is intended to stimulate the economy through different channels.
The first one is the portfolio rebalancing channel, in which the fall in yields on safe government bonds induces investors to buy riskier assets, thereby stimulating the economy. Since the launch of the programme yields have dropped not only for the assets purchased but also for bank and corporate bonds and equity prices have continued to rise.
A second channel works through lowering and flattening the yield curve. The purchase programme will continue until the end of September 2016 and beyond if we do not see a sustained adjustment in the path of inflation. This reinforces the guidance that interest rates will remain low for an extended period of time.
A third channel is related to increasing credit supply. If the Euro-system buys assets from non-banks or institutions, then these funds will reappear as deposits in the banking system. This creates an additional source of funding for banks which in turn increases the supply of credit to the private sector.
A last channel is the signaling effect. By acting decisively and more aggressively than anticipated, we hope to lift inflation expectations.
Let me conclude with my assessment of the impact of the QE. So far, recent indicators suggest that our quantitative easing has had a positive effect on inflation expectations. This is true when looking at market-based measures at short term horizons, although less so at long term horizons. Looking at the recent release of the ECB survey of professional forecasters, there is an upward trend in inflation expectations over the medium to long term.
But, as shown by the US experience, re-anchoring inflation expectations takes time. The Eurosystem is ready to go further if necessary to deliver on its mandate of maintaining inflation close to but below 2%. But monetary policy will be more effective if, at the same time, governments increase the potential growth rate by implementing ambitious structural reforms.
1 A recent BdF working paper argued that short-term unemployment is more meaningful than total unemployment when looking at the Phillips curve in the US . Cf. Ferrara L. and Sestieri G. in Banque de France • Quarterly Selection of Articles • No. 36 • Winter 2014-2015.
2 Bec, Frédérique and Annabelle De Gaye (Nov. 2014), “How do oil price forecast errors impact inflation forecast errors? An empirical analysis from French and U.S. inflation forecasts.” Banque de France Working Paper 523.
3 European commission forecast of Nov. 2012: 0.1 in 2013 and 1.4 in 2014. Realized growth: −0.4 in 2013 and 0.9 in 2014.
4 Peter Praet mentioned this evidence of a steepening in some countries of the Euro Area, in his recent speech “Price stability: a sinking will-o'-the-wisp?” during a panel at the IMF Spring Meetings Seminar in Washington D.C. in April 2015. See also the working paper of the Banca d’Italia by Riggi, Mariana and Fabrizio Venditti (September 2014): “Surprise! Euro area inflation has fallen” published in Questioni di Economia e Finanza 237,
5 Chatelais Nicolas, Annabelle De Gaye and Yannick Kalantzis (11 May 2015) “Low inflation in the Euro area: import prices and domestic slack”, Rue de la Banque 6.
6 Andrade and Le Bihan, 2013, Inattentive professional forecasters, Journal of Monetary Economics, volume 60, p. 967–982.