Monetary Policy and the Risk of a Lost Decade

Speech by Benoît Cœuré, Member of the Executive Board of the ECB
at the Amundi World Investment Forum 2013
“Mapping uncharted territories of investing”,
Paris, 5 July 2013

Ladies and Gentlemen [1],

It is a pleasure for me to be here with you.

Today, I would like to discuss the risks which characterise the long-term outlook for the euro area. I will also offer my view on some of the policy responses that are necessary to ensure that those risks do not materialise. The specific risk scenario that I wish to consider is that of a “lost decade” ‑ the possibility of a much more protracted period of stagnation than currently anticipated. Like others, in referring to a lost decade, I am drawing an analogy with Japan’s experience between 1991 and 2000, when average annual per capita GDP growth was less than 0.5%.

With Japan’s experience in mind, a particularly worrisome scenario for the euro area is related to the possibility that the banking sector’s efforts to reduce leverage and to restructure its balance sheets have not yet been completed. Troubled banks’ balance sheets have the potential to choke the engine of recovery and exert a more persistent drag on economic growth. A protracted slowdown may also produce longer-lasting consequences on potential output growth. This is why I will study the implications of such a scenario in some detail.

Let me repeat that this is a risk scenario. In the short term, the modal scenario in the Eurosystem staff macroeconomic projections is that the euro area economy will stabilise and recover in the course of the year, albeit at a subdued pace and with risks continuing to be on the downside. It is the policymakers’ duty, however, to identify the potential factors which may lead to a “lost decade” scenario and design the policy actions that are best suited to pre-empt it.

The euro area outlook

But, let me start first with the current outlook for the euro area.

The euro area is still suffering from the consequences of the financial crisis. Euro area real GDP declined, quarter on quarter, by 0.3% in the first quarter of 2013, following a contraction of 0.6% in the previous quarter. Since the trough in output in the second quarter of 2009, the euro area economy has barely grown: four years later, in the first quarter of 2013, euro area GDP stood only 2.5% higher. Compared with the previous peak in the first quarter of 2008, GDP is still down by 3.3%. Domestic demand remains subdued due to low confidence and high unemployment. Indeed, unemployment figures continue to disappoint. The latest euro area seasonally adjusted unemployment rate was 12.2% in May 2013.

Looking ahead to later in the year and to 2014, euro area export growth should benefit from a gradual recovery in global demand, while domestic demand should be supported by the accommodative monetary policy stance as well as the recent gains in real income owing to generally lower inflation. Furthermore, notwithstanding recent developments, the overall improvements in financial markets seen since last summer should work their way through to the real economy, as should the progress made in fiscal consolidation. This said, the remaining necessary balance sheet adjustments in the public and private sectors will continue to weigh on economic activity. Overall, euro area economic activity should stabilise and recover in the course of the year, albeit at a subdued pace.

The risks surrounding this economic outlook continue to be on the downside. The recent tightening of global money and financial market conditions and related uncertainties may have the potential to negatively affect economic conditions. Other downside risks include the possibility of weaker than expected domestic and global demand and slow or insufficient implementation of structural reforms in euro area countries.

The monetary policy stance in the euro area will remain accommodative for as long as necessary. The Governing Council of the ECB expects the key interest rates to remain at present or lower levels for an extended period of time. This expectation is based on the overall subdued outlook for inflation extending into the medium term, given the broad-based weakness in the real economy and subdued monetary dynamics. In the period ahead, the Governing Council will monitor all incoming information on economic and monetary developments and assess any impact on the outlook for price stability.

If the short term risks scenarios were to materialise, they might push the full recovery back by some quarters. A more worrisome possibility, however, is that of a persistent slowdown. Action is necessary to ensure that it does not materialise. This is what I would like to discuss now.

Zombie banks and the risks to potential output growth

Although total factor productivity growth had been slowing down even before the crisis, the financial crisis had a lasting impact on the level of potential output in the euro area, and maybe also on its trend growth rate. According to the 2013 spring report of the European Commission, the decline in the level of potential output has been structural and mainly driven by increases in structural unemployment, slower growth in the population of working age and weaker investment trends. The Commission concludes that the most likely long-run impact of the crisis will be a permanent loss in the level of the euro area's output of roughly 5%, with no long-run effect on its growth rate.

Various factors could contribute to determining such an outcome. [2] For example, long spells of unemployment could result in the destruction of human capital and, in the absence of active labour market policies, a permanent deterioration in the chances of finding a new job. “Unemployment hysteresis”, which refers to the possibility that any increases in unemployment become permanent, might also be observed, in the presence of segmented labour markets, where wages are set between workers and firms and unemployed workers are not represented. [3] The ensuing fall in the natural level of unemployment will tend to reduce the productive capacity of the economy. In addition to that, depressed demand for services in a recessionary environment could slow down the adoption of information technologies in industries such as retail, business services, and finance, resulting in a persistently lower growth in total factor productivity. The ageing population of the euro area could further depress potential output, for example by lowering labour-force participation and saving rates.

This has also implications for fiscal policy. In a slow potential growth environment, fiscal deficits have to become smaller to sustain a given debt-to-GDP ratio. This may reinforce the need to adjust our social models, so that they are not financed at the expense of future generations.

Clearly, the crisis makes it even more necessary to undertake reforms that support the level and growth rate of potential output growth. Such reforms are diverse and mutually supportive. Action can be taken to improve participation in the labour market, to make tax structures supportive of private investment and to make sure that fiscal consolidation plans do not sacrifice socially profitable public investment. Total factor productivity can benefit from support to higher education and research and from educating and training the young and low-skilled workers.

All of this is essential but it is not in the hands of central banks. The specific factors I wish to explore today are the impact of a delay in the banking sector’s balance sheet adjustments on potential growth and what monetary policy can do in this process.

Following a financial crisis triggered by the bursting of a bubble, which was originally fuelled by a build-up of massive leverage ratios in the banking sector, the lack of balance sheet restructuring can indeed have severe implications for real economic activity.

A well-identified danger is the emergence of so-called “zombie banks”. These are banks which are undercapitalised but can continue to operate by failing to recognise losses on their loan portfolios. Such banks remain weak and, in the short term, overly sensitive to adverse shocks. They are a problem for regulators because they have an incentive to take excessive risks, for example, by extending loans to ex ante risky new customers. A deeper concern, however, is that zombie banks may pose risks for medium to long-term growth if they engage in the so-called “evergreening” of loans, which is often discussed with reference to Japan. [4]

The collapse of the Japanese stock and real estate markets in the early 1990s led to severe losses for Japan’s non-financial corporations, a large number of which became unable to service their bank loans. If banks had decided to call in non-performing loans, they would have been forced to write off existing capital. The fear of falling below the regulatory capital standards gave them the perverse incentive to continue extending credit to insolvent borrowers, gambling on the hope that the firms would recover or that the government would bail them out. This resulted in the evergreening of loans to troubled firms. By keeping the loans current, banks’ balance sheets also looked healthier because the troubled loans were not reported as non-performing.

“Zombie lending” is dangerous, as it can stifle the physiological progress of creative destruction which contributes to fostering productivity growth. The adverse consequences on the Japanese real economy are well documented in the literature [5]. The normal process of industrial churning was severely disrupted in sectors populated by “zombie firms”. Under normal conditions, unprofitable enterprises would have lost market value or been driven out of the market, paving the way for profitable new entrants. By guaranteeing the liabilities of banks that supported such zombies, the government effectively kept wages high and prices low, reducing the profits that new productive firms generated and distorting market competition. As a result, even solvent banks had very few good lending opportunities and the economy remained stagnant throughout the 1990s.

The risks of a lost decade in the euro area

How likely is a full-blown Japan-style scenario in the euro area?

Many underlying conditions in the economic situation in Japan in the 1990s and in Europe today clearly differ. Let me mention four of them.

First, at the aggregate level, balance sheet problems are less obvious for the euro area in contrast to 1990’s Japan, which faced severe adjustment needs in the corporate sector. For example, the ratio of the debt of non-financial corporations relative to GDP declined from 130% to just 80% in the 15 years that followed the peak in economic activity in the case of Japan. For the euro area, the corresponding ratio has hovered around much lower levels of 75-80% since 2007. The indebtedness of euro area households is also lower compared with that of Japanese households.

The second key difference between the two episodes is that the adjustment of asset prices was more subdued in the euro area as a whole. Four years after the peak in residential property prices in the euro area in 2008, the cumulative decline at the aggregate level has been around 3%. Euro area equity prices have also rebounded from their post-Lehman trough. This contrasts with the larger pre-crisis bubble in Japan, which was deflated by the collapse of equity prices persisting for the next ten years and of real estate prices, which fell by 40% in the decade after their peak in 1991.

A third major difference concerns inflation expectations and the monetary policy response to the crisis. Interest rates in Japan remained at a considerably higher level than in the euro area after the start of the corrections in stock markets and land prices. The rapid decline in inflation in Japan in the 1990s also complicated monetary policy and ultimately prompted the Bank of Japan’s announcement of a quantitative and qualitative monetary easing. By contrast, inflation expectations in the euro area have remained very well anchored close to 2%.

Fourth, the impact on the supply side of the economy was also different. Unlike Japan during the 1990s, the euro area has avoided a significant slowdown in total factor productivity since the start of the financial crisis. In Japan, average annual growth in total factor productivity [6] had fallen from 1.3% in the 1980s to 0.2% in the 1990. Based on the previously discussed estimates by the European Commission, the decline in potential growth in the euro area since 2009 has largely been a result of lower contributions from capital and labour. Total factor productivity has held up relatively well, by declining relatively mildly from 0.7% over 2000-07 to 0.5% over 2008-12.

That said, the evidence presented must not lead to complacency among euro area policy makers. First, the aggregate euro area numbers mask significant heterogeneity among euro area countries. Second, challenges to reform the governance structure of Economic and Monetary Union make crisis resolution necessarily slower and more complex than in Japan.

Now, what about the risk of “zombie banks”? The risk that a few undercapitalised banks still operate in the euro area cannot be dismissed. Indeed, the financial crisis of 2008-09 left banks with relatively low capital ratios. In some cases, this problem was tackled by means of interventions that directly sought to restore appropriate capitalisation levels through incentives to restructure bad loans and free up resources for new economic activities. In other cases, governments have reacted by delaying recapitalisation, instead issuing guarantees on bank liabilities. It has been argued that such measures tend to provide banks with an incentive to roll over bad loans and shift risks to depositors or the government. [7]

It is however unclear whether bank capital constraints may be leading to quantitatively significant evergreening of loans or excessive risk taking. Academic research has identified cases of small undercapitalised banks providing loans to risky firms after the collapse of Lehman Brothers, but these cases were isolated and the amounts involved were not quantitatively significant. [8] Today, the ratio of non-performing loans continues to deteriorate in a number of jurisdictions, but it has remained stable at the aggregate level for large and complex banking groups. Solvency positions have improved as a result of from both rising core Tier 1 capital and reductions in risk-weighted assets. The median core Tier 1 ratio of the euro area large and complex banking groups reached 11.1% in the first quarter of 2013, up from 9.6% at the end of 2011. [9]

Furthermore, the results of the ECB’s Euro area bank lending survey, showing that euro area banks started tightening credit standards as early as the third quarter of 2007, may also be interpreted as going against the hypothesis of excessive risk taking. Also, the evidence of the deleveraging of non-financial corporations during the financial crisis points against the hypothesis of the evergreening of loans.

However, the observed large exposure of banks to sovereign debt may have longer-term effects on the non-financial sector of the economy. Recent research suggests that banks with sizeable balance sheet exposures to stressed government debt at the beginning of the sovereign debt crisis have extended substantially less credit to non-financial corporations than otherwise similar non-exposed banks. [10]

Indeed, what is of concern at the moment is that a number of euro area banks seem to have intentionally sought to increase their holdings of government debt, at least in the early stages of the crisis. Banks in countries under stress have, in particular, been investing in relatively risky government bonds, using cheap short-term funding, hoping to pocket the spread between the bond return and the cost of funding on the upside. [11] This phenomenon shares unsavoury features with the behaviour of banks during Japan’s lost decade. In Japan, bank portfolios were rebalanced away from lending to creditworthy firms because regulatory requirements incentivised banks to keep lending to non-viable firms in order not to recognise credit losses. In the euro area, bank portfolios may have been rebalanced away from lending to creditworthy firms because high yields have incentivised banks to undertake long sovereign bond positions.

This sovereign-bank nexus, whereby domestic banks become the main creditor to national governments, in return increasing the government’s incentive to intervene in its banking sector, can become a threat to long-term economic recovery in the euro area. Unless both bank balance sheets and public finances are sanitised, and the link between the banks and sovereigns is broken, the euro area may indeed have a long recovery ahead of it, with insufficient loan supply to creditworthy enterprises, depressed productivity and high unemployment rates.

Policy options and concluding remarks

A number of policy actions have to be implemented in order to avert such a scenario.

The most important action is to ensure that a broad quality review of banks’ assets is carried out, so as to ensure that any hidden balance sheet losses are brought to the surface. This is one of the priorities for the new Single Supervisory Mechanism (SSM), which is being established as the first pillar of the European banking union.

The SSM could start operating in the late summer of 2014 and will directly supervise the euro area’s largest and systemically important banks. The banking union in general and the SSM in particular are expected to reduce the risks of regulatory forbearance resulting from closeness between the financial industry and local regulators. The recognition of hidden losses in banks’ portfolios, starting with the inaugural balance sheet assessment of directly supervised banks, to be performed before the start of the SSM, would eliminate the incentives to evergreen non-performing loans and eventually ensure that credit is reallocated to its most productive users. The establishment of a central supervisory body should also help break the sovereign-bank nexus by allowing liquidity and capital to be allocated freely across participating jurisdictions, and by removing incentives for banks to act as main players in the market for domestic sovereign debt. This should further free resources that can be used to finance profitable activities in the real economy.

Finally, a credible framework for resolving failing banks is necessary to eliminate the risk of a long period of economic stagnation brought about by chronically weak bank balance sheets. The Council agreement on the Bank Recovery and Resolution Directive, including the new bail-in rules, is a step forward. The SSM should now be complemented with a Single Resolution Mechanism, with a single Authority and a common financial backstop, and at a later stage by a single deposit guarantee scheme.

The establishment of the SSM is an important step forward, but potential growth-enhancing policies should also be implemented. The agenda for structural reform must be implemented with renewed impetus. Obstacles to competition in product and even more importantly in service markets should be removed so as to curtail rent-seeking behaviours and to foster productivity improvements. Labour market segmentation between highly protected workers on permanent contracts and limited protection for those on fixed-term contracts perpetrates the insider-outsider structure which was arguably responsible for the high European unemployment rates of the 1980s and should be combated.

Labour market segmentation also increases the social costs of the recession for those workers that bear the brunt of the economic adjustment process. Youth unemployment is a case in point. Before the crisis, the youth unemployment rate was twice the average rate of overall unemployment; it is now 2.6 times higher. The emergence of a “lost generation” is morally unacceptable. It would be a failure of the European project and would affect potential output in the euro area in diverse and persistent ways.

What can monetary policy do in this process? In the short term, monetary policy can avert the short-term risk of fire sales and of a freeze in specific financial market segments. After the financial crisis of 2008-09, and throughout the sovereign debt crisis in the euro area, the ECB implemented a number of standard and non-standard measures to ensure that the transmission of monetary policy stimuli to the economy would remain active, in the face of impairments in different financial market segments. This is arguably an important difference with respect to the Japanese experience in the 1990s, where commentators have argued that non-standard measures could have been implemented earlier. [12]

Looking ahead, anchoring inflationary expectations is crucial to provide a stable environment for the decisions to invest and hire workers, and to avoid the deflationary pressures which exacerbated and prolonged economic stagnation during Japan’s lost decade. The firm commitment of the ECB to its medium-term inflation objective is its best contribution to sustainable economic growth. The recent decisions of the ECB and the forward guidance it has provided on its future monetary policy stance are geared towards this objective.

I thank you for your attention.



[1]I wish to thank Georgi Krustev, Alexander Popov and Philip Vermeulen for their contributions to these remarks. I remain solely responsible for the opinions contained herein.

[2]See also European Commission (2009), “Impact of the current economic and financial crisis on potential output”, Occasional Papers, No 49; Haugh, D., Ollivaud, P. and Turner, D. (2009), “The macroeconomic consequences of banking crises in OECD countries”, OECD Economics Department Working Paper, No 683.

[3]Blanchard, O. and Summers, L. (1986), “Hysteresis and the European unemployment problem”, NBER Macroeconomics Annual, Vol. 1, pp. 15 – 90.

[4]Peek, J. and Rosengren, E. (2005), “Unnatural selection: Perverse incentives and the misallocation of credit in Japan”, American Economic Review, Vol. 95, pp. 1144-1166.

[5]Caballero, R., Hoshi, T. and Kashyap, A. (2008), “Zombie lending and depressed restructuring in Japan”, American Economic Review, Vol. 98, pp. 1943-1977.

[6]See ECB Monthly Bulletin May, 2012 on “Comparing the recent financial crisis in the United States and the Euro Area with the experience of Japan in the 1990s”. Estimates are based on data from the Japanese Cabinet Office, and Ministry of Internal Affairs and Communication.

[7]Homar, T. and van Wijbergen, S. (2013), “On zombie banks and recessions after systemic banking crises: It does matter how governments intervene”, University of Amsterdam mimeo.

[8]Albertazzi, U. and Marchetti, D. (2010), “Credit supply, flight to quality and evergreening: An analysis of bank-firm relationships after Lehman”, Banca d’Italia, Working Papers, No 756.

[9]See ECB Financial Stability Review, May 2013.

[10]Popov, A. and van Horen, N. (2012), “The impact of sovereign debt exposure on bank lending: Evidence from the European debt crisis”, ECB mimeo; Bofondi, M., Carpinelli, L. and Sette, E. (2012), “Credit supply during a sovereign crisis”, Banca d’Italia mimeo.

[11]Acharya, V. and Steffen, S. (2013), “The “greatest” carry trade ever? Understanding Eurozone bank risks”, NBER Working Paper Series, No 19039.

[12]See, for example, Ueda (2012), “Deleveraging and monetary policy: Japan since the 1990s and the United States since 2007”, Journal of Economic Perspectives, Vol. 26, pp. 177-202.

Media contacts