Outright Monetary Transactions, One year On

Ladies and Gentlemen [1],

One year after the inception of the ECB’s Outright Monetary Transactions (OMTs), financial markets in the euro area are in much calmer conditions and fragmentation is receding. Yet, OMTs have created turmoil elsewhere: in the hallways of economics faculties, newspaper commentaries and, maybe, dinner tables across Europe; today, it is not government bond spreads but emotions that are running high.

I very much welcome this debate. It is important for citizens to inform their view on monetary policy decisions, and public debate among experts can help them arrive at a balanced assessment. The euro being shared by 330 million Europeans, it is crucial that such debate brings together different European perspectives. It cannot serve the euro well to take a parochial view to its future.

I am therefore grateful to the organisers for setting up this conference and inviting me to explain the rationale and key design features of OMTs.

Let me address three issues – and come back to them in turn:

• First, the necessity of OMTs. Here, I will recall the circumstances that made OMT necessary.
• Second, the effectiveness of OMTs. I will explain why OMTs were effective from a monetary policy perspective.
• Finally, the robustness of OMTs. I will show that the design of OMTs is robust to the criticisms that have been raised in the public debate.

So let me start by addressing why OMTs are necessary. Here, two things have to be appreciated: first, all ECB’s policy actions and instruments fully serve our objective of price stability in the euro area as a whole and, second, one year ago, we were facing serious challenges in achieving precisely this objective. So, let me elaborate.

At the time when OMTs were announced in August 2012, financial fragmentation had created widely divergent borrowing costs for firms and households across euro area countries. As a result, the transmission of monetary policy was severely impaired.

First, we saw impairments to the traditional inter-temporal arbitrage mechanism, which has curtailed the effectiveness of our standard monetary policy.

Second, horizontal and spatial fragmentation – that is, impairments across banks, markets and countries resulting from structural impairments have challenged monetary policy operating with a single instrument. These dynamics were exacerbated by an adverse feedback due to the close association of banks and their sovereigns.

These two dimensions of fragmentation have been mutually reinforcing. For example, the lack of cross-border liquidity in the interbank market impaired the price-finding mechanism along the yield curve. As a result, our monetary policy impulses were not evenly transmitted across countries or adequately along the yield curve.

Furthermore, we saw the emergence of a tail-risk in the euro area, which triggered self-perpetuating dynamics in the economy. [2] A “bad equilibrium” of an adverse scenario was possible, triggered by self-fulfilling and reinforcing expectations. [3] In this adverse scenario, the expectation of one or several countries exiting the euro would have driven public and private financing costs in these countries at such a high level that they would have had no other option than to actually exit.

It is clear that the root causes of the adverse scenario were associated with substantial losses in competitiveness, lack of fiscal sustainability and financial risks in national banking sectors. But an additional factor became prominent: redenomination risk, reflecting fears of a possible imminent euro area break-up.

While exchange rate risk across euro area countries should have disappeared permanently with the creation of the single currency, there were signs, since the start of the year and increasingly so during the summer 2012, that investors had started pricing in redenomination risk. In investors’ perception, the monetary union had turned from a single currency area into a fixed nominal exchange rate system cursed with the classical “peso problem”. Investors required an interest rate premium as a compensation for the risk that the euro might not remain the irreversible currency of the euro area – at least in its current composition.

Indeed, redenomination risk was apparent in the market for government bonds. The inversion of the slope of the term structure of sovereign bond spreads observed in early summer 2012, for instance, for Spain and Italy was consistent with expectations of imminent break-up risks. Markets perceived a high probability that obligations may not be repaid in full, or equivalently be repaid in a different, lower-valued currency, which translated into commands of higher spreads. If the probability of this event concentrates over the short horizon, then the cumulative default probability for longer horizons cannot rise much further, and inversion of the spread curve necessarily follows.

Overall, in the first half of 2012, the pricing of redenomination risk led to a significant and rapid increase in spreads between government bonds in euro area countries.

These divergences could not be explained by changes in fiscal or macroeconomic fundamentals.

For example, the spreads of Spanish and Italian ten-year government bonds relative to Germany had increased by 250 basis points and 200 basis points respectively in July 2012 compared to one year before.

In neither one of the two countries, fundamentals had changed so spectacularly to justify such drastic re-pricing of sovereign risk. The Italian government had taken measures which would lead to a reduction in the deficit below the reference value of 3%. The Spanish government had just embarked on a series of reforms re-dressing long-standing problems in the labour market and in the banking sector.

Despite this, the compensation that investors were asking for the risk to hold Spanish or Italian bonds was almost twice as large as one year before.

The distortions were not limited to long maturities but observed in all segments of the yield curve. For example, as of July 2012, the spreads on Spanish and Italian two-year government bonds relative to Germany were averaging around 500 basis points and 400 basis points respectively.

Why were these sovereign bond market developments relevant from an ECB perspective?

In any economy, the government bond market plays a prominent role in the transmission of monetary policy and ultimately matters for the effective achievement of the central bank’s objective – in our case, price stability.

Changes in long-term government bond yields are an important driver of corporate bond yields and bank lending rates – either through arbitrage relations or through sovereign bonds directly serving as a benchmark for the pricing of loans and other assets.

Through this and other channels, the sovereign bond market turmoil rapidly spilled-over to banks, and ultimately it reached households and firms. As a consequence, the entire economy’s lending conditions deteriorated significantly in the countries under stress.

The implications of the crisis were not confined to countries under stress but also visible in core euro area countries. With fears of a euro area break up growing, core euro area economies became the destination of safe haven flows.

Liquidity was suddenly drying up in stressed countries, threatening to lead to a destructive downward spiral of abrupt deleveraging, fire sales and ultimately deflation. By contrast, there was a risk that abundant liquidity in core countries would unduly drive down financing conditions and fuel asset price bubbles. In fact, as market interest rates were surging in countries under stress, they declined in core countries. German two-year government bond yields turned negative in early July 2012.

Deposit outflows and augmented recourse to central bank borrowing as a substitute for the lack of alternative funding were reflected in intra-euro area payment imbalances. Target balances, an indicator of euro area fragmentation and capital market malfunction, reached an all-time high at levels above 1,000 billion euro in the summer of 2012, even though current account imbalances were already being reduced.

In this challenging environment, we were losing the steering ability of our monetary policy. The single monetary policy had shifted from “one size fits all” to “one size fits none”.

Let me make a general remark at this stage. How should central bankers act in such an environment? Policy makers have always the option of doing nothing. Within their mandate, they can choose responding actively to the crisis by rethinking and potentially revising the appropriate set of policy instruments, or, alternatively, they can be passive. Being passive is, however, not an innocent policy stance, but a policy choice itself with potentially harmful implications for price stability. As John F. Kennedy once observed, “There are risks and costs to action. But they are far less than the long range risks of comfortable inaction.” And in the ECB case, passivity is not an option if it means failing to reach our objective of price stability. It is our unequivocal duty, enshrined in the Treaty, to do whatever it takes within our mandate to fulfil our objective.

So how should active monetary policy being conducted in such an environment, within the central bank’s mandate?

First and foremost, monetary policy should aim at repairing the transmission of monetary policy by reducing fragmentation, insofar it is not related to structural causes.

In this respect, it has been argued that reducing fragmentation can imply a redistribution of risk in times of crisis. [4] For example, central banks can provide insurance policies against a tail event associated a catastrophic scenario. The main insight here is that redistribution of risk is not a zero-sum game, but that the overall risk in the economy, in our case in the monetary union, can be reduced. I agree with this view, but I would also like to stress that any such insurance provided by the central bank should come with appropriate safeguards to mitigate moral hazard. [5] I will come back to that point later.

As our main policy response to the crisis, we announced our decision to implement OMTs in secondary government bond markets. Let me be clear: OMTs are not to interfere with the pricing of sovereign bonds on the basis of economic fundamentals and the respective credit and liquidity risks of the sovereign. The goal of OMTs is a narrow one: to eliminate unwarranted and self-reinforcing fears of a euro area breakup that have undermined our ability to effectively conduct monetary policy in the pursuit of price stability. In this respect, OMTs are an instrument tailored to the multiple equilibrium problem outlined above.

Why OMTs are effective

Have OMT been effective? OMTs were able to address the impairments to the transmission mechanism of monetary policy, by reducing fragmentation and restoring the distributional neutrality of monetary policy. It has eliminated fears of disaster events and removed denomination risk from the market.

How was this possible without spending a single euro? OMTs are an insurance device against redenomination risk, in the sense of reducing the probability attached to worst-case scenarios. As for any insurance mechanism, OMTs face a trade-off between insurance and incentives, but their specific design was effective in aligning ex-ante incentives with ex-post efficiency. I will come back later to the reasons why OMTs are incentive-compatible.

So, what is the evidence that OMTs have been effective? One year after the announcement, the positive effects of OMTs are visible in several key indicators.

First, distortions on the sovereign debt markets have receded. Compared to their peak before the OMT announcement spreads on ten-year government bonds for Italy and Spain, for example, are now back around levels observed in the summer of 2011. Moreover, CDS spreads have gone down in all countries.

Second, bank and firm borrowing conditions have relaxed. Corporate bond spreads in stressed countries have declined substantially, affording creditworthy companies better access to market funding. This has been observed not only for highly rated large corporations but also for other corporations, for the financial as well as the non-financial sector.

Third, banks have been able to re-access the market, for both funding and raising capital. Also, the strong divergence in funding costs across countries has fallen. Deposits outflows have been reversed: deposits by euro area residents at banks in stressed countries have increased by about 210 billion euros since August 2012.

These improvements in the form of reduced fragmentation can be summarized by the downward trend in Target balances. These have been reduced by about 300 billion euros or some 30% decline from their peak one year ago. They have now returned to levels observed before the two three-year LTROs at the end of 2011.
In sum, these indicators point to significant benefits from OMT for corporations, banks as well as individuals in all euro area countries.

Fears of a possible euro area break-up have receded and impairments to the transmission mechanism have been partly reduced.

Why OMTs are robust

Does this favourable market response mean that OMTs do not raise concerns? The answer is: no, but these concerns have been intensely discussed in the Governing Council, and adequately mitigated.

The relaxation of financial turmoil is, in itself, not sufficient for a positive assessment of OMTs.

In fact, for OMTs to be successful, they have to not only eliminate risk in the market, but they have to do so in a controlled and incentive compatible way.
It has been argued that:

• first, government bond purchases undermine incentives for reform at national level;
• second, government bond purchases expose the ECB to balance sheet risk that may eventually lead to fiscal redistribution among euro area countries; and
• and third, government bond purchases violate the monetary financing prohibition because it directly affects the conditions at which governments can issue debt.

I don’t address here a fourth objection, namely that buying government bonds would be conducive to a higher inflation. The ECB has all instruments at hand to retain control of its monetary stance, whether OMTs are activated or not.

Let me explain how OMTs address each of these concerns.

Incentives for reform

The argument on incentive compatibility goes as follows: sovereign bond markets act as a disciplining device; if governments adopt imprudent economic policies they face higher interest rates in the market; as a consequence, they are forced to take corrective action. [6]

But the incentive for such corrective action is undermined if the central bank shields governments from market pressure. In this case, governments can, to some extent, choose whether they adopt painful fiscal and structural adjustment – or delay their reform efforts and just count on further support from the central bank.

On a general level, this argument is convincing. Certainly, the euro area needs an active and freely functioning government bond market, and lack of market discipline was one of the failures that have led to the crisis.

But when applying it to the case at hand, one should not ignore the most important feature of OMTs, namely: its explicit link to policy conditionality.

This feature distinguishes OMTs from other historical episodes of central bank intervention in government bond markets. And it renders the argument on incentive incompatibility invalid.

This is because conditionality removes the privilege for governments to choose between economic adjustment and central bank intervention.

Even if the ECB were to decide - at its full discretion - to buy bonds under OMTs, governments will have to continue their reform efforts as required by the respective ESM programme and by IMF involvement. Otherwise, they would simply become ineligible for OMTs. Hence, no reforms, no OMTs.

In fact, the policy conditionality not only aligns incentives in the countries already subject to an ESM programme. It also aligns incentives in countries that are at risk of requiring support.

This is because, in the programme context, national authorities face a substantial loss of sovereignty with regard to economic policy. The political cost related to this loss of sovereignty constitutes a forceful deterrent.

Indeed, the experience with existing EU/IMF programmes shows that governments request official assistance only if it is strictly unavoidable through own efforts.

In other words, there is no incentive for governments to wait and do nothing until they move into an ESM programme. By contrast, they have an incentive to put all their efforts into solving their problems without having to request a programme.

Besides, purchases under OMTs are limited to shorter dated bonds, namely, up to the three year tenor. Government yields at longer maturities would be entirely determined by market forces, and at shorter maturities, fear of the programme running off-track and OMT eligibility being lost would be reflected in market prices.

Overall, these considerations demonstrate that OMTs do not eliminate the market as a disciplining device for governments. What OMTs do is to complement the market, which in the current crisis has been abrupt and unreliable, by another incentive mechanism.

Risks to the ECB balance sheet

Let me turn to the second argument, which is related to the ECB’s risk exposure. Thus far, we are in the comfortable situation that desirable effects of OMTs have materialised without a single euro being spent. In fact, risks to the balance sheet of the ECB have vastly decreased since mid-2012, thanks to the convergence in yields and marked decrease in central bank borrowing.

But indeed: once we were to conduct bond purchases under OMTs, the Eurosystem would assume the concomitant balance sheet risk.

Under an extreme scenario with the default of the sovereign concerned, the resulting losses would have to be shared across the Eurosystem. Ultimately, this could indeed constitute a transfer of risk among euro area countries.

Let me make three remarks on this:

First, it is precisely the existence of OMT that contains the risk of such catastrophic scenario to materialise, since:

• by its very existence OMT eliminates the undue spread components and hence self-fulfilling default trajectories.
• through conditionality OMT fosters incentives for sound economic and fiscal policies.
• and by focusing on bonds with shorter residual maturity, it limits the duration of the risk exposure to a given creditor.

Second, any risk comes with a reward. Under a successful OMT programme, interest payments on the bonds purchased by the ECB amount to a transfer from the benefiting country to the ECB, then to the national central banks.

Third, and more fundamentally, one can argue that a certain transfer of risk among different geographical areas is inevitable and inherent in the implementation of a single monetary policy for any currency area. Indeed, for the euro area geographically diverse risk exposure is not new and has always been part of the way we conduct and implement monetary policy.

Think, for example, about our main instrument for monetary policy: collateralised liquidity provision via our repo operations with banks.

Here, the decision on the geographical allocation of liquidity across countries is not at the discretion of the ECB: it is essentially market-driven and depends on the liquidity needs of the banks. Hence the transfer of risk across countries is inherent.

Moreover, the transfer of risk is inevitable because a decentralised allocation of liquidity is a pre-condition for achieving our mandate of price stability. If we were to impose a specific distribution of liquidity across countries we would have to renege on our mandate. In sum, and as discussed above, monetary policy can be redistributive provided that it aims at restoring distributive neutrality.

For OMT, it is in essence the same as for our standard framework.

The transfer of risks is inherent to the motivation of OMT because sources of impaired monetary policy transmission are unevenly distributed geographically.

It is inevitable because imposing constraints on OMT due to risk sharing considerations would weaken its effectiveness and hamper the ECB in its pursuit of price stability.

Let me say a more general word on the distributional neutrality of monetary policy [7]. Central banks are untrusted with a narrow mandate and, being independent, they are shielded from representative politics. They should refrain from engaging in income redistribution, which should be sanctioned by Parliaments. This does not imply that monetary policy actions do not have distributive consequences – in fact, they always have. But these are side-effects of a strategy that aim at ensuring price stability, which is by essence neutral as regards income distribution. OMTs are just an aspect of such a strategy. The emergence of redenomination risk resulted into a situation where monetary policy decisions were not evenly transmitted to specific market segments and regions in the euro area. OMTs supported the distributional neutrality of our monetary policy by addressing the underlying impairments.

The prohibition of monetary financing

A final concern that has been raised regarding OMTs is whether it violates the monetary financing prohibition. Let me focus on the economic aspects.

The economic rationale of the monetary financing prohibition is clear: as history has shown, central banks cannot ensure price stability if they have to permanently make up for weak performance in other policy domains. Article 123 TFEU transports this historical lesson into the architecture of EMU.

The aim and design of OMTs are in full respect of this economic requirement. To see this, we again have to recall the concrete design elements of OMTs.

The first design element that’s relevant here is, once more, the link to conditionality.

As mentioned before, the link to policy conditionality of an ESM programme ensures that central bank intervention under OMTs does not replace reform efforts by other policy domains.

Instead, OMT can – by design – only be a complement to national reform efforts.

Hence, the design of OMTs clearly prevents a scenario of harmful central bank support, or fiscal dominance over the central bank that has motivated the monetary financing prohibition; that is: the central bank intervenes, although – or even because – other economic policies fail to deliver.

The second element to keep in mind is the rationale of OMT.

To be clear: OMT would never be used to indiscriminately push down government bond spreads. By contrast, spreads should continue to reflect the underlying country specific economic fundamentals, fiscal positions and market risk perceptions that incentivise governments to engage in sustainable fiscal spending and competitiveness enhancing structural reforms.

Hence also on this count, OMT does not correspond to the harmful scenario that motivated the monetary financing prohibition; a scenario, in which the very purpose of central bank intervention is to reduce yields below the fundamentally justified level so as to preserve debt sustainability despite weak policy performance.

The aim of OMTs is the exact opposite: a deterioration of fundamentals would induce the ECB to lower or even stop OMT intervention. OMTs would only aim at that portion of bond yield spreads that are not fundamentally justified and based on undue risks of a euro area break up.

As I have said elsewhere [8], central bank independence and a clear focus on price stability are necessary but not sufficient to ensure monetary dominance. The fiscal authority must be ready and willing to adjust its revenues and primary spending to stabilise its debt at any level of the interest rate that the central bank may choose. In academic parlance, for monetary policy to remain active, fiscal policy needs to be passive, or “Ricardian” [9].

By creating the right environment, and providing appropriate incentives, for governments to take action to ensure fiscal solvency, OMTs create the conditions to affirm the monetary dominance regime which is at the heart of the Treaty.

Conclusion

Let me just conclude, by repeating: OMTs are necessary from a monetary policy perspective; they were effective in counteracting the destructive market turmoil in the summer of last year; and they are designed in a way that overcomes the key concerns associated with government bond purchases by the central bank.

OMTs are not just words: the ECB is fully prepared to use them. But even if euro area member states comply with the conditions, there is no automatism to activate OMTs. For the Governing Council of the ECB, there is only one single deciding criterion and that is the maintenance of price stability in the euro area. This overarching criterion has shaped the design of OMTs, and it will decide over its potential activation.

Thank you very much for your attention.

[1] I am grateful to F. Holm-Hadulla and A. Saint-Guilhem for their contribution to this speech; all views expressed remain mine.
[2]This resembled the environment described in the academic literature on second generation currency crises, see M. Obstfeld (1986), “Rational and self-fulfilling balance-of-payments crises“. American Economic Review 76 (1), pp. 72-81.
[3]See G. Corsetti and L. Dedola, 2013, “The Mystery of the Printing Press: Self-fullfilling Debt Crises and Monetary Sovereignty”, CEPR Discussion Paper No. DP9358.
[4]See M. Brunnermeier, and Y. Sannikov, 2012. “Redistributive Monetary Policy”, paper prepared for the 2012 Jackson Hole Symposium, Princeton University.
[5]See B. Coeuré, 2012. “Central banking, insurance and incentives”, speech at the ECB conference on “Debt, Growth and Macroeconomic Policies” Frankfurt, 6 December.
[6]This line of reasoning has also received empirical support. See, for example, O. de Groot, Holm-Hadulla, F. and Leiner-Killinger, N. (2012). “Cost of borrowing shocks and fiscal adjustment”. ECB Working Paper No. 1503.
[7]For a detailed discussion, see B. Cœuré (2013), “Monetary policy in a fragmented world”, speech at the 41st Economic Conference of the Oesterreichische Nationalbank. Vienna, 10 June.
[8]See B. Cœuré (2012), “Challenges to the single monetary policy and the ECB response”, speech at the Institut d’études politiques. Paris, 20 September.
[9]See E. Leeper (1991), “Equilibria under ‘active’ and ‘passive’ monetary and fiscal policies”, Journal of Monetary Economics, 27, pp. 129-147; and M. Woodford (2001), “Fiscal requirements for price stability”, NBER Working Paper, No. 8072.