The price of uncertainty and uncertainty about prices: monetary policy in the post-COVID-19 economy

Keynote speech by Fabio Panetta at a Capital Markets webinar organised by the European Investment Bank and the European Stability Mechanism

Frankfurt am Main, 1 July 2020

The coronavirus (COVID-19) shock creates a uniquely high degree of uncertainty about the economic outlook[1]. We lack clarity about when the virus will be tamed and when social distancing measures can be relaxed; about the effects of the lockdown on confidence and growth; and about how the structure of the economy itself will evolve in response to the shock.

Policymakers need nonetheless to form a view on the balance of risks in the economy. This is essential to communicate their reaction function, to decide how to act today and to influence expectations about tomorrow. Demonstrating to the public that macroeconomic policies are in control is crucial in conditions of high uncertainty. Insofar as it creates confidence, this is itself a form of stimulus. This imperative has guided the ECB’s response to the crisis, and Europe’s response more broadly.

Today I would like to discuss, first, the economic outlook for the euro area; second, the competing forces weighing on inflation; and third, how macroeconomic policies should react in this environment of heightened uncertainty. My main message is that downside risks are prominent, but a pre-emptive and decisive policy response has put us in a much better position to avoid more severe scenarios.

The uncertain economic outlook

Since the start of the crisis, the ECB’s monetary policy has stepped in forcefully to prevent vicious feedback loops between financial markets and the real economy that could have derailed the outlook for inflation. The expansion of our pandemic emergency purchase programme (PEPP) in June further secures the financial conditions currently needed to shore up the economy through the crisis.

Fiscal policy has in turn capitalised on supportive financial conditions to absorb, in large part, the loss in private sector income, preventing a much steeper surge in unemployment. Guarantee schemes and banking supervision crisis-relief measures have avoided a credit freeze, supporting a record uptake of liquidity during the latest round of the ECB’s targeted long-term refinancing operations (TLTROs).

As a result, the euro area has so far weathered the consequences of the shock about as well as could be expected. Recent survey data suggest that the economy may have bottomed out and we are already seeing tentative signs of recovery.

But these positive indications need to be taken with a pinch of salt, since the high levels of prevailing uncertainty make it difficult to rely on past regularities. Moreover, recent PMI releases only signal a stabilisation or improvement in the rate of contraction of activity from the very low levels reached in April. We should bear in mind that the Eurosystem’s macroeconomic projections require a strong growth bounce back in the third quarter to achieve our baseline scenario.

The key question remains whether we are really on course to quickly recover our pre-crisis GDP level and then return to a healthy growth and inflation path.

The uncertainty surrounding this question is evident in our projections. Our baseline scenario sees the economy broadly returning to its pre-crisis level of output by the end of 2022. In our severe scenario, however, output would still be around 5% below its pre-crisis level at that time (Chart 1).[2] Likewise, survey data indicate that uncertainty about the outlook for inflation over the medium term remains elevated (Chart 2).

I see three main risks to growth and inflation that warrant a cautious outlook. The first risk is that the synchronised contraction in economies worldwide – and the likely asynchronous pace of recovery – could severely depress growth and inflation.[3] This puts a much greater onus on achieving a strong domestic recovery.

We saw a similarly synchronised European recession after the Lehman crash, but an effective policy response by the Chinese authorities supported the global economy at large and emerging markets recovered relatively quickly, allowing euro area countries to redirect their production to satisfy external demand. Today, that does not appear to be feasible to the same extent. Indeed, the sharp fall in the euro area trade surplus and the weakness of forward looking indicators for world trade – such as the global PMI for export orders – corroborate the view that we will not be able to export our way out of this crisis (Chart 3).

This leads to the second major risk to growth, namely the risk that a tepid recovery could set off endogenous recessionary forces which take longer to reverse, and which could adversely affect the financial sector.

The main factor underlying this risk is that the worst of the impact on labour markets may be yet to come. According to ECB estimates, the euro area unemployment rate could exceed 10% in the third quarter. Some workers on short-time work schemes and temporary lay-offs may not be able to return to their jobs and hiring looks likely to stay subdued. Survey data already point to a sharp deterioration in firms’ employment expectations (Chart 4).

If job losses were to continue to rise, it could upset the employment-income-consumption cycle that was the driver of the post-2013 recovery.[4] That could in turn spread the pain of the downturn from the sectors most severely affected by COVID-19 – such as retail, transportation and tourism – to more cyclical services like business services and real estate.[5] Recent evidence suggests that there have been significant spillovers, reducing demand also in sectors that have not been directly affected by the lockdown measures.[6]

These second-round domestic risks to the economic outlook could in turn be exacerbated by the third factor, which is heightened uncertainty and its effect on investment and consumption. Faced with such uncertainty, many firms are postponing capital spending.[7] Likewise, household consumption has fallen, bringing the propensity to save to historical highs (Chart 5). This partly reflects “forced saving”, as during the lockdown people have been unable to spend to the same extent in shops or restaurants. But we are also seeing significant precautionary saving.

Uncertainty is partially being driven by concerns over the structure of the economy after the crisis, especially in such vulnerable sectors as tourism, entertainment or transportation. It is possible that people working in these sectors will start to revise down not only their current earnings, but also their permanent income. COVID-induced uncertainty might leave scars even after the economy recovers: after a shock to industrial production, for instance, two-thirds of the effect persisting after five years is due to the response to uncertainty.[8]

Such effects typically diminish only slowly, because firms and households need to be convinced that the shock was a tail event.[9] As a result, the consequences of the COVID-19 crisis will likely be more persistent than those of standard recessions.[10]

Demand- and supply-side forces on inflation

For monetary policymakers, the key question in this environment is what impact these developments will have on medium-term inflation.

The various risks weighing on aggregate demand that I have described, if they materialise, will be unambiguously disinflationary. ECB research finds that in a situation of labour market disruptions, heightened uncertainty and financial amplification, inflation would be 0.8 percentage points lower by 2022 than under the pre-crisis macroeconomic outlook.[11] Absent appropriate policy responses, a revival of COVID-19-related tensions could therefore push the euro area into deflation. Indeed, after the outbreak of the crisis in March, the probability of deflation over the next five years increased abruptly; it has subsequently more than halved following the announcement of the ECB’s policy measures.[12]

But it has also been argued that COVID-19 may have inflationary effects because the strong demand-side stimulus by central banks and fiscal authorities could run up against a less efficient supply side.[13] In my view, however, this possibility does not correspond with the situation we face.

Looking at the fundamental determinants of inflation, the euro area economy entered the crisis with a widening output gap and inflation hovering around 1%; inflation expectations were at historically low levels. The correlation between inflation and economic activity had markedly decreased (i.e. the Phillips curve had flattened).[14] And natural rates of interest had fallen continuously, pushing policy rates towards their effective lower bound.

This suggests that a severe downsizing in production capacity would be necessary to produce broad-based supply shortages, since there was already excess supply. And even if capacity constraints were to appear, they would not necessarily produce significant inflationary pressures, owing to the flatness of the Phillips curve, the rise in the unemployment rate and the downward trend of inflation expectations – although there may be some heterogeneity across countries depending on their starting positions.[15]

In other words, faced with a large negative supply shock, the forces that have been holding back inflation over the last decade would likely continue to do so.

And, in fact, these dampening forces might even be strengthened insofar as COVID-19 triggers feedback loops from supply to demand. The COVID-19 shock has the peculiar feature of being a negative supply shock that, paradoxically, could end up producing excess supply. For example, a set of firms closing their doors due to the lockdown might not only compress production, but simultaneously reduce demand in other sectors producing complementary goods.[16]

Moreover, past experience suggests that negative supply shocks can worsen “supply-side secular stagnation”,[17] where firms become pessimistic about future growth and invest less, causing natural interest rates to fall and inflation expectations to weaken. Past pandemics – going back to the 14th century – have been found to lead to lasting declines in the European real natural interest rate of almost 2 percentage points, which are only fully recovered after 40 years.[18]

But inflation can also be affected by one-off structural factors that emerge gradually and therefore have persistent effects on prices. Over the past 20 years, both globalisation and technology have tended to depress prices in this way. COVID-19, however, could now lead to these forces pulling in opposite directions. We may well see de-globalisation as adoption of digital technology accelerates.

How this shift in the balance of forces will play out is uncertain. But my suspicion is that the net effect will still be disinflationary.

Certainly, if firms seek to increase supply chain resilience and manufacturing is re-shored, it could increase cost pressures by mitigating international competition, reducing margins and inducing firms to pass on cost increases.[19]

But reshoring manufacturing firms tend to accelerate automation, in order to avoid higher labour costs.[20] And the impact of COVID-19 on digitalisation could contain inflation through several channels.

First, it may broaden the market for tradeable services and expand the pool of workers available to firms, generating downward wage, cost and price pressures. Second, it could increase the market share of “superstar” firms, which tend to have a lower share of labour in value-added[21] and higher negotiating power in the labour market, which in turn weakens workers’ bargaining power. Third, the pandemic is accelerating the shift towards e-commerce, which will further sharpen competition between online and traditional retailers;[22] for the euro area, the disinflationary effect of e-commerce is estimated to have been 0.8 percentage points cumulatively between 2006 and 2018, with larger effects in more digitalised economies.[23]

Finally, the COVID-19 crisis is influencing work patterns in ways that could have persistent downward effects on prices in specific sectors. For example, remote working will probably become more widespread, which could depress rents for commercial real estate (since less office space will be required) and for residential real estate in large cities (as people can now live outside towns). This could have a meaningful impact in countries such as Germany, where rents make up close to 15% of core inflation and have added 0.2 percentage points to inflation on average since 1999.

Our latest projections are consistent with the view that medium-term pressures on inflation will ultimately be downward: after the outbreak of the crisis, our inflation projections for 2021 and 2022 were revised down by 0.6 and 0.3 percentage points, respectively; moreover, projected inflation falls monotonically with the severity of the crisis scenario (from 1.7% in 2022 under the mild scenario to 0.9% in the severe scenario – Chart 6).

Financial markets agree with this view: five-year forward five-year ahead inflation swaps are currently trading at around 1.1%. Accordingly, policy rates are expected to remain low, with the EONIA forward rate standing at around -0.3% and 0.1% in five and ten years’ time, respectively (Chart 7).

Policy responses under uncertainty

Disinflationary risks and lower growth justified – indeed, required – the vigorous reaction of macroeconomic policies in the euro area. With interest rates already close to their effective lower bound, providing the necessary accommodation for the ECB meant engaging in large-scale asset purchases under the PEPP.

But the argument for acting decisively was also strengthened by a further factor: the singular features of the crisis mean that the benefits of acting aggressively to ward off downside risks exceed the costs of potential overshooting.

To start with, heightened uncertainty increases the likelihood that the economy could shift between multiple equilibria, including a scenario of protracted low growth, low employment and low inflation. In these conditions, providing policy certainty through forceful actions becomes critical.

Macroeconomic policies are today playing a pivotal role – perhaps more so than ever before[24] – in stabilising financial markets and supporting private incomes. For example, how quickly precautionary saving diminishes and investment recovers hinges crucially on whether policies are perceived to be commanding and credible. This is visible in survey data collected by the ECB, which show that households with more confidence in government support expect lower future unemployment, higher income growth and display lower precautionary behaviour.

Likewise – with net private and public debt issuance in the euro area expected to increase dramatically in the coming years – financial market dynamics rest heavily on confidence in the overall policy mix. Deploying a rapid, determined and credible response – as the ECB has done – can produce a virtuous circle, where risk taking is restored and more investors choose to buy alongside the central bank, reinforcing its policy, rather than using its interventions to sell to it.

Such a response in turn eases financial conditions for all sectors, removing doubts about agents’ ability to finance necessary spending, and thereby lifts private expectations of future demand, moving the real economy into a better equilibrium.

By contrast, if financial conditions were allowed to tighten, evidence points to a significant non-linearity in the reaction of the economy – the effects are much greater during downturns and recessions. Had the ECB therefore not reacted strongly, the impact on GDP and inflation of the shock in financial conditions could have been much greater, up to three times larger than what we typically see.

The case for acting pre-emptively and forcefully becomes even clearer when one considers the potential “hysteresis” effects which the COVID-19 shock could trigger.

Since the sectors currently most affected by the pandemic are those that traditionally invest the most in productive capital,[25] there is a tangible threat of “capital hysteresis”. And if the most severe scenarios for growth materialise, employment could fall significantly over the coming years, presenting a serious risk of rising structural unemployment. Participation rates might also be impacted, not least because women – who have contributed in large numbers to rising labour force participation in recent decades[26] – are represented to a greater extent in those sectors that are primarily affected by the pandemic.[27]

Yet another important reason for forceful monetary policy measures is the high degree of complementarity among macroeconomic policies today. It is clear that monetary policy cannot be the only force countering the shock.

This is partly due to the asymmetry of monetary tools in dealing with downside risks when policy rates are close to their effective lower bound. It also reflects the unique conditions of the COVID-19 recession. For example, with the private sector essentially shut down for a period of time, banks are facing higher credit risk and may therefore be reluctant to make new loans, irrespective of financial conditions; similarly, monetary policy cannot improve health conditions or induce people to go on holiday. These problems can only be addressed by fiscal policy.[28]

In this environment, monetary policy needs to act proactively to prevent fragmentation of financing conditions, as in doing so we create the conditions for underpinning aggregate demand. This, in turn, allows us to achieve our price stability goal.[29]

Since the start of the crisis monetary and fiscal policies in the euro area have acted – independently – as complements, reinforcing each other and effectively removing tail risk from the market. The monetary policy measures taken by the ECB, especially the introduction of the PEPP, and the fiscal measures at European level – from the €540 billion safety nets agreed by finance ministers for workers, businesses and sovereigns, relying on the expertise of the European Investment Bank, the European Stability Mechanism and the Commission, to the announcement by the Commission of the €750 billion Next Generation EU recovery plan – have guided investors away from a bad equilibrium. As a result, systemic stress has decreased significantly.

By acting vigorously to avert financial fragmentation, the ECB has made its credit support policies more effective, allowing the transmission of monetary policy to regain traction faster.


The European response to COVID-19 has been remarkable. The ECB’s policy measures have avoided fragmentation in the euro area, dispelled tail risks in financial markets and stabilised the economy, thereby preventing inflation from falling even further below our aim. As a result of our actions so far, we are now in a better position to watch how key developments play out.

Our reaction function should nonetheless be clear: the ECB will respond to any significant tightening in financing conditions for as long as the negative effects of the COVID-shock persist. We remain ready to adjust all our policies depending on how incoming data affect our assessment of the medium-term inflation outlook.

We may be now moving from the first phase of the crisis – which was about protecting our productive capacity – to the second phase, which is about rebuilding and modernising the European economy in response to the shock. COVID-19 will inevitably lead to structural changes. It is the responsibility of governments to master and accompany them with the necessary reforms. 

In order to avoid exiting this crisis with further economic divergence, concerns about competitiveness and long-term sustainability will have to be addressed. We now have the opportunity to move towards an economy that is more agile, more digital, more green and more inclusive – supported by substantial European resources provided by the recovery fund. Seizing this chance is crucial to ensure a more resilient recovery today and more sustainable growth tomorrow.

[1]I wish to thank Alessandro Giovannini, Jean Francois Jamet and Jonathan Yiangou for their excellent support in the preparation of this speech.
[2]Eurosystem staff macroeconomic projections, June 2020.
[3]IMF research finds that highly synchronised recessions last on average 40% longer than other recessions, and their amplitude is 45% greater. Recoveries from such recessions are also very slow. See Kose, A. and Terrones, M. (2015), “Synchronization of Recessions”, Collapse and Revival: Understanding Global Recessions and Recoveries, IMF.
[4]See Anderton, R., Aranki, T., Bonthuis, B. and Jarvis, V. (2014), “Disaggregating Okun’s law − Decomposing the impact of the expenditure components of GDP on euro area unemployment”, Working Paper Series, No 1747, ECB, December.
[5]See Forsells, M., Kennedy, N. and Timm, L. (2019), “Developments in the services sector and its relationship with manufacturing”, Economic Bulletin, Issue 7, ECB.
[6]Brinca, P., Duarte, J.B. and Faria-e-Castro, M. (2020), “Measuring Sectoral Supply and Demand Shocks during COVID-19”, COVID Economics, Issue 20, London: CEPR Press.
[7]According to estimates by the European Investment Bank, corporate investment could fall by around twice as much as it did during the financial crisis, when a decline of 19% was recorded. See EU firms in the post-COVID-19 environment: investment-debt trade-offs and the optimal sequencing of policy responses, VoxEu column.
[8]Ludvigson, S., Ma, S. and Ng, S. (2020), “Uncertainty and Business Cycles: Exogenous Impulse or Endogenous Response?”, American Economic Journal: Macroeconomics, forthcoming.
[9]Kozlowski, J., Veldkamp, L. and Venkateswaran, V. (2019), “The tail that keeps the riskless rate low”, NBER Macroeconomics Annual, 33(1), pp. 253-283.
[10]One parallel is the effect on consumption of disaster episodes, such as world wars or global depressions. Research finds that only half of the peak drop in consumption due to disasters is subsequently reversed. The estimated permanent effect on consumption is -14%. See Nakamura, E., Steinsson, J., Barro, R. and Ursúa, J. (2013), “Crises and Recoveries in an Empirical Model of Consumption Disasters,” American Economic Journal: Macroeconomics, 5(3), pp. 35-74.
[11]Angelini, E., Damjanovi, M., Darracq Paries, M., Zimic, S. (2020), “ECB-BASIR: a primer on the macroeconomic implications of the COVID-19 pandemic”, Working Paper Series, No 2431, ECB.
[12]Probability implied by five-year zero-coupon inflation options, adjusted for risk premia.
[13]In a recent survey of investors, 37% saw inflation in developed economies as an under-priced risk globally. See BofA Merrill Lynch Fund Manager Survey, May 2020.
[14]See Eser, F., Karadi, P., Lane, P., Moretti, L. and Osbat, C. (2020), “The Phillips Curve at the ECB”, Working Paper Series, No 2400, ECB.
[15]That said, Germany had an estimated positive output gap for four years from 2016-19 and also experienced relatively weak inflationary pressures.
[16]Guerrieri, V., Lorenzoni, G., Straub, L. and Werning, I. (2020), “Macroeconomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Shortages?”, NBER Working Papers, No 26918, National Bureau of Economic Research.
[17]Gordon, R. (2016), “Secular Stagnation on the Supply Side: U.S. Productivity Growth in the Long Run”, Digiworld Economic Journal, No 100, 4th quarter.
[18]Jordá, Ó., Singh, S.R. and Taylor, A.M. (2020), “Longer-run economic consequences of pandemics”, CEPR working paper, No 14543.
[19]Looking at the components of core inflation in major economies over the past 20 years, while average inflation in the goods component has fluctuated around 0%, that of services has fluctuated around 2%. A possible explanation is that higher trade openness and competition in the goods sector has held down inflation, as well as putting pressure on firms to raise productivity.
[20]One recent study argues that COVID-19 will reduce supply chains by 35% and accelerate robot adoption by 76%. See Kilic, K. and Marin, D. (2020), How COVID-19 is transforming the world economy, VoxEu column.
[21]Autor, D., Dorn, D., Katz, L.F., Patterson, C. and Van Reenen, J. (2019), “The Fall of the Labor Share and the Rise of Superstar Firms”, Quarterly Journal of Economics, forthcoming.
[22]Belz, S., Wessel, D. and Yellen, J. (2020), “What’s (not) up with Inflation?”, Brookings Institute working paper.
[23]Anderton, R., Jarvis, V., Labhard, V., Morgan, J., Petroulakis, F. and Vivian, L. (2020), “Virtually everywhere? Digitalisation and the euro area and EU economies: Degree, effects, and key issues”, Occasional Paper Series, No 244, ECB, June.
[24]Governments are playing an unprecedented role in supporting household incomes. If headline unemployment figures were adjusted for people who are temporarily not working and are being subsidised under government schemes, unemployment rates could be up to two to three times higher than the headline rate in several countries.
[25]Manufacturing and retail trade, transport (including travel), accommodation (including hotels) and food and beverage sectors have been among the largest contributors to developments in investment in machinery and equipment.
[26]ECB (2017), “Recent developments in euro area labour supply”, Economic Bulletin, Issue 6.
[27]For example, accommodation and food service activities; arts, entertainment, recreation. See International Labour Organization (2020), COVID-19 and world of work: Impacts and responses.
[28]Governments have addressed these problems by giving guarantees for bank loans and tax incentives for holidays.
[29]Jarociński, M. and Maćkowiak, B. (2017), “Monetary-fiscal interactions and the euro area’s malaise”, Working Paper Series, No 2072, ECB, June.
1 July 2020

Media contacts