The following is an English translation of the Bundesbank's letter to the German Constitutional Court that was published by Handelsblatt. We welcome comments on translation issues and otherwise.
Bundesbank December 21, 2012
Statement to the Federal Constitutional Court of Germany regarding the lawsuits with file reference 2 BvR 1390/12, 2 BvR 1421/12, 2 BvR 1439/12, 2 BvR 1824/12, 2 BvR 6/12
Within the Monetary Union of the European Union, the monetary policy framework is given by the Maastricht Treaty and the legal acts based thereon. The general principle is a stability-oriented monetary policy with the goal of price level stability which is implemented by independent central banks and to whom monetary government financing is prohibited. This reflects the experiences of those central banks which prior to the Monetary Union were independent, and able to ensure monetary stability with a focus on price level stability. The narrow and clearly defined mandate of the central bank system recognizes the particular constellation of the Monetary Union: a community of countries which have assigned responsibility for monetary policy over to the supranational level, but which continue to decide on fiscal and economic policy primarily at a national level, and which deliberately did not enter into a liability or transfer union. Within this scope, the protection of the common monetary policy, from, for example unsound government financing of some member states, is ensured by the exclusion of liability for other member states, the prohibition of monetary government financing as well as the independent role of markets in the evaluation of the solvency of member states of the Monetary Union, which derives from their individual fiscal responsibility. The latter is expressed by the respective risk premia governments incur when borrowing on capital markets.
The financial and economic crisis since 2007 as well as the debt crisis in some member states of the common currency area since 2009 witnessed a considerable enlargement of the range of monetary policy instruments employed, and a strong expansion of the balance sheets of Eurosystem central banks. With these measures, the Eurosystem has made a substantial contribution to the containment of the crisis. The Bundesbank endorsed many of the measures taken. However, the Bundesbank considers some decisions as very problematic, and has also publicly stated its criticism.
Against the background of the arguments and views newly presented during the cases at the Federal Constitutional Court of Germany, the Bundesbank hereby completes its statement. This statement focuses on the purchase of government bonds by the Eurosystem, the TARGET2 balances, and the resulting risks of losses for the federal budget.
B. The OMT program
Following a brief description of the enacted OMT program, the monetary policy justification for this program is discussed with a particular regard to a disruption of the transmission mechanism. The subsequent section elaborates in more detail on the instrument of government bond purchases and its relation to monetary government finance.
I. The current operating framework of the OMT program
On September 6, 2012, the Council of the European Central Bank ended the SMP. It had been initiated in May 2010 and the Eurosystem’s current stock of government bonds which were acquired in the context of the SMP amounts to an acquisition volume of approximately 210 bn €. The OMT program was enacted at the same time as the SMP was ended.
As was the case for the SMP, the purpose of the OMT program is to ensure a properly functioning monetary policy transmission and the uniformity of monetary policy. In contrast to the SMP, the extent of OMT purchases is explicitly not limited, but tied to certain conditions. A complete macroeconomic EFSF/ESM adjustment program or a preventive EFSF/ESM program, both of which include the possibility of EFSF/ESM primary market purchases, are mentioned as a necessary, but not sufficient requirements. However, the ECB Governing Council reserves the option of setting further conditions beyond what was mentioned. As a basic principle, OMT purchases shall be stopped if a government breaches a program’s conditions. No purchases are to take place during an ongoing program review. Furthermore, purchases are only allowed to take place if the respective government has broad access to the capital markets.
Further details of OMT include the organizational and monitoring involvement of the IMF in the adjustment program, as well as the possibility of bond purchases from the current program countries (Greece, Portugal and Ireland) once they have regained access to the bond market. Moreover, and as in the case of the SMP, liquidity provided through OMT purchases shall be fully sterilized, regardless of how long the current full allotment of monetary policy refinancing operations lasts. In contrast to the SMP framework, OMT purchases shall be focused on the shorter end of the yield curve which is understood as bonds with remaining maturities from one to three years. As it is not desired that beneficiary member states shift possible issuances towards shorter maturities in response to purchases, the issuing behavior of these states shall be closely observed. The Eurosystem does not want to claim a preferential creditor status with respect to government bonds acquired through OMT purchases. The public shall be informed about the total amount and market value of the OMT portfolio on a weekly basis, and in addition there shall be a monthly disclosure of average times to maturity and a breakdown with respect to states. The start, continuation and suspension of government bond purchases are at the sole discretion of the ECB Governing Council and shall be decided in accordance with the monetary policy mandate.
II. Monetary policy justification of the OMT program regarding the transmission
The ECB justifies the necessity to guarantee monetary policy transmission via the OMT by the following argument: the current situation is characterized by risk premia on government bond returns which are attributed in particular to the unjustified fear on the part of investors regarding the reversibility of the euro. As markets for government bonds would be important for different parts of the transmission mechanism, these risk premia would undermine the functionality of monetary policy transmission. The effectiveness of monetary policy measures would hence be limited particularly in those countries of the euro zone where (unjustified) risk premia on government bonds exist. Due to tensions on the markets for government bonds the banks’ lending capacity is considerably limited, which would have negative consequences for the real economy. Thus, OMT purchases shall serve to better align the financing conditions of the real economy with the ECB’s policy rates. Hence, the following sections discuss the concept of monetary policy transmission as well as the necessity of a correction in case of disruption and the role of conditionality of OMT purchases.
1. The concept of the transmission mechanism
The monetary transmission mechanism is a process through which monetary policy decisions influence the economy in general and the price level in particular. Monetary policy decisions are transmitted through different channels which impact the various phases of the transmission process. In general, monetary policy decisions impact the prices with a time lag and have different impacts depending on the economic situation. In principle, it is not possible to give a precise estimate (point estimate) on the consequences of monetary policy measures due to uncertainties regarding the data used (for example, revisions, under-reporting), models (nonconsideration of important channels of transmission) and parameters (for example, regarding structural changes within an economy). It is only possible to provide probabilistic statements, typically using confidence intervals. The only thing one can say with certainty is that central banks are generally confronted with long, variable and not precisely predictable delays in monetary policy effects; a delay of one to two years until the full impact of monetary policy actions is a rough estimate. Moreover, it has to be considered that the economy is in a constant state of change. For monetary policy transmission in particular this means that one cannot assume it remains constant over time, but rather is altered by increasing globalization, structural reforms or behavioral changes, for instance. Whether and how these changes manifest can only be empirically assessed sometime after they occur, sufficient amount of data is required.
At the beginning of the transmission or impact channel usually stands a change in the main instrument used by the central bank, i.e. the policy rates. The central bank decides on interest rates applying to its monetary policy operations and, due to its monopoly on central bank money creation, thereby determines the refinancing costs of commercial banks. By steering the refinancing costs the central bank exerts a considerable influence on money market rates. In turn, changes in money market rates have an effect on other interest rates to a different extent, such as the banks’ interest rates on short-term loans and deposits (“interest channel”). Studies within the Monetary Transmission Network of the Eurosystem show that it is the interest channel of monetary policy that primarily influences economies in the euro area.
Interest rates for longer duration securities (such as on government bonds or private borrowing) are only indirectly influenced by changes in money market rates, as expectations of market participants play an important role in determining long-term rates. As such, long-term market rates are determined to a considerable extent by the expectations of market participants regarding the long-term development of growth and inflation, and thus by the long-term perspectives of an economy (or a currency area). Accordingly, expectations regarding future policy rates and thus monetary policy pursued are reflected in long-term market rates. As a result, changes in policy rates only influence long-term markets rates once they induce a change in market expectations regarding the long-term development of prices and the economy.
Via the outlined interest channel, changes in interest rates influence the saving, consumption and investment decisions of private households and firms in various ways. The resulting demand effects temporarily influence economic activity and, moreover price development. Due to their impact on financing conditions, monetary policy decisions also influence financial parameters such as asset prices (“asset price channel”) and exchange rates (“exchange rate channel”). Finally, changes in policy rates can also affect credit supply (“bank lending channel”) as well as firms’ balance sheets (“balance sheet channel”).
Some important aspects of the monetary transmission mechanism and its transmission channels are still not fully understood, even after a series of recent empirical studies. In particular, there is only limited knowledge as to how institutional changes and financial innovations since the beginning of the monetary union in 1999 changed the dynamic relations between different economic variables in the euro area (which in turn may have had an impact on monetary policy transmission). On balance, there is only incomplete knowledge as to how monetary policy transmission has developed over time – this always applies in particular to recent data, as the demand for data analytics is large and only a small amount of data is available to begin with. Since the outbreak of the financial crisis, this knowledge gap has expanded rather than diminished.
The following stylized facts can be summarized: first, there still exist indeterminate delays regarding the transmission of monetary policy to the price level. Second, as a rule monetary policy influences the overall economic development essentially via the interest channel. As already explained tightening monetary policy temporarily leads to a decrease in production which reaches its peak approximately one to two years after the respective interest rate increase. This finding is essentially based on studies using data from the time before the crisis. Price decreases tend to take place more slowly and prices react more sluggishly to monetary changes than production. Third, changes in interest rates influence the business cycle through firms’ cash flows and banks’ credit supply and thereby underline the importance of the bank lending channel of monetary policy.
2. Existence of a current disruption of the monetary transmission channel
Just as the transmission of monetary policy to the real economy encompasses a series of different mechanisms and reactions of economic actors, so too, disruptions of the monetary policy transmission channel can arise in different places. A disruption of the transmission process can be described as an unexpected effect of a monetary policy impulse. Plainly speaking a disruption of monetary policy occurs if the impulses do not evoke their “normal” or “usually to be expected” effect regarding sign, timing or strength. Or, to describe it in a more technical manner: a disruption can exist if current estimates regarding the monetary policy transmission lie outside the range of those confidence intervals which one “usually” obtains from econometric estimations. An obvious reason for a disruption can be that one or several of the aforementioned transmission channels work either just partially or not at all. This may be the case if, for example, the formation of prices in certain markets no longer follow the traditional empirical pattern. For instance, the risk assessment of market participants may be detached from traditional factors in a crisis-like situation on financial markets.
However, it is not possible to determine whether a current deviation represents a temporary disruption or a fundamental, long-term change of monetary policy transmission, since the economic context has changed such that the assumed consistency of effects is no longer given. If the change is fundamental then the use of the term “disruption” would not be justified. Moreover, an exact diagnosis is impeded by the fact that according to consensus opinion such a disruption can only be evaluated with certainty in hindsight. An empirical analysis of the monetary transmission mechanism in real time is therefore afflicted with considerable uncertainties.
3. Indications for a disruption of the monetary transmission mechanism
Hence, a diagnosis in times of crisis depends on the evaluation of indications. Given the prevailing uncertainty and also the potential for serious long-term risks resulting from monetary policy measures taken, these indications should be subject to heightened scrutiny.
Risk premia on government bonds as an indication of a disruption
As an indication of the presence of a disruption of monetary policy transmission, one could, study returns on government bonds of the members of the euro system which, in comparison to the pre-crisis time, developed differently. With the outbreak of the sovereign debt crisis, the yields on government bonds of the so-called periphery countries developed in a clearly different way than the yields on, for instance, German government bonds.
However, recent developments in the market for government bonds cannot with certainty be used as an argument for a disruption of monetary transmission, as it is not possible to determine whether a “disruption” in the development of yields on government bonds can be attributed to fundamentally justified causes or whether possible exaggerations, irrationalities or other forms of inefficiencies are present. Such an argument would require proof of both an inaccurate market valuation of government bonds of individual member states as well as a transmission of this mispricing to financing conditions of the private sector. In order to determine that there is a mispricing, it is indeed possible to resort to observable data on fundamentals such as governmental or macroeconomic debt or deficit ratios. However, quite different results can arise depending on the the specifications of the model. Moreover for a forward-looking analysis of the sovereign debt crisis, already implemented reform and consolidation measures should be considered, which adds to the uncertainty and subjectivity. In addition, a crucial factor in the market’s valuation is the extent to which the further implementation of reforms, the compliance with conditionalities and ultimately the service of private and governmental debt are seen as safe. Thus, focusing on risk premia of selected government bonds is not sufficient. If it is not possible to quantify individual risk components with certainty and to attribute and interpret them unambiguously, that is, to analytically disentangle risk premia, then as a result any interpretation and resulting recommendation for actions can be justified through respective assumptions.
High financing costs for the real sector as an indication for a disruption
Another indication of a disruption of the monetary transmission might be derived from the national differences regarding interest on borrowing to nonfinancial corporations and private households. It is undisputed that with the emergence of the sovereign debt crisis, the financing environment of credit institutions deteriorated as a result of harm done to national banking systems by the impaired credit-worthiness of individual states.
This applies all the more as noncompliance with reform obligations shall lead to a discontinuation of OMT purchases. In such a case, the solvency of a state would be in acute danger and risk premia can therefore be rational. It is obvious that such doubts regarding the sovereign solvency can impact the private sector, as the financial outlook and thus credit-worthiness for parts of the private sector would be extremely damaged by the potential for state insolvency. Higher financing costs for the private sector can thus mirror higher sovereign fiscal risks. This would not be a development that is to be dealt with by monetary policy, but is the direct consequence of the independent national fiscal policy.
In this respect different market rates within the euro area do not contradict a uniform monetary policy. In principle, different economic fundamentals should result in different equilibrium market rates for member states of the EMU. Therefore, it is highly doubtful whether a uniform market rate within the currency union is a desirable economic situation: these doubts are justified as long as individual countries differ in their economic fundamentals. With this in mind, a possible disruption of the transmission mechanism should not be diagnosed by the absolute level of the market rate but rather by the change of the general interest rate level in response to changes in the policy rate, i.e. by the transmission of interest rates. The question is whether the financing conditions of the real economy are in accord with the Eurosystem’s policy rates, such that price level stability can be guaranteed through the resulting impact on economic activity.
Just as differences in interest rates per se cannot serve as an indication for a disruption in the transmission mechanism, one cannot necessarily speak of a well-functioning transmission mechanism in case of a uniform level and reaction of the private sector’s financing costs. From the present point of view, it should be indisputable that the risk premia of periphery states’ government bonds were clearly too low compared to returns on German bonds prior to the outbreak of the sovereign debt crisis in 2010 – be it due to inaccurate risk assessment or be it that the exclusion of liability stipulated by EU contract was not considered to be credible. Due to different data on fundamentals (economic structure, business situation, expectations regarding the future business situation and political development as well as the realization of shocks etc.), a heterogeneous transmission (despite a uniform policy rate of monetary policy) could be expected and even economically in order. An (enforced or artificial) uniform interest rate or an (enforced) uniform transmission of interest rates could impede just these necessary adjustments.
Finally, it is also necessary to consider the relationship between the refinancing conditions of the financial sector and the supply of credit for the real economy more carefully. Even if the banks’ financing terms depend on the level of government bond returns, one does not only have to examine the relationship’s actual strength (and whether it is the same in each state). Beyond that, one has to examine the importance of a possible increase in risk premia for lending activities and thus for the aggregate demand for loans and the price development. The level of government bond returns on its own does not allow for conclusions in this respect. Moreover, the importance of government returns for monetary policy transmission might have to be relativized, as the banking sector is, at least in the short run, less dependent on refinancing through the money or capital market since the transition to full allotment. Eventually also the implication of high risk premia is less obvious as it first seems: do high risk premia which are not fundamentally justified put more pressure on bank balance sheets and thus monetary policy transmission as if they were fundamentally justified? Does monetary policy have to intervene in the first case, but not in the second?
4. Necessity of monetary policy adjustment of a disruption
Even if one would abstract from the problems associated with a practical assessment of a disruption of the monetary transmission mechanism and affirm such a disruption, the question is whether and why such a development has to be adjusted – and whether and why this has to be realized through monetary policy. In the end, economic developments in the various states are connected with potentially different risks as long as decisions on economic and financial policies remain on a national level. These differences also justify different risk premia for private credit relationships. In this respect, monetary policy (interest rate) being targeted at the euro area can be dominated by country specific developments without it being irrational or calling for correcitve monetary policy action. The uniformity of monetary policy within the Eurosystem thus rules out measures and decisions which are only aimed at the elimination of national disruptions.
These considerations especially apply to the argument raised in the context of OMT that a currency reversibility of individual member states would lead to additional interest premia and that this would not be acceptable from the perspective of monetary policy. However, in light of the continued sovereignty of national states, the current composition of the currency union cannot be guaranteed – at least not by the central bank. Hypothetically, this would only be possible if it agrees on unconditional, unlimited financing for each state in order to prevent withdrawal. This does not belong to the field of responsibility of monetary policy, though. And even in the case of monetary policy assistance, the majority of a state’s population can urge to a state’s withdrawal from the currency union and the elected representatives can democratically decide on withdrawal, as they are not able or not willing to create the economic foundations to stay in the currency union. An opinion on the probabilities of such political developments and the related appropriateness of governmental and private debt instrument prices necessarily has to be highly subjective.
Just as the diagnosis of a disruption in monetary transmission, this subjectivity translates to a government bond purchasing program whose purchases which are in principle unlimited should be based on such considerations. And if it is not beyond any doubt that individual states remain in the currency union, the occurrence of reversibility premia per se is no justification for unlimited monetary policy interventions aimed at eliminating these premia. Decisions on how the euro area is composed or whether and how this composition is guaranteed fall to other actors, above all governments and parliaments. The respective risks have to be evaluated by them and, if necessary, be borne via assistance measures.
5. Conditionality of OMT purchases
Furthermore, it is argued that conditionality and the compliance to it would be of particular importance for the program’s legitimacy. That is to say, if conditionality were adhered to and the assistance program realized, redenomination or insolvency premia would obviously not be justified: on the one hand, a sustainability analysis would have to come to a positive conclusion, and on the other hand, the expected successful program realization would make it seem irrational for the state to “derail”. The assumption that the program terms agreed upon within a comprehensive political compromise and a certain level of information are realistic and thus have to convince rational investors is vital for this line of argument. The case of Greece however illustrates that program terms do not have to be able to exclude insolvency. It remains open whether the program’s assumptions, being definitely too optimistic in retrospect, could have been avoided. In any case, against the background of the experiences made so far, it is extremely ambitious to argue that monetary policy would be able to assume compliance to conditionality in any case and, based on these grounds, guarantee that the respective state remains solvent permanently and risk premia decrease in a fundamentally justified manner.
A state’s solvency and low refinancing costs would only then be definitely guaranteed if the option of an eventually unlimited and unconditional monetary financing of the state existed which, however, is incompatible with the Eurosystem’s mandate. In addition, based on the announced conditionality the transaction of OMT purchases would have to be suspended if the program terms can no longer be met. It is obvious, however, that the Eurosystem would then be deemed in a dilemma and inevitably, the question arises of why just in this situation, where savings programs can no longer be met and tensions start rising, monetary policy will suspend its interventions and no longer describe monetary transmission as being disrupted.
6. Findings so far
Even if monetary policy works differently within the euro area, it is doubtful whether these differences constitute a disruption that has to be eliminated by monetary policy. The finding of a substantial disruption of the monetary transmission process will always involve strong subjective elements. It is inarguable that secondary market purchases can enable a temporary reduction of risk premia. Whether these purchases serve a sustainable stability-oriented development of the currency union is doubtful: due to comprehensive risk transfer to the Eurosystem related to OMT purchases, the reversibility of these measures becomes increasingly difficult, while fiscal policy can feel less burdened regarding the discharge of its duties. The assessment of the appropriateness of a country specific risk premium, the transfer of risk in connection with assistance measures (on the primary or secondary market) and in relation to potential misjudgments regarding the future economic and political development in a country is indeed the prerogative of fiscal policy. Another reason for this is that it is fiscal policy which ultimately decides on program terms in the particular case, as well as on the organization and further development of European integration as a whole, and since fiscal policy is subject to the direct parliamentary control. If monetary policy is loaded with this function, it is in danger of being dominated by fiscal policy considerations potentially endangering the goal of stability.
III. Government bond purchases by the Eurosystem
1. Classification of government bond purchases
Independent of a disruption of the transmission process and the question whether this disruption shall or can be eliminated through monetary policy, the question arises whether government bond purchases in terms of OMT are an instrument whose use is allowed for the Eurosystem.
In general, also nowadays open market purchases of government bonds by central banks are not unusual. However, current government bond purchases by, for instance, the Bank of England (“BoE”) or the Federal Reserve System (“Fed”) deviate in a fundamental manner from scheduled bond purchases by the Eurosystem in terms of OMT purchases. These bond purchases do not aim at securing the solvency of states, but rather at open market operations. These aim at influencing the risk-free interest rate rather than at the solvency risk premium of individual member states of a currency union. In this respect, the Fed, the BoE and the Bank of Japan (“BoJ”) purchase those bonds of the respective central government with a high credit-worthiness. In contrast, the Eurosystem intends to decrease high risk premia of individual poorly rated member states of a currency union by means of government bond purchases. Furthermore, it has to be taken into account that the aforementioned states are federal states or central governments where the central state level holds the central bank’s equity and bears gains or losses. In this respect, purchases do not provoke redistributions between tax payers of different independent member states.
The discussion about government bond purchases by the Eurosystem also has to take into consideration that, in contrast to the mandates of the central banks mentioned, the European primary law imposes certain obligations in this respect. The prohibition of government bond purchases on the primary market is a key element of the prohibition of monetary state finance. It represents a necessary element to guarantee the independence of ECB and national central banks of the Eurosystem.
The prohibition of primary market purchases leads to the understanding that the ECB and the Eurosystem can purchase government debt instruments on the secondary market within their monetary policy mandate. In this respect, it is assumed that in principle, the purchase of government bonds on the secondary market is captured by the rule in article 18 of the statute of the European System of Central Banks and the European Central Bank (ESCB statute) regarding monetary policy operations by the Eurosystem. However, these secondary markets are required to move within monetary policy assigned to the Eurosystem and are not used to circumvent the prohibition of primary market purchases. Against the background of a possible evasion of the prohibition of monetary state finance, the ECB is not allowed to purchase such government bonds on the secondary market which aim at financing federal budgets independently of capital markets.
The concepts of “monetary policy” and “monetary policy measure” are however not sufficiently defined. In this way, all measures taken by a central bank could be perceived as monetary policy in general, including comprehensive government bond purchases (such as the Fed’s quantitative easing, for instance) or even a comprehensive monetary state finance aimed at guaranteeing a state’s solvency. However, that implies neither this type of monetary policy to be necessarily sensible and beneficial to price level stability or possible minor objectives, nor does it imply that such monetary policy measures would be backed by a central bank’s mandate. If a central bank was authorized to pursue every action that with some plausibility might be regarded as beneficial to monetary policy and price level stability and that can be realized by the central bank itself, then normative limitations of monetary policy would be ineffective ex ante and thus unnecessary. However, this would rigorously challenge the legitimacy of monetary policy independence. The reason is that legitimacy is based on the fact that monetary policy operates within a binding and externally given, clearly defined framework. This is respected not also, but especially during times of crises and external pressure. Also, this may very well have been the understanding that prevailed during the currency union’s foundation and that was communicated to the public.
2. Regarding the prohibition of monetary state finance
When done on a large scale, secondary market purchases also support the financing of government budgets. But then these purchases are accompanied by considerable stability risks. This reasoning is based on the economically explicable and historically supported experience that governments are in principle more short-term oriented and thus inclined to accumulate debt. In the absence of limitations, this disposition to indebtedness can lead to an undesired and continuous accumulation of debt which can eventually endanger government solvency.
Such a development also compromises monetary policy targeted at the objective of monetary stability: indeed, if public debt in domestic currency is getting out of hand, monetary policy is generally able to guarantee government solvency by eventually covering government financing needs by monetary government financing. The more monetary policy uses its instruments to ensure government solvency, the more its hands are tied to pursue its actual objective of price stability at the same time. Measures such as increasing policy rates make it more difficult and expensive to meet government financing needs. Thus, the more monetary government financing proceeds, the more it dominates the objective of price level stability.
If a situation arises where public finance threatens to be getting out of hand, monetary policy faces strong pressure to guarantee government solvency without regard to the future risk of inflation. Naturally, in the case of government insolvency there are severe risks for financial stability and the broader economy (and the monetary transmission mechanism does not work as in normal times). At the same time, it is often politically very difficult to take fiscal policy measures aimed at guaranteeing solvency on an adequate scale. The central bank is then faced with increasing expectations as the only agent alleged to be capable of acting. In the most extreme case, monetary policy is left no choice but to abandon price level stability or to accept the dangers arising from government insolvency. In the former case, government insolvency can be avoided if government debt is denominated in domestic currency and can be financed by the central bank (this is not possible if the government debt is denominated in foreign currency). But even in situations long before government insolvency looms, it will be difficult for a stability-oriented central bank to guarantee price level stability in the case of unsound fiscal policy. If economic agents expect the situation to escalate with a certain probability, and therefore put additional pressure on monetary policy, then there can come a point in time when inflation expectations rise rapidly and by a large amount, which makes it much more difficult for the central bank to ensure price level stability.
At the moment, no such development is seems to be present. Long-term expected inflation identified by means of surveys or financial market data remain anchored, although, according to surveys, the uncertainty regarding the future development of inflation has increased. Unanchoring [of inflation expectations] in the Eurozone (which is designed as a stability union) does not have to have occurred to prove monetary government financing. A government, which in case of emergency is able to rely on the central bank to finance its expenditure policy, is as a rule more inclined to debt financing. Awareness of this incentive problem was indeed present during the foundation of the monetary union and molded the organization of the economic and monetary union. Fiscal rules were adopted to prevent imbalances of government finances; to strengthen them the Stability and Growth Pact was created. Furthermore, exclusion of liability was decided upon, which would generally cause investors to assess risks more carefully also in the case of government bonds. This, in turn, would lead to risk premia for less sound government finances which would discipline fiscal policy. Moreover, a prohibition of monetary government financing was put in the EU Treaty in order to guarantee the objective of price level stability. This clarified ex ante that monetary financing of governments would not be an option, and that monetary policy would not come to the rescue even in case of strong political and economic pressure. The goal was to make it in the self-interest of fiscal policy to independently prevent the fiscal situation from deteriorating sharply. Shifting this responsibility to monetary policy was intended to be impossible.
In retrospect, these precautionary measures did not have the desired success. In the time before the crisis, fiscal rules were insufficiently adhered to and the SGP increasingly eroded, such that it ultimately was no longer able to fulfill its stabilizing function. Moreover, serious undesirable developments occurred in other areas of numerous economies, and in addition a strong increase in private sector debt in some economies.
In the European Monetary Union, all member states were to be protected from disregarding the objective of price level stability within the currency union (as well being protected from redistribution of member states' solvency risks among national taxpayers). Even a single precedent which damages the basic framework and undermines the fundamental prohibition of monetary government financing is capable of considerably impairing the desired monetary predominance; the precedent could subsequently result in tempting other member states to rely on monetary government financing. In such a situation, the protection of price level stability as a monetary policy objective would stealthily retreat behind fiscal interests. The common monetary policy would be increasingly dominated by solvency ensuring measures and credibility would be endangered. Once monetary policy is on such a precipitous path, reversing course is very difficult and associated with large costs: this is so even if, subsequently, monetary policy were to adopt a strict stability policy thereby (unexpectedly) tolerating a member state’s insolvency, the credibility and legitimacy of monetary policy would be damaged since the prior misinterpretation would be manifested in large losses with the resulting burden for EMU taxpayers via central bank balance sheets.[17 Compare with D]
In order to prevent such a gradual, but difficult to reverse dynamic it would be particularly important to demonstrate early on that monetary policy does not allow itself to be encroached upon and that the central bank strictly interprets the prohibition on monetary government financing. However, in a fiscal crisis this is potentially at odds with the arguments put forward regarding the disruption of the monetary transmission mechanism and also with the necessity of containing a possible financial market crisis. Monetary policy then gives the impression that it can be used for fiscal policy purposes insofar as it tries to compensate for insufficient fiscal policy actions; here, the development in the case of Greece (see following section) is an alarming example. In this context, the presentation of unlimited government bond purchases as an option is as alarming a signal as the argument commonly heard in public debate that timely, appropriate actions by fiscal policy are, as a practical matter, impossible; thus, only monetary policy would be in a position to act. Here, the seeming lack of alternatives leads to the conclusion that monetary policy has to act irrespective of legal, economic and institutional constraints. In such a situation, actions taken by monetary policy can reinforce governments’ impression that shifting the pressure for action to monetary policy is, from a national perspective, a beneficial strategy as the associated political costs can be reduced or even completely shifted. Thus, the lack of alternatives for central bank action becomes a self-fulfilling prophecy.
The liquidity provision to Greece
The Eurosystem’s liquidity provision to meet the financing need in Greece is especially problematic, and shows that monetary policy has assumed responsibilities of fiscal policy. The Greek need for liquidity was financed through the Eurosystem, without regard for the solvency of the country or its banks. While there was a tendency to transfer the risk accruing from the Eurosystem’s liquidity provision to the Bank of Greece, the latter is hardly in a position to independently assume potential losses on a large scale. In the case of a complete insolvency of Greece and its banking system, and in particular in the case of a withdrawal from the currency union – which, was not unthinkable amidst the political tensions present during the implementation of the adjustment program – the Eurosystem has exposed itself to substantial loss risks.
It seems particularly alarming that ultimately, financing of barely solvent banks without adequate collateral was permitted via ELA, where banks also used the obtained liquidity to finance the Greek government. The problems arising from the encroachment of monetary policy into fiscal policy became especially clear during the actions taken in connection with the looming insolvency of Greece in August 2012:
· The implementation of the Greek assistance program fell considerably short of its targets, which was later confirmed by the adoption of a new program.
· European fiscal policy refused to pay out the promised assistance, when conditions were not met.
· But Greece had considerable financing needs (especially refinancing maturing debt).
· Plainly, Greece had decided not to meet its financing needs by means of fiscal policy (e.g. by freezing expenditures).
· Instead, Greece issued additional T-Bills. However, this was not possible due to continued, regular access to capital markets. Instead most of these T-Bills were undoubtedly purchased by Greek banks. However the Greek banks had no reserves left, as they had already been relying on liquidity assistance by the central bank (ELA).
· Greek banks were only able to acquire these T-Bills since the Eurosystem had, in an ad hoc manner, considerably extended the volume of T-Bills, which were accepted as collateral for liquidity operations. In this way, Greek banks were able to purchase additional Greek T-Bills, and immediately refinanced themselves at the Greek central bank.
Thus monetary policy enabled a state’s financing through the provision of liquidity despite that the conditions of a fiscal assistance program were not adhered to and fiscal policy had stopped the disbursement of further assistance. These experiences also provide a basis for fearing that the handling of conditionality within the OMT program, even in dubious cases, will not to guard against vast purchases, and thus not against against redistribution of risk through the Eurosystem’s balance sheets.
The preceding considerations regarding monetary government financing certainly offer some room for interpretation regarding purchases of government bonds on the secondary market. From our point of view, these have to be critically evaluated due to the following reasons:
· As already mentioned, government bond purchases by other central banks differ from those of the Eurosystem insofar as the Eurosystem targets bonds of poorly-rated sovereigns and thus the balance sheets risks are considerably higher. Moreover, the purchases of other central banks are of bonds of their own federal or central government where the central government owns the central bank’s equity and bears gains or losses, thus purchases do not imply redistribution between taxpayers of different autonomous member states.
· If the Eurosystem effectively imposes an upper limit on the interest rate (or interest rate spreads) of a member state’s bonds through purchases on the secondary market, this will also have an effect on the primary market and thus on new issues of government bonds. The reason is that financial market participants can be sure that they can sell a newly issued bond to the Eurosystem at a minimum price at any time. Thus, a member state’s financing access is partially detached from the market prices, since these are now predetermined by interventions of the Eurosystem.
· This becomes more problematic if the majority of a member state’s current borrowing is purchased by the Eurosystem without the guarantee of a considerable time lag from issuance. If there existed a substantial time lag, then the circumvention of financial market pricing would be mitigated insofar as buyers would still hold risks over a noticeable period of time and could not presume to quickly sell the bonds to the Eurosystem. Conversely, the shorter the time lag between issuances and the purchases, and the larger the Eurosystem’s purchase volumes are, the smaller the residual risk for the original buyers. At its extreme, these purchases reduce the role of buyers to simply handing-off the newly issued bonds to the Eurosystem with short delay. The access to capital markets of the respective member state would then largely be protected from market forces.
· This problem of circumventing financial markets also arises even more severely if it is state-owned banks or banks under government control or even the ESM / the EFSF performing the hand-off. This would be all the more problematic if in turn banks were dependent on financing from the central bank.
· Furthermore, purchases can compromise central banks’ independence, which is necessary for accomplishing their main purpose – ensuring price level stability. For one thing, this occurs through the organization of scheduled purchases of government bonds on secondary markets, where a necessary requirement for such purchases is the conclusion of a complete macroeconomic EFSF/ESM adjustment program, or a preventive EFSF/ESM program which includes the possibility of EFSF/ESM primary market purchases. For another, unconditional continuance of the euro zone in its current composition, which is at least implied by the organization and justification of secondary market purchases, ultimately implies that member state can also be financed independently of markets in order to ensure that it can remain in the currency union. This provides the governments in question with a special capability to blackmail the Eurosystem which endangers the independence of monetary policy.
· To justify the recently dormant SMP created in 2010, it was then argued that it was limited in both scope and volume. These limitations shall not apply to the newly adopted OMT program for the purchase government bonds on the secondary market. So over time, and also given the aforementioned guarantee, the debt of a beneficiary member state could to a large extent enter the balance sheet of the Eurosystem. Thus a country could end up with the central bank being by far its largest creditor. This endangers the independence of monetary policy decision-making in the Eurosystem, since the insolvency of a member state would strongly damage the credibility of the Eurosystem.
C. The TARGET2 problems
I. Why are the TARGET2 balances of interest for the case?
There has been a large public debate about so-called TARGET2 balances, which occurred at Eurosystem national central banks since the outbreak of the financial crisis. The public debate has focused on the causes and risks associated with these balances. The guiding principle of this statement is to describe the TARGET2 program and to clarify potential risks. The underlying assumption of this statement is that the Bundesbank constitutes an integral part of the Eurosystem, which consists of the ECB and the national central banks of the member states whose currency is the euro. The Bundesbank assumes that this system will endure and that Germany will remain part of the currency union. This is the basis for the Bundesbank’s considerations regarding risk.
II. What is TARGET2?
Although the national central banks of the Eurosystem were not merged into one central bank upon entering the third stage of the European Monetary Union, their monetary policy competences were transferred to the Eurosystem. Hence, the Bundesbank’s competences in monetary policy were transferred to the Eurosystem with the introduction of the euro as electronic currency in 1999. Decisions on monetary policy are taken by the ECB Governing Council. Changes in the monetary policy framework for action can only be decided upon by the ECB Governing Council. However, the national central banks are still responsible for the implementation of monetary policy according to the principle of decentralization, see article 12.1 of the statute of the European System of Central Banks and the European Central Bank (in the following: ESCB statute). Given this institutional framework, it is necessary to ensure an efficient and reliable execution of monetary policy operations. To this end, a safe and fast payment system was essential in order to provide and absorb central bank money within the unitary currency area without restrictions. For this purpose, TARGET was implemented as a payment system where central bank liquidity could be transferred within the common currency area without restrictions. Banks are supplied with central bank money primarily through refinancing operations, but also through operations which are the responsibility of the national central banks including the accumulation of security holdings and other operations.
TARGET was first launched as a network system on January 4, 1999. To this end, the national RTGS systems were linked by an interlinking component. In 2007, the network system was replaced by a uniform technical infrastructure (TARGET2). The underlying decisions of the ECB Governing Council as well as the respective implementing guidelines of the ECB Governing Council are based on article 22 of the ESCB statute. These guidelines were decreed by the ECB Governing Council pursuant to article 12.1 of the ESCB statute and are binding for all national central banks of the Eurosystem. Therefore, all central banks of the Eurosystem are legally required to participate in TARGET.
Each day, on average, about 350,000 payments amounting to roughly 2.5 trillion € are processed via TARGET2. This approximately corresponds to the German gross national product. The Eurosystem has decided to allow the system, which was created for its own needs, to process transactions other than those related to monetary policy in accordance with its mandate, derived from article 22 of the ESCB statute, and to obtain a better utilization of the system . These payments can be very different in nature. For instance, one can think of the payment for a delivery of goods, the purchase or sale of a security, the granting or repayment of a maturing loan, the financial investment at a bank, and much more. Here, TARGET2 does not exclusively serve for the execution of cross-border payments. For example, in 2011 about seventy percent of all payments executed were domestic and only about thirty percent were cross-border within the German component. Besides TARGET2, there exist other payment systems in Europe (e.g. EURO1) which often have large turnovers. At the end of a business day, the remaining balances from these private payment systems are also settled by TARGET2, since the latter settles payments safely in central bank money, thus the resulting amounts from these other systems are also incorporated into the TARGET2 system.
III. What causes TARGET2 balances?
The decentralized implementation of common monetary policy in the eurozone results in the fact that central bank money can be created in different countries and can flow across borders. If net central bank money flows to a member state’s financial institutions, a positive balance is generated at the respective national central bank, which currently the case at the Bundesbank, for example. If there are net central bank money flows to a member state’s financial institutions, a negative balance results, which is currently the case at the Bank of Greece, for instance. In the TARGET2 system, the counterpart of this negative balance is not another national central bank, but the ECB. The reason is that the legal relationships between the central banks are organized in such a way that a real settlement of accounts takes place where the ECB acts as a clearing center between national central banks. The individual relationships between national central banks and the payment balances generated between them are thus legally eliminated each night; hence, only their payment balances with the ECB matter. In any case, however, bilateral relationships between Germany and the periphery states are only able to explain a small fraction of generated balances. Instead, complex multilateral transactions are responsible for their formation. From a legal perspective, this process is ongoing where balances are updated daily. As long as the system exists, these balances are updated on a daily basis. In a legal sense, balances are claims by or against the ECB for which a regular settlement outside the circular flow of central bank money is not intended. As the continued existence of the system is presumed, its rules and standards do not include rights of termination which could be used to effectuate a repayment of these claims. In our judgment, the relationships between the ECB and the national central banks are not to be understood from a legal perspective as a standard credit agreement as evidenced by the fact that no credit terms are negotiated.
As national central banks involved maintain checking accounts for financial institutions participating in TARGET2, liabilities of national central banks against national banking systems as well as claims against the ECB (TARGET balances) can thus be generated by cross-border liquidity flows, as observed in the case of the Bundesbank. This applies if financial institutions have credit at national central banks. Conversely, in the case of Greece for example, the national central bank can have liabilities against the ECB as well as claims to its national financial institutions, which may be due to refinancing, for instance.
There is no liquidity provision through TARGET2 itself and no credit is granted. TARGET2 balances rather mirror the decentralized implementation of monetary policy decisions of the ECB Council as well as the money market environment, and thus does not allow for limitations on money flows. In normal times, the respective balances were always settled through capital flows between national banking systems, as capital requirements by one banking system are settled by means of interbank lending from creditor countries. Hence, between 1999 and the outbreak of the crisis in 2007 there existed no TARGET2 balances which might have indicated fundamental problems.
IV. Did TARGET2 balances grow in absolute value with the outbreak of the financial crisis?
This situation fundamentally changed with the onset of the financial crisis. Since 2007, larger positive and negative TARGET2 balances were generated within the Eurosystem as a consequence of the financial crisis. Mistrust among banks is responsible for the fact that from that time on, the money market did not function to equalize liquidity balances from surplus to deficit banking systems. At the same time, market based refinancing of banks became more difficult and more expensive. Some institutions were, and continue to be, isolated from the market and draw on liquidity assistances provided by central banks which are inexpensive for them. This applies in particular since the sovereign debt crisis created mistrust of entire national markets.
Ultimately, the current development of TARGET2 balances can be traced back to disequilibria in balances of payments in several EMU states. Here, both current account deficits and capital exports of the private sector can play a role. These are mirrored in outflows of liquidity from these countries, which are ultimately financed by the Eurosystem. Spain and Italy as well as the program countries of Greece, Ireland and Portugal currently have the largest TARGET2 liabilities against the ECB. Apart from Germany (with 715 bn € on November 30, 2012), the Netherlands, Luxembourg and Finland currently have large TARGET2 claims against the ECB. TARGET2 balances increased considerably during 2011, especially during the second half of the year when the financial and sovereign debt crisis grew more acute. Following a further considerable expansion during the first half of 2012, when balances increased by more than 200 bn €, they have averaged slightly above 1000 bn € since June. Within this time period, the Bundesbank’s TARGET2 claims oscillated around a value of approximately 740 bn €. However, the overall stabilization that was observed was accompanied by some considerable daily fluctuations. There were already several phases during the financial and sovereign debt crisis where temporarily, TARGET2 balances did not increase further or even decreased (e.g in the first half of 2009).
1. In the Surplus Countries
Banking systems that receive central bank money via TARGET2, in part because their domestic markets are perceived as safe havens during the sovereign debt crisis, have a lower need for refinancing at “their” central bank. Thus, institutions in Germany have continually decreased their refinancing volume at the Bundesbank and now have large net deposits at the Bundesbank. The German banks direct further inflows of central bank money to their deposit facility or invest them in liquidity absorbing operations of the Eurosystem; an enlargement of the Bundesbank’s balance sheet occurs. It is not necessary to sell assets such as currency reserves in order to compensate for inflowing central bank money. The reason is that the TARGET2 asset vs. the ECB balances against the inflow of central bank money in the balance sheet of the Bundesbank. This TARGET2 asset represents an important part of German foreign assets.
2. In the deficit countries
The above mentioned outflows of liquidity were reflected in financial institutions of deficit countries. However, and in contrast to the pre-crisis situation, they stopped borrowing from financial institutions in surplus states due the loss of trust. The latter were not willing to provide funds, at least not at a low interest rate. The banking sectors in the deficit countries thus financed themselves at their central banks which provided liquidity at prevailing central bank interest rates.
a) Full allotment and the extended collateral framework of the Eurosystem
In the course of the crisis, the demand for central bank financing strongly increased and the provision of liquidity by the Eurosystem was massively expanded. In this way, substantial liquidity was provided within the framework of full allotment. The collateral framework against which the Eurosystem is permitted to provide credit for financial institutions – article 18.1 of the ESCB statute only allows lending on the basis of appropriate collateral – was relaxed several times in order to be able to provide more liquidity and thus meet the deficit countries’ need. These crisis measures of the Eurosystem were meant to be short-term in nature in order to keep the finance system working under difficult conditions. Without this additional provision of liquidity, extremely short-term adjustment processes would have taken place. It turned out that not only a few financial institutions were concerned, but that increasing outflows of payments from the entire banking system were settled. On one hand, this allowed for a prolonged process of necessary adjustments in periphery countries. On the other hand, increased balance sheet risks were incurred due to lowered requirements for solvency of counterparties and collateral accepted. Here, the boundary between liquidity and solvency assistance was shifted and the tasks of monetary policy were greatly extended.
This situation was even more aggravated by the fact that besides this regular source of central bank money, financial institutions experiencing difficulties also drew on additional emergency liquidity (so-called Emergency Liquidity Assistance: ELA) provided by national central banks. This is a traditional instrument of national central banks, and, since it is not specifically mentioned in the ESCB statute, it can be used by national central banks in their own responsibility in accordance with article 14.4 of the ESCB statute. Similar to normal refinancing, this emergency liquidity has to be collateralized. However, in this case the common collateral requirements of the Eurosystem do not apply because these operations are taken at the responsibility of the national central bank in question. Hence, flexible credit can be granted on this basis. In order not to collide with state aid rules of the EU, this is only allowed for the coverage of a temporary liquidity need of illiquid but not insolvent financial institutions. The granting of emergency liquidity through national central banks does not completely lie beyond the ECB Governing Council’s control. According to article 14.4 of the ESCB statute, the ECB council can decide by a two-thirds majority that the granting of ELA is not in line with the objectives and duties of the ESCB. If the Governing Council so decides, the national central bank could no longer grant ELA. In fact, substantial ELA assistance is currently granted for long periods of time, for example the Greek banking sector heavily draws on ELA provided by the Greek central bank.
3. Ways to reverse this development
Within the Eurosystem, an increase in balances can be limited by reducing the strong usage of central bank refinancing by individual financial institutions. This was repeatedly emphasized by the Bundesbank. To this end, based on a majority vote of the ECB Council, the collateral framework of the Eurosystem would need to return to more strictly enforced standards. At the same time, the respective financial institutions either need to be recapitalized or if necessary be liquidated if they cannot survive without the special refinancing by the central bank. Ultimately, it is only fiscal policy which should decide on taking over risks via the provision of credit to banks which are in danger of insolvency. The ongoing lateral movement of TARGET2 balances since the middle of 2012 does not imply that problems are less severe. However, this shows that, in particular, the additional external financing need of periphery countries is not (or no longer) primarily covered by national central banks. At the same time, countries receiving funds from assistance programs have not shown noticeable increases in (negative) TARGET2 balances for a long time.
Measures to directly limit or collateralize TARGET2 balances are currently not planned. They could only be introduced as a change to the TARGET-Guideline on the basis of a majority within the ECB Governing Council. Under no circumstances should such measures contribute to a segmentation of the money market or to a limitation of free capital flows.
The creation of balances can also arise at the US Federal Reserve’s payments system Fedwire. These are registered in the so-called Interdistrict Settlement Account (ISA). Usually, ISA balances are only partially settled. The Fed carries out these settlements by redistributing security holdings within the so-called System Open Market Account (SOMA) in April of each year based on average changes in balances during the past year; hence, security holdings increase for a surplus Fed whose ISA claims decrease accordingly. Such a settlement cannot readily be applied to the TARGET2 system, as the Eurosystem – unlike the Fed – does not generally implement monetary policy by purchasing securities (“Outright Transactions”), but rather by granting credit against collateral in a decentralized manner through national central banks. National central banks are not able to transfer these securities among each other, as they are bound by the respective credit relationship. Moreover, in the US these purchases are almost always targeted at federal government securities or at least at marketable securities of high quality. For the federal government of the United States, the redistribution of interest bearing securities within SOMA is effectively irrelevant. This is because gains of all districts are distributed to the US Treasury (approximately 98% of distributed gains in 2011) after a dividend of 6% on capital deposited is distributed to all member banks.
VI. What information do TARGET2 balances convey?
TARGET2 balances are a good instrument for the assessment of the situation of the banking sector in different countries. They mirror disequilibria in the balance of payments (resulting from trade and capital flows) which are settled by relatively inexpensive central bank financing. If a country’s banks receive net central bank money, then a positive balance is created at the respective national central bank as in the case of the Bundesbank. However, if there is a net outflow of central bank money from a banking sector, then a negative balance is created as in the case of the Bank of Greece.
VII. Risks from TARGET2 balances
1. Continued existence of the Eurosystem
As long as the Eurosystem exists in its current form there are no direct risks from TARGET2 balances, but only from the measures taken to provide liquidity described above. If trust in the banking sector of the euro area and in individual banks is reestablished, and if banks with currently large liquidity problems are either improved or have exited the market, tensions at financial markets should decrease. Recapitalizing solvent banks and liquidating financially unviable ones is an essential requirement in this respect. In addition, those countries which have lost the trust of capital markets need to remove their structural deficits and strengthen their competitiveness in order to improve their public finances and current account situation and to become attractive for private capital again. As soon as this is the case, equalizing capital flows will result between national markets of the Eurosystem which will potentially cause the balances to return to normal.
2. Risks that result from a member state’s withdrawal from the currency union
The withdrawal of a member state of the currency union is a hypothetical case discussed in public debate. In such a case, part of the negative TARGET2 balances can manifest as a balance sheet risk. A withdrawal is not envisioned in the Treaty on the Functioning of the European Union (TFEU). However, such an occurrence would lead to the withdrawal of the respective central bank from its current status in the TARGET2 system, even though this is not discussed within the rules of TARGET2. In such a case, the ECB’s Target2 claims against the national central bank of the departing state become due. The potential for offsetting these claims against paid-in capital and transferred currency reserves is rather negligible in light of the given dimensions. If the withdrawing central bank is not able to fulfill its obligations, an arrangement would have to be found for the remaining difference. In this respect, these obligations of the national central bank against the ECB are in principle the counterpart to claims against its national financial institutions, which are collateralized. However, the ECB as a creditor of TARGET2 balances does not have direct access to this collateral. Within the decentralized organization of monetary policy, collateral is provided to the national central bank. Moreover, a realization of the collateral [,i.e., a default event,] is not necessarily given in such a scenario, as the national financial institution does not have to default against its central bank. Moreover the collateral’s value measured in euro might be relatively limited.
Regarding the level of possible losses, it is ultimately decisive to what extent the withdrawing central bank (or government) is willing or able to fulfill its obligations. As an accounting matter, the remaining claim only needs to be written down by the ECB if it is deemed permanently unrecoverable. A write-down, or a provision taken because of a foreseeable write-down would cause a loss on the ECB’s balance sheet. The risk thus primarily lies with the ECB. Regardless of the timing of the respective payment flows, losses generated within the Eurosystem will eventually have to be borne by the taxpayers of the remaining member states from an economic perspective.
From a risk perspective, the level of TARGET2 claims of the Bundesbank is only relevant in a scenario that assumes the withdrawal of Germany from, or the collapse of the currency union. Here, the claims’ recoverability would depend on the willingness to bring about a negotiated solution on the European level. As already mentioned earlier, the Bundesbank assumes the Eurosystem will survive and that Germany will remain part of the currency union. This is the basis for its risk considerations.
D. Risks for the federal budget
I. General risks of monetary policy
A requirement for the participation in refinancing operations of the Eurosystem is the counterparty’s solvency on the one hand and the provision of sufficient collateral on the other. Within traditional central bank operations, this results in a double safety mechanism, as losses can only occur if both the counterparty defaults and the liquidation of collateral provided is insufficient to cover the claim. In case of adequate provision of collateral according to article 18.1 of the ESCB statute, losses should normally not occur. Hence, refinancing usually generates gains for the Eurosystem. Based on the principle of decentralization and generally uniform collateral, refinancing can take place everywhere within the Eurosystem. If national central banks were able to fully retain these gains, a race regarding these refinancing operations might emerge within the Eurosystem. In order to avoid this, gains generated by refinancing operations are divided among national central banks using the capital key according to article 32 of the ESCB statute.
Conversely, and as a complement to this distribution of gains, losses generated by refinancing operations are ordinarily borne by national central banks according to their capital share in the ECB, as settled by the decision of the ECB Governing Council. The underlying reason is that national central banks have no influence on the choice of collateral, as these are jointly defined for the entire Eurosystem by the ECB Governing Council. Furthermore, there are refinancing operations unrelated to monetary policy, which are excluded from risk sharing. An example is the short-term provision of liquidity assistance (ELA) by a national central bank that chooses the collateral.
In the context of the crisis, the ECB considerably lowered requirements imposed on collateral for monetary policy refinancing operations. This allowed for a comprehensive provision of liquidity in tandem with full allotment and for the long-term provision of liquidity to credit institutions. This leads to risks resulting from operations intended to create central bank money, as higher default risks for this collateral are accepted due to the described expansion of approved collateral. Despite larger haircuts it is thus possibly no longer guaranteed that losses can be fully absorbed.
The decisions made by the ECB Governing Council regarding the distribution of losses remain in force. Hence, financial risks resulting from the expanded monetary policy refinancing operations are in principle borne by national central banks according to their capital share, independent of which national central bank suffered the losses.
In addition, monetary policy includes the accumulation of security portfolios. These are held at the national central banks of the Eurosystem in proportion to the ratio of the so-called banknotes allocation key. This implies that the ECB itself holds 8% and that the remaining amounts are distributed to national central banks of the Eurosystem according to their capital key at the ECB. If such securities default, for example Greek government bonds, the resulting risks are thus also distributed according to the capital key. This results in increased risk provisions on the part of the Bundesbank, leading to a lower distribution of profits to the federal government.
II. Losses in the Eurosystem from a balance sheet perspective
If based on the profit and loss statement, a loss is determined at a national central bank of the Eurosystem, it is handled as follows:
1. Loss at the ECB
If a loss directly occurs at the ECB, it has to be settled using the general reserve fund and the loss provision of the ECB according to article 33.2 of the ESCB statute. If these means are not sufficient to cover the loss, the national central banks, in their position as shareholders, can decide by capital majority (according to article 10.3 of the ESCB Statute) within the ECB Governing Council that monetary income that is to be distributed to them will remain at the ECB for the purpose of settling remaining losses of the ECB. A coverage of losses by national central banks beyond that is not envisioned within the ESCB Statute. Loss sharing would decrease profits of national central banks.
National central banks are not obliged to settle existing losses. However, national central banks are obliged to deposit an amount according to their subscribed share in ECB capital if the ECB’s nominal capital is raised; but the amount deposited for the purpose of capital increases cannot directly be used for loss compensation. In this respect, losses can only be reduced indirectly by using gains to build up additional provisions.
For the case that possible losses of the ECB are not settled with this procedure, the ECB has to disclose loss carry forwards in its annual financial statement until, based on relevant decisions of the ECB Governing Council, the loss is compensated by future gains of the ECB or by future monetary income of national central banks of the Eurosystem.
2. Loss of a national central bank of the Eurosystem
If losses result from monetary policy operations at a national central bank of the Eurosystem, the ECB Governing Council can decide to compensate the national central bank for these losses. According to article 32.4 of the ESCB statute, this compensation can be settled with earnings accruing from the monetary income of national central banks.
If a national central bank suffers a loss due to the Eurosystem’s internal loss distribution mechanism or due to other reasons, it can resort to its current income as well as to its own provisions and in the event that there is a loss on the profit and loss statement to its reserves. The total value of risk provisions of the Bundesbank amounted to 7.709 bn € on December 31, 2011. The increase in reserves of the Bundesbank in the current year corresponds to the legal upper limit of 2.5 bn € as defined by § 27 no. 1 of the Bundesbank Act.
For the case where the losses of a national central bank of the Eurosystem are not completely settled, the national central bank has to disclose a loss carry forward in its annual financial statement until the loss is covered by future gains.
There is no obligation for the owners of national central banks – as in Germany, for most member states of the Eurosystem these are the respective states – to directly compensate losses of the central banks. However, a margin call might be required if a large loss carry forward continues over several years. According to the ECB’s convergence reports, a duty to compensate is assumed if, based on the level and sustainability of losses, doubts regarding the national central bank’s ability to carry out its tasks arise. (However, the same also has to be assumed for a large loss carry forward of the ECB). In its convergence report, the ECB elaborates on the issue as follows:
“For all the reasons mentioned above, ﬁnancial independence also implies that an NCB should always be sufﬁciently capitalised. In particular, any situation should be avoided whereby for a prolonged period of time an NCB’s net equity is below the level of its statutory capital or is even negative, including where losses beyond the level of capital and the reserves are carried over. Any such situation may negatively impact on the NCB’s ability to perform its ESCB-related tasks but also its national tasks. Moreover, such a situation may affect the credibility of the Eurosystem’s monetary policy. Therefore, the event of an NCB’s net equity becoming less than its statutory capital or even negative would require that the respective Member State provides the NCB with an appropriate amount of capital at least up to the level of the statutory capital within a reasonable period of time so as to comply with the principle of ﬁ nancial independence.”
Thus, if risk provisions of the Bundesbank are absorbed by losses resulting from the aforementioned monetary policy operations, this can lead to losses for the Bundesbank which need to be disclosed. In such a case, the federation as capital proprietor would face the question whether to recapitalize the Bundesbank’s losses in order to allow for a solvent balance sheet. Depending on the volume of the respective losses, this can imply considerable obligations for the federal budget though the recapitalization may be considered unavoidable for the preservation of a stability union. The German Bundestag would then have to appropriate the necessary funds.
III. The effect of Eurosystem losses on the federal budget from an economic perspective
A central bank generally has the possibility to create money and could always relieve the national budget by providing liquidity independent of its own gains or losses. However, the Eurosystem is not allowed to engage in government financing. Moreover, the Monetary Union is designed as a stability union, and thus monetary policy is not governed by a government’s financial need, but rather by stability requirements.
Ultimately, the national budgets of member states are financially connected with the Eurosystem. Gains of national central banks are sent to the respective member state, while in the case of losses, national treasuries have to go without this income in the respective year and may face several years without this income and the possibility of having to recapitalize their national central bank at some point. In this way, losses enter the federal budget through the Bundesbank. Regarding the timing of payments between the national central bank and national treasuries, delays are possible if, for instance, gains are retained, loss carry forwards are effectuated or (at first) a loss only enters the balance sheet of the ECB. Given stability-oriented monetary policy, it does not matter for the federal budget from an economic perspective at what point in time possible losses resulting from government bond purchases, for instance, become effective. A prompt recapitalization by the federal government would be necessary if the central bank’s credibility in maintaining its primary objective of price stability was undermined by large losses. Ultimately losses generated by Eurosystem government bond purchases would put the same pressure on the long-term sustainability of government finances as losses sustained by the EFSF or the ESM; and accordingly limit the financial flexibility of the federal government. In this respect, the impact of Eurosystem government bond purchases do not fundamentally differ from secondary market purchases by the EFSF or the ESM. However, the latter require the approval of parliaments and are subject to possible judicial supervision.
Occasionally, it is argued that losses for the federal budget would be even larger if the Eurosystem were not to intervene with secondary market purchases to stabilize the situation and thus prevent a further escalation of the crisis. Thus it naturally follows, in keeping with the spirit of the guiding principles of EMU, that the volume and organization of assistance programs are to be decided within the framework of the EFSF and the ESM. The decision would then lie with national governments and parliaments preventing the dividing line between fiscal and monetary policy from blurring further and ensuring monetary policy’s capacity to achieve its primary objective of price level stability. In this way, decisions on individual assistance measures taken for the benefit of individual countries can be consistent with the current discussion on the basic orientation and further development of the currency union. Decisions on the assumption of large liabilities, appropriate control mechanisms or even rights to intervene in national budget sovereignty, as in the case of a fiscal union, for instance, should in general be considered in tandem. Here, the order of events is in danger of being confused due to the potential of large bond purchases by the Eurosystem, which is not eliminated by the announcement of principally unlimited purchases. In this way, the Eurosystem would actually buy government bonds and thereby collectivize risks without counterbalancing with enhanced rights to intervene which might be viewed as potentially desirable. This increases the danger that the disequilibrium between liability and control will be increased.
 As acquisition prices deviated – typically downwards – from the nominal value of bonds purchased, the specified acquisition volume, that is the amount of liquidity being provided through the purchases, deviated from the nominal value of bonds purchased.
 However, it cannot be ruled out that in such a case the Eurosystem takes other measures with regard to the member state in order to counteract a potential exit from the common currency – even if conditionality is not adhered to.
 Opening statement by ECB President Draghi at the press conference on October 4, 2012: “OMTs would not take place while a given programme is under review and would resume after the review period once programme compliance has been assured.”
 See, for instance, the editorial to the ECB’s monthly report of September 2012.
 Some argue that the stability of the euro will be viewed in a negative light if a euro exit is seen as more likely in even just a single member state. At the press conference following the ECB Council meeting at the beginning of August, ECB President Draghi emphasized the irreversibility of the euro in each member state and explicitly pointed out that drachma and lira would not come back. Self-reinforcing dynamics which are attributed to “irrational interpretations of the market” shall be countered by government bond purchases. It is argued that there are multiple equilibria with harmful consequences which monetary policy could avoid via OMT.
 ECB (2011), Die Geldpolitik der EZB, p. 62. (The monetary policy of the ECB)
 When describing the monetary transmission mechanism, it is usually an implicit assumption that the prices of goods are not perfectly flexible, although this assumption represents an essential premise as to how changes in nominal interest rates can trigger effects in the real economy. For the following explanations, it is not necessary to describe the individual transmission channels. For more detail, see Worms, A. (2004), Monetary policy transmission and the financial system in Germany, in: J.-P. Krahnen & R. Schmidt (ed.), The German Financial System, Oxford University Press or J. Bolvin, M. Kiley and F. Mishkin (2011), How has the monetary transmission mechanism changed over time?, in: B. Friedman & M. Woodford (ed.), Handbook of Monetary Economics, North-Holland.
 Commercial banks demand central bank money to cover their need for cash, to settle interbank loans and to fulfill their obligations with respect to minimum reserve requirements.
 See, for instance, B. Hofmann and A. Worms (2008), Financial Structure and Monetary Transmission in the EMU, in: X. Freixas, P. Hartmann & C. Mayer (ed.), Handbook of European Financial Markets and Institutions, Oxford.
 Besides the traditional bank credit channel which relies on the extent of the credit supply – higher policy rates reduces or raise the cost of capital at banks’ disposal for lending – the recent literature also discusses the so-called “risk taking channel”. This channel unfolds its effects if the banks’ incentive to take risks during lending is influenced by monetary policy decisions.
 The balance sheet channel applies to the fact that an increase in policy rates puts pressure on assets of firms’ balance sheets, for example as certain claims become relatively less valuable. However, a lower net value puts pressure on the firm’s credit-worthiness as a borrower and thus cushions lending.
 First, possible secondary reserves of the banking sector in terms of liquid securities (especially government bonds) decreased, through a rise in risk premia of the securities concerned. Hence, the refinancing of loans through secondary reserves was impeded. Second, higher risk premia reduced the value as a security. Third, the downgrading of a state typically led to a worse rating of the banks in the state concerned. Fourth, implicit or explicit governmental promises of guarantee towards banks diminished with increasing risk premia (Panetta et al., 2011. p. 1).
 In this respect, it can be shown, for instance within the framework of a neoclassical growth model (to be more precise: using the Euler equation), that there exists a tight relationship between the long-term equilibrium growth path of an economy and the level of the (natural) real interest rate in this economy.
 See ECB monthly report, September 2012, p. 10.
 See European Court of Justice decision, 199/2012, of November 27, 2012, point 135 f.
 See the decision of the Federal Constitutional Court, 2 BvR 1390/12, of September 12, 2012, paragraph no. 278.
 See D.
 See Hermann, Europäische Zeitschrift für Wirtschaftsrecht 2012, pp. 805, 811.
 While the refinancing of public banks by the Eurosystem is in line with the EU treaty such an approach can no longer be considered as justified if public banks or banks under government control are used to undermine the prohibition of primary market purchases. If in fact hardly any private investor is at risk of possible losses occurring during the intermittent period (but in fact the government is the proprietor), it stands to reason that this is a means to circumvent the prohibition of monetary government financing (just as in the case of ESM).
 Real-time Gross Settlement-Systems
 Guideline ECB/2007/2 and guideline ECB/2011/NP17.
 Figures according to the ECB announcement: http://www.ecb.int/paym/t2/html/index.en.html.
 TARGET2 has 976 direct participants, 3465 indirect participants and 13083 correspondents, http://www.ecb.int/paym/t2/html/index.en.html.
 The share of TARGET2 in large-value payments in euro amounts to 91% in absolute value and to 59% in volume, http://www.ecb.int/paym/t2/html/index.en.html.
 See the development of other claims of the Bundesbank to the Eurosystem including TARGET balances – available at: http://bundesbank.de/Redaktion/DE/Downloads/Statistiken/Aussenwirtschaft/Auslandspositionen_Bundesbank/S201ATB39697A.pdf?__blob=publicationFile – which fluctuated from -31 to +71 bn between 1999 and 2007, but which were often within a range of a few bn.
 For the Bundesbank, see for instance http://bundesbank.de/Redaktion/DE/Downloads/Statistiken/Aussenwirtschaft/Auslandspositionen_Bundesbank/S201ATB39697A.pdf?__blob=publicationFile.
 Sum of all TARGET2 claims or, as applies, sum of all TARGET2 liabilities.
 Taken from the monthly report of the Bundesbank, November 2012, p. 52.
 Schlesinger, ifo-Schnelldienst 16/2011, p. 9, 10.
 Bundesbank, monthly report, November 2012, p. 52.
 Krämer, Wirtschaftswoche, February 18, 2012, p. 38 and Board of Governors of the Federal Reserve System (2012), Financial Accounting Manual for Federal Reserve Banks, 40.40 SOMA Participation, Revision Set 52, July 2012.
 In 2011, 1.6 bn $ flowed to the district banks and 76.9 bn $ flowed to the central government. (http://www.federalreserve.gov/newsevents/press/other/2012011a.htm)
 This is also a reason why TARGET2 balances pay interest on the basis of the refinancing rate, thus the same interest rate which also accrues directly for refinancing. In this way, returns are transferred to the national central bank to whom the created central bank money flowed according to capital shares in the ECB
 ECB press release of October 15, 2008.
 The decision of the European Central Bank of December 13, 2010 regarding the issuance of euro banknotes, ABI, of February 9, 2011, L35/26, defines in its Appendix I the banknote allocation key as of January 1, 2011, which is assumed for the distribution of the total value of the issued banknotes among the participants of the Eurosystem (according to article 4 paragraph 1).
 ECB Convergence Report, May 2012, p. 28/29.