Peter Praet: The European Central Bank's monetary policy - past and present
Speech by Mr Peter Praet, Member of the Executive Board of the European Central Bank, at the Febelfin Connect event, Brussels/Londerzeel, 16 March 2017.
The views expressed in this speech are those of the speaker and not the view of the BIS.
In recent years it has been often said that the ECB has become the "only game in town" for stabilising the euro area economy. As governments have had to consolidate their fiscal positions, there has been an unprecedented onus on monetary policy to support aggregate demand. The ECB has responded to this challenge and acted decisively to secure price stability in face of an economic and financial crisis unparalleled in post-war history.
But this role, and the unconventional measures we have adopted to execute it, has inevitably put monetary policy more in the spotlight. We are facing intense scrutiny as to how our policy works, its necessity and the side effects it causes.
So what I would like to do in my remarks is, first, to explain the current monetary policy of the ECB and its evolution during the crisis. The unconventional measures we have deployed, such as negative rates and asset purchases, are the outcome of a protracted crisis that has unfolded in multiple waves - one that has both worsened the economic situation facing monetary policy and constrained the instruments we have available to respond to it.
I will then go on to describe how these unconventional measures are working to ease financing conditions for firms and households, especially through the bank lending channel, and ensure a sustained recovery for the euro area economy. Finally, I will touch on the side effects of our measures on the banking sector, and how they can be addressed to ensure a continued, robust transmission of our monetary policy.
While my focus will be on what monetary policy has achieved during the crisis, it is also important to underline what we cannot do: central banks cannot remain the "only game in town" indefinitely. Monetary policy can bring output back to its potential level, and it may even be able to affect that potential by unwinding hysteresis effects, but it cannot durably raise long-term growth - which was on a declining trend even before the crisis. That requires further, determined progress with structural and institutional reforms.
The phases of the crisis
In 2008 the global economy faced a crisis caused by the coincidence of two trends.
The first was the bout of over-optimistic expectations which took hold in several advanced economies in the pre-crisis years, reinforced in the euro area by a renewed sense of security and economic prosperity following the launch of monetary union. Despite slowing potential growth, agents in a number of economies overestimated their future income and borrowed against it, accumulating excessive debt. In some countries this over-leveraging was centred on firms, in other countries on households and in others still on the state. Banks acted as the intermediaries in both creditor and debtor economies.
This new environment would probably not have produced such aggressive risk-taking, however, had it not been for the second trend: the parallel process of financial liberalisation and deregulation across advanced economies which encouraged financial overextension and rapid credit growth. In the euro area, financial integration also triggered an expansion in cross-border financial flows, adding to the forces created by the waxing global financial cycle. The consequent misallocation of resources into unproductive sectors only worsened the underlying growth picture.
So when this reality finally bit in 2008, monetary policy was left facing a uniquely challenging set of conditions: an "expectations gap" which had left firms and households overleveraged and in urgent need of balance sheet repair, coupled with a "credit gap" as overextended banks began their necessary restructuring and deleveraging - in other words, a so-called balance sheet recession. These conditions laid the ground for the subsequent evolution of the crisis in the euro area, which has unfolded in three main phases, each of which has required a different monetary policy response.
The first phase was the immediate liquidity crisis triggered by the turning of the global financial cycle and the subsequent collapse of Lehman Brothers. As in other advanced economies, euro area banks faced uncertainty about the scale of their exposure to subprime products, prompting a sudden stop in the availability of money market funding for many financial institutions and a systemic rush to cash. Central banks faced an imminent risk of asset fire sales by credit institutions to meet maturing obligations, in the process depleting their balance sheets and initiating a steep contraction in credit.
The ECB's response was to lower its main refinancing rate to the then record low of 1% in May 2009 and to provide liquidity elastically to the banking sector, at both increasingly long durations and against a wider range of collateral. Our balance sheet expanded as central bank liquidity substituted for the malfunctioning interbank market.
But though monetary policy was able to avert the collapse of the financial system and prevented a much deeper recession, the Lehman episode nonetheless left deep scars on the banking sector, which the euro area was unable to heal as swiftly as other jurisdictions. Crucially missing was a truly integrated banking system that could share risk across countries, with a harmonised resolution framework and a European public backstop. Thus the onus fell onto sovereigns to support their national banking sectors, strengthening the two-way risk between European banks and their national governments.
This created the conditions for the second phase of the crisis: the sovereign debt crisis of 2011-12 and its amplification through the "bank-sovereign" nexus. As sovereign borrowing costs spiked, banks' exposures to selected governments came under intense market scrutiny. The banking systems in vulnerable countries, representing almost 35% of euro area GDP, lost access to affordable wholesale funding. Intermediation retreated behind national borders, and with perceptions of redenomination risk on the rise, financial markets underwent severe geographical fragmentation.
The result was a serious disruption in monetary transmission across different parts of the euro area, which prevented our intended accommodative policy stance from reaching the economy. Hence we responded to this fragmentation in two ways.
First, as per the Lehman shock, by compensating for sources of market funding that had suddenly become unavailable, which now also included bank bonds. We achieved this by lengthening the maturity of our operations to three years in November 2011, as well as by further expanding the range of eligible collateral that could be pledged against central bank money. Without such action there would, once more, have been a risk of banks being forced into fire sales and rapidly contracting credit to the economy. Generalised capital controls in vulnerable countries would have been inevitable.
Second, by removing the redenomination risk expressed in sovereign spreads, which was distorting the local pricing benchmarks used to price assets and credit. Our main tool here was the announcement of our conditional Outright Monetary Transactions in September 2012, which acted as a powerful circuit breaker against self-reinforcing fears in sovereign bond markets. This successfully truncated the worst tail of the distribution of possible macroeconomic outcomes. But the sovereign debt crisis nonetheless left a damaging legacy for the economy - and this laid the ground for the third phase.
The ECB's measures since summer 2014
As the euro area entered a prolonged slump, banks in large parts of the euro area began to engage in a drawn-out process of deleveraging. This represented a necessary - and indeed overdue - process of adjustment for banks that had overleveraged in the run-up to the crisis. But rather than reducing risk-weighted assets by raising capital or disposing of non-core assets, banks instead began to reduce lending. In those countries most severely hit by the sovereign debt crisis, the risk-adjusted returns on loans had declined to the point where banks had a strong incentive to redeem loans and downsize.
The euro area faced a chronic malfunctioning of bank-based transmission and growing signs of a credit crunch: by end-2013 the annual growth rate of loans to the private sector had contracted by more than 2%. Against that background, inflation began a prolonged downward drift. The ECB cut rates twice in 2013 in response, reaching 0.25% in November, and introduced forward guidance to better insulate euro area financial conditions from the diverging US monetary policy cycle. This policy helped decouple the risk-free curve from outside influences and made it more appropriate for the underlying conditions we were facing. In early 2014 the recovery appeared to be back on a firming path.
By mid-2014, however, it was clear that the economic recovery had lost momentum, removing a key driver of the reflation scenario we had anticipated. The sharp fall-off in oil prices that began in late summer added further disinflationary pressures. Given the weak underlying trend in inflation, we saw a growing risk that low inflation could de-anchor inflation expectations and produce second-round effects, creating a self-reinforcing negative spiral. Without decisive monetary policy action, the euro area could have faced deflation.
The ECB therefore faced a pressing need to reinforce the degree of policy accommodation in view of persistently weak inflation. Yet here we confronted an instrument limitation.
With interest rates already not far from zero, our ability to use conventional monetary policy to provide additional accommodation was now constrained. And moreover, the pre-crisis slowdown in potential growth, coupled with the negative effects of the crisis on investment and private sector balance sheets, appeared to have imbalanced saving and investment, and depressed the real equilibrium interest rate to very low levels. Hence even cutting rates to low (positive) levels would provide insufficient stimulus.
Under these conditions, the ECB had to resort to a new approach to expand its monetary stance, one based less on adjusting its main refinancing rate, and more on directly influencing the whole constellation of interest rates in the economy that are relevant for private sector financing conditions. This strategy has been articulated in three instruments.
The first is the negative interest rate on our deposit facility, which - in a context of excess reserves - has brought overnight rates down to negative levels and hence provided additional effective stimulus. The fact that policy rates can indeed turn negative has also contributed to flattening the short to medium end of the yield curve, thereby easing broader financing conditions by removing the upward bias to yields that comes from the perception that rates can only go up, not down.
The second instrument is the asset purchase programme (APP) of private and public sector securities, which has helped further depress the term structure of interest rates by compressing risk premia out along the yield curve. This effect takes place directly in the market segments where we intervene, namely covered bonds, asset-backed securities (ABS), sovereign and corporate bonds. And higher asset prices spill over, through portfolio rebalancing, into non-targeted markets as investors move up the risk and maturity ladder.
Importantly, the stimulus from these two instruments reaches the economy independently of whether financing is dominated by banks or capital markets. A broad easing of financial conditions leads to lower funding costs for banks, while asset purchases depress the pricing kernels used in each country to price local credit, in particular the risk-free yield curve. Both lead to lower interest rates for borrowers. At the same time, as asset prices rise banks have a stronger incentive to provide loans to the real economy in the first place.
Indeed, banks operate on the basis of the same relative-value logic as non-bank market participants: a decline in fixed-income securities earnings encourages banks into the same type of portfolio reallocations that affect investors in the capital market, rebalancing their portfolios towards assets with higher risk-adjusted returns, including loans. And this also applies to the excess cash reserves that are created as result of our asset purchases, where the incentive to invest in higher-yielding assets is intensified by our negative rate policy.
The third instrument is our targeted longer-term refinancing operations (TLTROs), which are specifically aimed at galvanising bank lending to the private sector in a context of ongoing bank deleveraging. Banks have been able to borrow at the interest rate on our deposit facility, but only on condition that they demonstrate strong performance in loan origination. The intention was to introduce more competition in the market for bank loans, which in turn squeezes unit lending margins and borrowing costs for the real economy.
This set of measures - which works as a mutually reinforcing package - has proved adaptable to the shocks the euro area has faced since its launch. Indeed, the third stage of the crisis has not only included the disinflationary consequences of the sovereign debt crisis, but also an unprecedented collapse in energy and commodity prices, with its potentially destabilising effect on inflation expectations; and a slowdown in emerging market economies in 2015 which weighed on global trade. In each case we have been able to recalibrate our measures in terms of size and across instruments so as to continue providing sufficient accommodation to the euro area economy.
The impact of our measures
The evidence for the effectiveness of this strategy is clear. Our monetary policy has contributed to a major easing of euro area financing conditions and, through this channel, to a more robust and sustained economic recovery.
Since mid-2014 bank lending rates have dropped to historical lows for both firms and households, and pass-through regularities appear to have returned. Bank lending volumes have also been on a continuous upward trend, though they remain relatively weak: in January this year lending to households grew at the fastest rate since May 2011 - 2.2% on an annualised basis - while corporate lending, as in December, grew by 2.3%, its highest reading since 2009. And behind this picture of overall credit easing is one of substantial convergence - both inter-country and intra-country.
Granular data confirms that both the level and dispersion of lending rates has declined considerably in vulnerable countries and more strongly than elsewhere, thereby further reducing asymmetries in the overall policy transmission across country groups. Rate dispersion between large and small loans also largely vanished between vulnerable and other countries towards the end of last year, implying a convergence in financing conditions among different types of firms, especially for small and medium-sized enterprises (SMEs).
Indeed, our latest survey on access to finance by SMEs shows further improvements in the availability of external sources of finance and the willingness of banks to provide credit at lower interest rates. "Access to finance" is now considered the least important concern for SMEs, admittedly with some divergence across countries. SMEs in Greece continued to perceive it as a very important problem but, encouragingly, a score close to the euro area average was reported by SMEs in Italy, Ireland, Portugal and Spain.
Our measures have helped induce this broad easing of borrowing conditions very much through the channels we expected.
On the liability side, bank funding conditions have eased considerably, with the average cost of funding in the euro area reaching historical lows of around 40 basis points. Very low short-term rates have encouraged banks to rebalance their liability structures away from more expensive debt securities and towards deposits, providing funding cost relief. But there are of course limits to such substitution - namely the incoming requirement to hold a minimum amount of eligible liabilities and own funds (TLAC/MREL) - so it is important that wholesale funding costs have also come down substantially. Our outright purchases under the APP have tightened spreads on covered bonds and ABS, while rates on senior unsecured bank bonds have fallen due to the portfolio rebalancing effects of the APP.
This positive effect of monetary policy on wholesale funding conditions is confirmed by the Bank Lending Survey (BLS). Close to two-fifths of banks surveyed in the October 2016 BLS reported a net improvement in their market financing conditions owing to the APP - up from one-fifth in the April survey - in particular for financing via covered bonds and unsecured bank bonds, but also via ABS issuance.
The TLTROs, too, have directly lowered term funding costs, especially for banks in vulnerable countries, since those that draw on these operations have been able to substitute more expensive wholesale debt for longer-term central bank liquidity. 40% of the banks surveyed in the latest BLS round reported that they have used TLTRO funds for substituting maturing debt. And the incentives embedded in the TLTROs have ensured a pass-through to customers: ECB analysis confirms that TLTRO bidders in vulnerable countries have lowered lending rates more than their non-participating peers.
On the asset side, the flattening of the term structure of interest rates seems to have exerted strong downward pressure on bank lending rates, especially in countries that were more affected by the sovereign debt crisis. ECB pass-through models show that, since the launch of our credit easing package, the decline in lending rates in Spain and Italy in particular has been much stronger than can be explained by falling market rates alone. Tightening sovereign spreads help explain much of the observed fall.
This effect, paired with the credit impulse provided by the TLTROs, has also helped boost competition among banks and put further downward pressure on lending rates. Evidence from the BLS confirms that rising competitive pressures have been a key contributor to margin compressions over the last two years. Purchases of corporate bonds are also a factor: banks facing substitution of bank loans by corporate bonds have reported higher competitive pressures and stronger margin compressions on loans to large firms.
There is some evidence, too, that as prices have risen banks have rebalanced their portfolios away from sovereign bonds and towards loans. The increase in bank credit to the private sector has been counterbalanced by a decrease in credit to general government. Banks responding to the BLS last year indicated that they used the additional liquidity from APP-related sales for granting loans, potentially responding to the incentives created by our asset purchases and negative interest rate policy.
ECB staff research finds that banks' reactions to holdings of excess liquidity have changed due to the charge on the deposit facility: cumulated bank lending in less vulnerable countries has been one percentage point higher than it would have been otherwise.
As borrowing conditions for the real economy have loosened, economic activity has followed. Whereas the 2010-11 recovery was based almost exclusively on net exports, the euro area is now experiencing a recovery based predominantly on domestic demand. In fact, growth in the euro area was previously closely correlated with the strength of international trade, but that relationship has weakened recently. Monetary policy is contributing to higher domestic absorption through all three classical channels.
The strong recovery in consumption has been bolstered through the income channel, since falling interest rates have supported net borrowers, who tend to be liquidity-restrained and have a higher marginal propensity to consume. Expansionary effects are also visible through the intertemporal substitution channel: in particular, easier borrowing conditions have encouraged households to bring forward durable consumption, and firms' investment, through credit.
Consumption of durables has rebounded in recent years and its growth, especially in countries where credit was previously very tight, is closely correlated with the improvement in credit conditions as recorded in the BLS. Investment has responded more slowly to low interest rates, yet monetary policy is helping create the conditions for a stronger investment revival through its overall effect on the economy, sometimes called the "accelerator effect". Production and selling price expectations are rising as aggregate demand picks up; corporate profitability is increasing; and financing conditions remain extremely favourable.
Finally, housing markets are now moving into an expansionary phase in most euro area countries, which should also support consumption going forward via the wealth channel.
As a result, we now see the economic recovery continuing at a moderate, but steadily firming pace, and broadening gradually across sectors and countries. Sentiment indicators suggest that the cyclical recovery may be gaining momentum. Consequently, our latest staff macroeconomic projections foresee annual real GDP increasing by 1.8% in 2017, 1.7% in 2018 and 1.6% in 2019 - a slight upward revision this year and next compared with the previous exercise. The risks surrounding the euro area growth outlook have also become less pronounced, although remain tilted to the downside due to global factors.
Implications for the financial sector
So there is no doubt that our measures are having their desired effect. Yet unconventional monetary policy can have side effects as well, and especially for the financial sector.
One key concern is their impact on banks' profitability - which remains generally weak in the euro area - and especially on net interest income. Asset purchases flatten the yield curve and reduce the returns on maturity transformation; and the negative interest rate policy de facto flattens the curve further, since the zero lower bound that banks typically apply on deposits means that a large share of bank liabilities no longer reprice when market rates fall. Bank lending rates however typically do, compressing loan-deposit margins.
Consistent with this, margins between loan and deposit rates on new business narrowed in 2016 and may tighten further as loan rates continue to decline but deposit rates increasingly stack up against zero. More than half of the new deposits held by corporations and close to 40% of those held by households are de facto no longer remunerated. Slightly more than 40% of the stock of corporate deposits and 35% of household deposits have virtually zero return.
This is clearly a challenging situation for banks, but so far its impact on profitability has been contained, since monetary policy has countervailing effects on bank balance sheets. ECB analysis suggests that lower net interest margins are being partly offset by higher asset valuations and a more robust economy - which in turn boosts lending volumes and leads to lower provisions and impairments for banks. And more generally, our policy measures create a faster recovery with higher inflation, which implies a faster return to monetary policy normalisation. Such expectations are now one factor buoying bank share prices.
Still, given the dominant role of net interest income in banks' earnings and the staggered pass-through of low rates into lending margins, profitability is likely to remain vulnerable for some years to come, which raises the question of how the banking sector should best adapt. In my view, it would be misguided for banks to wait for changes in monetary policy to come to the rescue. Even in a normalisation scenario, the pre-crisis banking world is unlikely to return. Banks face new structural and technological challenges that they have to confront, in particular in terms of raising operational efficiency.
Though the efficiency of the euro area banking sector has improved since the crisis, operating costs remain generally high. This is especially apparent when one compares the cost-to-asset ratios of some national banking sectors with those of banks in the Nordic countries. Sweden and Denmark, for example, have cost-to-asset ratios of around 1, whereas in France and Germany the ratio is closer to 1.5 and in Italy nearly 2. Such cost structures represent vulnerability in an environment where traditional bank business models are under threat, and where banks are facing increasing competitive pressures from new fintech entrants, especially in areas such as retail payments, where they previously enjoyed healthy margins.
The reasons for these differences are complex, but they are partly the result of overcapacity in some national markets, leaving banks with high fixed costs. Hence rationalisation and consolidation within the sector appears a key ingredient in relieving cost pressures. While there has been some movement in this area since the crisis - the number of credit institutions in the euro area fell by 19% between 2008 and 2015 and the number of branches declined by 16% - concentration indices in some markets remain low, and consolidation has largely taken place within countries, reducing the benefits of cross-country diversification.
This is concerning, since it suggests that euro area banks are not yet in a position to exploit the economies of scope and scale that should be possible in a truly integrated banking sector, leaving a banking landscape that remains fragmented along national lines. And despite the launch of banking union, this trend is actually worsening: cross-border mergers and acquisitions accounted for just 9% of total transactions in 2016, against 15% for the period 2000 to 2015.
The still-unfinished state of banking union is one reason why. Two factors make it difficult for banks to execute a European balance sheet strategy. First, the end scenario for banks in terms of public risk-sharing is not yet clear, not least for deposit guarantee schemes. Second, the euro area is not yet fully treated as a single jurisdiction for the purposes of bank regulation and supervision, which creates barriers to cross-border activity.
The latter means, for example, that cross-border euro area exposures are still considered as international exposures from a regulatory perspective, which may lead to additional capital charges for systemically important banks. Banking groups seeking to optimise their liquidity and capital within the euro area also face frictions to fungibility, be it liquidity waivers or requirements to relocate staff if different parts of the balance sheet are reshuffled among subsidiaries. All this reduces the gains to operating as European banks.
What I would ideally like to see - as a monetary policymaker - is a banking union that allows banks to operate as truly European entities, and banks taking advantage of the cost-cutting opportunities this provides. This would guarantee a healthier, more profitable sector that can transmit our monetary policy smoothly to the real economy across the cycle. And it would increase macroeconomic risk-sharing, leading to a more resilient monetary union.
The extent to which we are able to achieve such a new banking landscape is, in my view, the litmus test of whether we have completed a genuine banking union.
Looking forward
Faced with a prolonged crisis, the ECB's unconventional policy measures have been essential to provide additional accommodation to the economy and prevent a self-sustaining fall in inflation - and they have been a clear success. Easier credit conditions have fed into a domestic demand-led recovery that has spread across countries and sectors. The economic outlook today is now better than it has been for many years.
This assessment was confirmed by the Governing Council at its last monetary policy meeting. Compared with our previous forecast round in December, we are now more confident in the outlook, which is reflected in both the upward revision to the growth path and the upward shift in the balance of risks surrounding that path. Accordingly, we are no longer concerned about deflation risks, which is why the reference to "act[ing] by using all the instruments available within its mandate" was removed from the introductory statement. We no longer perceive a sense of urgency to take further measures to combat adverse tail risks.
Nevertheless, in important ways our assessment remains unchanged. The economic risks, though becoming less pronounced, remain tilted to the downside. And crucially, we have not yet seen the firming recovery translate into a durable strengthening of inflation dynamics. Headline inflation has increased, but largely on account of rising energy and food price inflation. Underlying inflation pressures continue to remain subdued.
Our latest projections foresee inflation moving towards levels below, but close to, 2% over the forecast horizon. Given the softness of underlying inflation, however, we cannot yet be sufficiently confident that inflation will converge to levels consistent with our aim in a durable manner. Inflation dynamics also remain reliant on the present, very substantial degree of monetary accommodation, so they have not yet become self-sustained.
Against this background, the Governing Council reaffirmed its monetary policy stance, including its forward guidance, in order to secure a sustained convergence in inflation towards our aim and insure the recovery against the downside risks that still persist.