Monetary policy in a prolonged period of low inflation
Speech by Mario Draghi, President of the ECB,
at the ECB Forum on Central Banking,
Sintra, 26 May 2014
In the context of a certain disconnect between economic performance and inflation the monetary policy response has to be carefully considered and precisely designed. We are not resigned to allowing inflation to remain too low for too long. But to understand “what is too low for too long”, we need to answer two questions.
First, why is inflation so low?
Second, once we have a decomposition of inflation, we can ask: how likely is it that it persists over the medium-term? Falling commodity prices have accounted for around 80% of the decline in euro area inflation since late 2011. But there are two factors specific to the euro area that contribute to low inflation: the rise in the euro exchange rate and the process of relative price adjustments in certain euro area countries.
At present, our expectation is that low inflation will be prolonged but gradually return to 2%. Our responsibility is nonetheless to be alert to the risks to this scenario that might emerge and prepared for action if they do. What we need to be particularly watchful for at the moment is the potential for a negative spiral to take hold between between low inflation, falling inflation expectations and credit, in particular in stressed countries.
There is a risk that disinflationary expectations take hold. This may then cause households and firms to defer expenditure in a classic deflationary cycle – especially when monetary policy is at the effective lower bound and so cannot steer the nominal rate down to compensate.
On aggregate, euro area firms and households do not seem to be particularly exposed to debt deflation dynamics. But this picture masks the heterogeneity within the euro area. Debt service-to-income ratios tend to be higher in stressed countries. Credit weakness appears to be contributing to economic weakness in these countries. Our analysis suggests that credit constraints are putting a brake on the recovery in stressed countries, which adds to the disinflationary pressures. And heterogeneity becomes a factor in assessing low inflation in the euro area.
In terms of the monetary policy response, the key issue is timing. We have to be mindful of mismatches between the various trends: the rise in demand for credit; the repair of bank balance sheets; and the development of capital markets as a complement to bank lending. At this point of the cycle, these considerations feature prominently in the discussions of the Governing Council members. There is no debate about our goal, which is to return inflation towards 2% in the medium-term, in line with our mandate.
Ladies and gentlemen,
When we first thought about launching this forum in the spring of last year, we saw a clear value in setting aside time to withdraw from the pressure of our daily routines and the immediacy of decisions. The idea was to devote time to in-depth reflection on how to address the fundamental challenges facing central banks.
In the event, the opportunity for in-depth reflection has proven as valuable as we expected – but our ability to detach from the pressures of the moment has been less than we thought. We are meeting against the backdrop of a complex economic situation: a slowly consolidating recovery, but one which has been accompanied by a gradual fall in inflation rates. Cyclical developments are also interacting with structural developments, notably the structural deleveraging of the banking system. All this warrants discussion.
To some extent, a disconnect between economic performance and inflation is to be expected in the very early stages of the economic upturn. As households and firms resume their spending plans in the aftermath of a long period of restraint, they tend initially to use existing resources more intensively. This raises measured productivity, but causes employment to return more slowly towards its potential, which reduces price pressures.
In this context, the monetary policy response has to be carefully considered and precisely designed. We do not want to be too reactive to those parts of the disinflationary process that are expected to self-correct. We neither want to be too forbearing towards those factors that, if left unchecked, can lastingly undermine price stability. We are not resigned to allowing inflation to remain too low for too long. But to understand “what is too low for too long”, we need to answer two questions.
First, why is inflation so low? This is essentially a question about the anatomy of inflation, meaning the nature of the shocks that are causing inflation to deviate from its intended level.
Second, once we have a decomposition of inflation, we can ask: how likely is it that it persists over the medium-term? This is a question related to the physiology of inflation, meaning how current low inflation influences economic behaviour and impacts the economy, notably through the formation of expectations.
Through the filter of these two questions, I would like to use my remarks this morning to discuss the forthcoming decisions facing the Governing Council.
The anatomy of disinflation
Low inflation is not particular to the euro area. Inflation is low across advanced economies, mainly due to the diminishing effect of oil prices on consumer prices. But looking at the anatomy of inflation, there are two factors specific to the euro area that contribute to especially low inflation here.
The first is a common factor: the rise in the euro exchange rate and its effect on the price of internationally traded commodities. The second is a local factor: the process of relative price adjustment in certain euro area countries that pulls down aggregate inflation.
Let me explain each of these in turn.
Falling commodity prices explain the lion’s share of the disinflation the euro area has experienced since the end of 2011. Brent crude oil prices were down by around 7% in euro terms in the first quarter of this year, compared with a year earlier. Food prices were sharply down as well. In fact, these two components have together accounted for around 80% of the decline in euro area HICP inflation since late 2011.
The bulk of the imported downside pressures on euro area consumer prices are explained by the strengthening of the effective euro exchange rate, in particular vis-à-vis the dollar. In the past year or so, oil prices in US dollars have fluctuated – by historical standards – over a relatively narrow range. And they have exhibited no clear downward or upward trend.
This creates a balance of forces that might affect future inflation. On the one hand, lower commodity prices driven by euro appreciation help compensate for the generally weak developments in disposable income in the euro area. Indeed, real disposable income declined at slower pace throughout 2013, and turned slightly positive in the fourth quarter, increasing by 0.6% year-on-year. To the extent that this supports domestic demand in the euro area it will also create upward pressure on inflation.
On the other hand, exchange rate appreciation affects external demand and reduces the competitiveness gains of price and cost adjustment in some euro area countries. This has a countervailing effect on real disposable incomes, while also making disinflation more broad-based. Indeed, if we look at prices of non-energy industrial goods, which are mainly tradable, we see a downward trend across all euro area countries.
To add to this, aggregate inflation has been dragged down by local factors linked to the sovereign debt crisis and the process of relative price adjustment in stressed countries. Several euro area countries are currently undergoing internal devaluation to regain price competitiveness, both internationally and within the currency union. The crucial adjustments vis-à-vis other euro area countries have to take place irrespective of changes in the external value of the euro.
This process began hesitantly in the early years of the crisis, largely due to nominal rigidities in wages and prices. The result was that adjustment took place more through quantities – i.e. unemployment – than through prices. Stressed countries thus experienced a protracted period of declining disposable incomes and long-drawn-out price adjustment. In this context, several have seen domestic core inflation – that is, excluding the energy and food price effects I just described – fall well below the euro area average. For example, the recent overall fall in services price inflation for the euro area is almost entirely accounted for by price declines in these components in stressed countries.
Nevertheless, in the last few years relative price adjustment has accelerated in stressed countries. While this may also have initially weighed on disposable incomes, by creating a closer alignment between relative wage and productivity developments, it should increasingly support future incomes through the competitiveness channel. Export growth has been impressive in several stressed countries. And indeed, nominal income growth in stressed countries turned positive in the fourth quarter of 2013.
The physiology of disinflation
So to sum up: falling energy and food prices, coupled with the effects of relative price adjustment in stressed countries, explain almost fully the disinflation we have seen in the euro area. We also see that disinflation produces counterveiling forces, which may in time cause it to self-correct. To what extent should monetary policy therefore react to these developments?
The answer relates to the physiology of inflation: whether or not these factors are likely to persist into the medium-term and therefore enter the horizon of monetary policy. Temporary movements in the exchange rate or relative price adjustments would not normally warrant a monetary policy response. Given the lag in monetary policy transmission, a monetary impulse would hit the economy just when the effect on inflation has faded out and the impulse is no longer necessary.
That said, shocks can change: in certain circumstances temporary shocks can morph into persistent shocks via second-round effects. In particular, a prolonged period of low or even negative inflation rates might destabilise inflation expectations. And we know from international experience this change can happen quite quickly, especially if the objective of monetary policy is not clear. Thus, we have to judge carefully how an apparently temporary shock is spreading through the economy and affecting expectations.
Moreover, the situation is more complex if there are impairments in monetary policy transmission that extend the lag between our decisions and their impact on prices, as we see in the euro area today. In these circumstances, there is a risk that, if a temporary shock turns more persistent, any monetary policy response might arrive too late to prevent a more serious downward shift in expectations. Thus, more pre-emptive action may be warranted.
At present, our expectation is that low inflation will be prolonged but gradually return to close to 2%. Our responsibility is nonetheless to be alert to the risks to this scenario that might emerge and prepared for action if they do. What we need to be particularly watchful for at the moment is, in my view, the potential for a negative spiral to take hold between between low inflation, falling inflation expectations and credit, in particular in stressed countries.
Let me explain.
Low inflation, expectations and credit
Remember that countries undertaking relative price adjustments have to adjust relative to the average rate of euro area inflation. Low inflation therefore lowers the “nominal bar” around which such adjustments across countries have to happen. In these circumstances, stressed countries will likely experience a temporary period of very low or negative inflation rates. This is what we see in the euro area at the moment.
But here the common and local sources of disinflation I mentioned earlier interact. The effect of an appreciating exchange rate is to hold down overall euro area inflation. The nominal bar around which adjustment takes place is lower: In particular, we see a rise in inflation in non-stressed countries that is insufficient to raise the euro area average back to 2%. And the downward adjustment in the stressed countries becomes probably harder and certainly longer, especially if nominal rigidities imply that prices and costs cannot adjust quickly.
In this situation, there is a risk that disinflationary expectations take hold. This may then cause households and firms to defer expenditure in a classic deflationary cycle – especially when monetary policy is at the effective lower bound and so cannot steer the nominal rate down to compensate.
In addition, an unexpected period of low inflation and low nominal income results in a higher actual and expected future real debt burden. Unless compensated for by expectations of higher future income, firms may reduce investment and households consumption. Banks may in turn respond to this situation with stricter credit standards, which reinforces disinflationary pressure and hence worsens debt burdens. This is fertile ground for a pernicious negative spiral, which then also affects expectations.
Propagation through the euro area
So do we see any signs that low inflation might propagate through the euro area in this way?
On aggregate, euro area firms and households do not seem to be particularly exposed to debt deflation dynamics. The interest payment burden of euro area firms – the ratio of their gross interest payments to gross operating surplus – has actually fallen from 22% in 2008 to less than 12% at the end of 2013, which suggests that firms are in a stronger position today to withstand a period of low inflation. For euro area households debt service-to-income ratios are similar – around 14% – while the median household holds the equivalent of around two months’ income in liquid assets to cushion nominal income shocks.
But importantly, this picture masks the heterogeneity mentioned previously. For firms in some countries the interest payment burden has in fact risen since 2008, in particular for firms based in stressed countries. Debt service-to-income ratios also tend to be higher in stressed countries, implying greater vulnerabilities in these jurisdictions if low inflation persists.
Bank lending also presents a mixed picture. Surveys of bank behaviour show a gradually improving aggregate situation. According to the latest Bank Lending Survey credit conditions generally stabilised at the start of this year, and even improved in some stressed countries. The incremental tightening process that banks reported throughout the crisis seems to have come to a halt.
That said, credit conditions remain very heterogeneous across countries and sectors. According to the latest ECB survey on credit access by small- and medium-sized enterprises (SMEs), supply constraints remain especially strong for SMEs in stressed countries. The percentage of financially constrained but viable SMEs – i.e. those with positive turnover in the last six months seeking a bank loan – varies from a minimum of 1% in Germany and Austria to a quarter of the total population in Spain and as much as a third in Portugal.
Importantly, credit weakness appears to be contributing to economic weakness in stressed countries. To show this, one can undertake a quantitative exercise to compute “normal credit”, similar to the notion of potential output. The difference between the actual volume of credit and normal credit offers a measure of the “credit gap”, analogous to the output gap. Not surprisingly, this exercise reveals that in non-stressed economies the credit gap is either insignificant or making a very small contribution to economic slack. In stressed economies, the same measure contributes to up to a third of economic slack.
This analysis suggests that credit constraints are putting a brake on the recovery in stressed countries, which adds to disinflationary pressures. You can also see why the heterogeneity becomes a factor in assessing low inflation in the euro area.
Calibrating the policy response
What is the right policy response to these developments?
Essentially, we are confronting three issues that might warrant a response. First, the common effect of exogenous factors, including the exchange rate, on euro area inflation. Second, the asymmetric effect of endogenous developments, such as tight access to credit for parts and sectors of the euro area. And third, the risk that those effects combine to generate a more persistent regime of excessively low inflation.
Let me elaborate on those three issues. First, to the extent that developments in the exchange rate, money or capital markets result in an unwarranted tightening of monetary and financial conditions, this would require adjustment of our conventional instruments, so as to secure the degree of monetary accommodation intended by the Governing Council.
At the other end of the spectrum would be a too prolonged downward departure of inflation and/or inflation expectations from our projected baseline scenario, for example due to the interaction between exchange rate developments and medium-term inflation expectations as I explained earlier. This would call for a more expansionary stance, which would be the context for a broad-based asset purchase programme.
An intermediate situation is one where credit supply constraints interfere with the transmission of monetary policy and impair the effects of our intended monetary stance. This would require targeted measures to help alleviate credit constraints. I would like to dwell shortly on this contingency because it relates to the important question of whether weak credit growth results from low credit demand or supply constraints.
Weak credit demand in the early stages of an economic recovery is not unusual. Credit growth typically lags the recovery by three to four quarters as firms draw down internal funds. But a recovery eventually results in growing credit demand. And at this point, for monetary policy to produce its full effects, there must be no binding constraints on credit supply through the banking system. The more the recovery progresses, the more important it is that supply constraints ease so that the recovery can gather steam. Given where we are in the cycle today, it is highly desirable that banks advance the structural adjustment of their balance sheets, so that they can meet demand for credit as it emerges.
It is in part for this reason that we early on placed a great deal of emphasis on a timely comprehensive assessment of bank balance sheets, in the context of the creation of the Single Supervisory Mechanism. Frontloading bank balance sheet adjustment addresses capital constraints on credit supply. Only banks that have fully accounted for legacy losses and hold sufficient equity can take on risk again and therefore originate and price loans in normal conditions. The comprehensive assessment, while still ongoing, has already had a catalytic effect on asset revaluation and provisioning and on capital raising. We expect it also to feed through into new credit as it reaches completion.
Meeting credit demand is likewise why we have publicly supported measures to revive high quality securitisation in Europe. This complements the comprehensive assessment by helping remove capital constraints to loan origination. And it supports the development of capital markets, which will be essential to sustain credit supply while the banking system transitions towards a less leveraged, less risky model.
The key issue today, however, is timing. We have to be mindful of mismatches between these various trends: the rise in demand for credit; the repair of bank balance sheets; and the development of capital markets as a complement to bank lending. Credit demand may pick up more quickly than the other trends gain traction.
If, in this context, availability of term funding is a limiting factor on loan origination, then monetary policy can play a bridging role. Term-funding of loans, be it on-balance sheet – that is, through refinancing operations – or off-balance sheet – that is, through purchases of asset-backed securities – could help reduce any drag on the recovery coming from temporary credit supply constraints.
Let me conclude.
What I have laid out today is a decomposition of the factors behind low inflation in the euro area, and how they interact, percolate through the real economy and may affect medium-term price stability.
At this point of the cycle, all three contingencies I have discussed remain topical and feature prominently in the discussions of the Governing Council members.
Certainly, any policy response requires a careful assessment of the costs and benefits of the various tools at hand. But there is no debate about our goal, which is to return inflation towards 2% in the medium-term, in line with our mandate.