Central banks are often faced with a cruel irony: they are (almost always) mandated not to lend to insolvent firms and, by nature of being the lender of last resort (LOLR), face a pool of potential borrowers that are uniquely likely to be insolvent. The question of solvency is therefore one of considerable importance to central bankers and illustrates the important role of bank examiners and supervisory authorities. In principle, central bankers should not veer into fiscal policy by lending to insolvent firms. In practice, central banks frequently lend to insolvent firms, sometimes because they thought the firm was solvent when it wasn’t, and sometimes despite an ex ante acknowledgement of insolvency. This often creates challenges for central banks:[1] Depending on their operating framework and legal system, they may be in violation of legal rules or norms; publicization of their intervention might hurt their credibility; their political independence might be threatened; they may create perverse incentives for certain classes of creditors in other banks; and in extremis they risk taking capital losses above what is viewed as appropriate for central bank interventions.
The aim of this note is not to provide a playbook for assessing solvency—how it should be measured, by which authority, etc. Nor will it deal with the related consideration of viability or normative questions about whether a central bank should care about solvency, viability, etc. This note stipulates that in the vast majority of legal and regulatory frameworks around the world, particularly those in Europe, central banks generally do face constraints—legal and regulatory—on lending to insolvent and/or unviable firms. The question, then, is what to make of lending to insolvent firms gone wrong.
The European Union (EU) provides a particularly fruitful setting for this question, since in the EU, central bank financing of an insolvent firm is legally prohibited as monetary financing, a term normally associated with the direct purchase of government debt by monetary authorities. This note will explore through stylized case studies how the EU handles the legal and procedural limits on monetary financing as it relates to emergency liquidity assistance (ELA).[2] (To be clear, ELA has a specific meaning in the Eurosystem: while central bank lending for monetary policy is the responsibility of the ECB, central bank lending for the purposes of financial stability, termed ELA, is the responsibility of the central banks of the member nations that are part of the Eurosystem.) Then, drawing on those case studies, the note will sketch out some advantages and vulnerabilities associated with a common policy intervention employed in such situations: fiscal guarantees.
In Principle: The Brussels View on ELA and Monetary Financing
The EU is a particularly useful setting to analyze norms and rules about monetary financing because the EU’s prohibition of it is a, and sometimes the, binding constraint on emergency lending.
Article 123 of the Treaty on the Functioning of the European Union (TFEU)
Discussion of European regulations necessarily involves the Venn diagram that is Europe: Member States of the European Union (not all in the euro area); members of the euro area (all of which are also members of the EU); and European nations that are part of neither the EU nor the euro area. Here, the EU is the operative group. The EU’s general prohibition on monetary financing is sometimes discussed as European Central Bank (ECB)—and therefore euro area—regulation. That characterization, however, is misleading. While the ECB has implemented procedural arrangements on the provision on ELA and, in 2013, provided a public communication with respect to ELA operations in the form of an agreement among ECB member nations, ECB agreements are not legally binding, but rather are soft law.[3]
In contrast, the prohibition on monetary financing is found in Article 123(1) of the Treaty on the Functioning of the European Union (TFEU[4]), which reads (emphasis added):
"Overdraft facilities or any other type of credit facility with the European Central Bank or with the central banks of the Member States (hereinafter referred to as ‘national central banks’) in favour of Union institutions, bodies, offices or agencies, central governments, regional, local or other public authorities, other bodies governed by public law, or public undertakings of Member States shall be prohibited, as shall the purchase directly from them by the European Central Bank or national central banks of debt instruments."
As such, Article 123 is binding on all Member States, irrespective of their participation in the Eurosystem. In other words, this is a statutory implementation of EU-level law to which even non-Eurozone EU members[5] (e.g., Denmark, the UK before 2020) are subject.[6]
The Long Arm of Article 123
The operative phrase of TFEU Article 123 is “public undertakings.” That is public action that should legally be taken by the Member State governments (i.e., through fiscal resources rather than monetary resources). The Brussels View—what I’ve termed the EU’s view on monetary financing—is that bank recapitalizations and restructurings, which lending to an ex ante insolvent firm would be, are a public undertaking and thus should be supported with public (i.e., state) resources. Various public documents make this clear. For example, in a 2018 speech, ECB executive board member Yves Mersch said (emphasis in original):
"The ECB has repeatedly stated . . . that financing by central banks, even when granted independently and at their full discretion, of credit institutions other than in connection with their central banking tasks, in particular the support of insolvent credit and/or other financial institutions, is incompatible with the monetary financing prohibition. Financing insolvent institutions is a government task – indeed the ECB has identified criteria to distinguish between government tasks and central banking tasks. The ECB has also repeatedly clarified that while central banks may be involved in administering resolution measures, they should not finance them. National central banks could also administer resolution measures on a genuine agency basis on behalf and for the account of a third party, that is, the government or one of its entities. In this case, however, the central bank would not itself provide liquidity, but simply carry out these tasks like an agent."
In other words, the Brussels View is that central banks may be as involved as they find convenient in administering liquidity—even to insolvent institutions—but they cannot finance lending to an insolvent institution. Operationalized, this means that the fiscal authority (or some other organ of the state, e.g. ministry of economy, sovereign wealth fund, etc.) must bear the counterparty risk.
It is also worth noting that monetary policy operations do not fall within the scope of State Aid rules pursuant to Article 107(1) of the TFEU; ELA operations are also excluded, subject to specific LOLR conditions being met.[7] Accordingly, the provision of ELA to an insolvent credit institution would constitute prohibited State Aid.
In 2017, the ECB published its ELA Agreement. In the agreement—which, for the avoidance of doubt, is soft law in the EU and not legally binding as such—the ECB clarifies its views on ELA and monetary financing, writing (paras. 5.2–5.3):
"The violation of the prohibition of monetary financing laid down in Article 123 TFEU may constitute an interference with the objectives and tasks of the ESCB [European System of Central Banks]. The provision of ELA as notified under Sections 3.2(b) and 3.3 is, therefore, assessed ex ante as regards compliance with the prohibition of monetary financing. ELA transactions akin to an overdraft facility or any other type of credit facility for the State, in particular, any financing of the public sector’s obligations vis-à-vis third parties, or the central bank de facto taking over a State task, violate the prohibition of monetary financing."
The agreement references the ECB Governing Council’s ability to prohibit the extension of ELA by a national central bank (NCB, a central bank of a country participating in the Eurosystem) on the basis of monetary financing, insofar as ELA inconsistent with the prohibition on monetary financing interferes with the ECB’s tasks and responsibilities, as laid out in Article 14.4 of the Statue of the European System of Central Banks and of the European Central Bank (ESCB Statute; see OJ C 202, 7.6.2016, pp. 230–250). The prohibition on monetary financing is binding hard law at the EU level, even when other aspects of the ELA Agreement are not.[8]
The ECB has since updated its ELA guidance, most recently as of the time of writing in September 2024. In relevant part, the 2024 ELA Agreement preserves the 2017 ELA Agreement’s explicit permitting of the ECB Governing Council to prohibit the extension of ELA by an NCB on the basis of monetary financing.[9]
While this agreement was published and amended after the resolution of the cases examined here, it is instrumental for understanding the ECB’s view on ELA interacting with the monetary financing prohibition.
In Practice: Stylized Cases of Alleged Monetary Financing via the ELA Mechanism
Determination of a bank’s solvency and/or viability is deeply complex and requires two things in short supply during a crisis: time and detailed information. In practice, during the acute phase of a crisis, distinguishing between solvent and insolvent[10] may be impractical (see, e.g., Goodhart 1999).
The nightmare scenario for central banks, then, happens not infrequently: a potentially systemic institution applies to the central bank for emergency liquidity at the eleventh hour, and the authorities have limited or insufficient information about the institution’s condition. Many central banks in such scenarios choose to err on the side of caution but balance the risks they take in lending to a potentially insolvent firm by obtaining guarantees from the state.
Following are some stylized examples of central banks’ handling of lending to potentially insolvent banks, and how they handled risks associated with monetary financing (all before the publication of the ECB’s 2013 ELA Procedures).
Ireland 2009
In mid-January of 2009, the Irish government nationalized Anglo Irish Bank (Anglo) after it faced severe losses and deposit outflows. In March, the Central Bank of Ireland (CBIR) provided Anglo with ELA, under which the CBIR accepted a broader range of collateral than would be permitted in standing ECB liquidity operations. It continued this lending for years. When it became clear that Anglo was running out of even non-ECB-eligible collateral, the finance ministry recapitalized the bank with nonmarketable promissory notes, which drew much criticism from the European Commission (EC). Through 2010 and 2011, that ELA enjoyed a state guarantee. Eventually, Irish authorities resolved the bank. On the Anglo case generally, see Honohan (2010).
The Anglo case was a controversial touchstone for European and international thinking on monetary financing via ELA. While the initial intent of the ELA liquidity was for it to be short-term, it ended up being extended for four years and accounting for a large share of Anglo’s funding needs. The Anglo case even prompted some Bank for International Settlements (BIS) researchers to argue that, despite frequent use of ELA by central banks to help unwind insolvent banks, it might be time to retire the practice and provide ELA only to going-concern institutions. BIS researchers also argued that central banks should be transparent about the fiscal implications of any liquidity advanced to banks.
The finance ministry’s issuance of nonmarketable government promissory notes to Anglo received particular criticism with respect to monetary financing. In this case, the monetary financing allegation relates to both the more familiar form of monetary financing—when the central bank buys government debt directly[11]—and to the Brussels View interpretation of monetary financing as it relates to ELA specifically. With respect to the Brussels View, the monetary authority was advancing liquidity to a potentially insolvent institution, potentially in violation of TFEU Article 123.
With respect to the direct financing of government debt, in essence, the fiscal authority owned Anglo at the time (after nationalization) and needed to provide it with a capital injection so that it could obtain liquidity at the CBIR. To do so, the fiscal authority would typically issue debt publicly, in standard auction form. However, Ireland had very limited fiscal space and, as a result of the crisis, was facing a nearly nonexistent market for its sovereign debt; issuing debt publicly, then, was impractical. To avoid those financing conditions, the government provided nonmarketable promissory notes to Anglo, against which the CBIR readily advanced liquidity. As a result, in function if not form, the CBIR was financing government debt directly (i.e., without market intermediation). In essence, the fiscal cost saved by the CBIR discounting of nonmarketable promissory notes relative to standard debt issuance would be the spread between the rate on the promissory notes and the rate that the Irish state would have obtained in the market (unknown, but sufficiently high for the Irish state to not even try). That said, when Anglo was put into resolution, the Irish state exchanged outstanding promissory notes with the CBIR for marketable floating-rate sovereign notes, thereby limiting the monetization impacts of the promissory note issuance. For more on the fiscal costs of the Anglo case related to ELA and promissory notes, see Whelan (2012).
In October 2010, the president of the ECB wrote to the Irish minister of finance that ELA could be prohibited on the basis of monetary financing:
"The provision of [ELA] by the Central Bank of Ireland, as by any National Central Bank of the Eurosystem, is closely monitored by the ECB’s Governing Council as it may interfere with the objectives and tasks of the Eurosystem and the prohibition of monetary financing under the Treaties."
The next month, the president of the ECB warned again about lending to insolvent institutions as being subject to monetary financing prohibitions and conditioned further approval for CBIR ELA on Irish fiscal reforms and a state guarantee of the ELA.
Denmark 2008
In Denmark in 2008, Roskilde Bank took larger-than-expected losses and appealed to the Danmarks Nationalbank (DNB) for emergency liquidity facilities. The DNB’s lending to Roskilde was guaranteed by both the private deposit insurer (PCA) for the first 750 million Danish krone (DKK) and the fiscal authority for any losses over DKK 750 million. While the DNB initially thought Roskilde was solvent, the bank fell below its statutory solvency requirement a few weeks after the provision of emergency liquidity. Eventually, the PCA converted its guarantee into an equity stake when the bank was restructured. Monetary financing concerns were not raised by the DNB, the Danish government, the ECB, or the EC. Notably, this was not a case of ELA, given that the DNB was not a member of the Eurosystem. On the Roskilde case generally, see Rigsrevisionen (2009).
While it appears the DNB was never officially charged with monetary financing, Danish authorities later said (para. 110) they “had no knowledge of the European Central Bank’s strict interpretation of monetary financing rules relative to insolvent banks given that [the DNB’s] risk was to be covered by a state guarantee.” Illustratively, Denmark was not a member of the euro area at the time. The EC noted that the state would likely bear any losses on the facility, so there would be no monetary financing concerns (to the extent there was incompatible State Aid, it would be in the domain of government competition distortion).
Latvia 2008
When the large Latvian bank Parex began showing signs of trouble in 2008, the Latvian ministry of finance provided government securities (sovereign bonds) to Parex as term deposits, taking loans as collateral. (These government securities were presumed to be standard marketable securities, unlike in the case of Anglo Irish in Ireland.) Parex was in turn able to use those government securities as collateral at the Bank of Latvia (BOL), obtaining liquidity in both Latvian lats and euros.
In this case, the EC did not raise monetary financing concerns. As we know, these concerns come in two main forms: (1) a central bank lending to an insolvent institution; and (2) the more traditional monetary financing concern, the central bank's discounting of unfunded sovereign debt. Here, the first form of monetary financing was not possible (despite Parex’s ultimate disposition as insolvent), as the initial and ultimate liability was to the fiscal authority. The second form, though, familiar from Anglo’s case, could theoretically be applicable, given that Parex was able to make use of the BOL’s liquidity-provision capacity, specifically its ability to provide euro liquidity, on the basis of providing government securities as collateral. Documentation here is unclear, but again, in the absence of any evidence to the contrary, these securities were presumed to be standard marketable securities, which would explain the EC’s silence on this point. (This would not have been a case of ELA, given that the BOL was not a member of the Eurosystem in 2008; Latvia adopted the euro in 2014.) On the Parex case generally, see European Commission (2008).
Cyprus 2012
Laiki Bank—one of Cyprus’s largest banks—faced severe challenges in 2010–2011, during the European sovereign debt crisis. Laiki was able to provide acceptable collateral to the ECB and obtain liquidity through monetary policy operations until October 2011, when the EU decided to haircut Greek sovereign bonds, resulting in significant losses for Laiki. Throughout the rest of 2011 and through 2012, as Laiki’s deposit outflows worsened, Laiki continued to rely on Central Bank of Cyprus (CBC) ELA for its funding, with the ELA eventually replacing deposit funding. In May 2012, the Cypriot government capitalized Laiki with an unfunded sovereign bond, becoming a majority shareholder. In June 2012, the ECB stopped accepting Cypriot sovereign bonds as collateral, exhausting Laiki’s ability to obtain any ECB liquidity; outstanding ECB liquidity was transferred to CBC ELA. As such, Laiki became wholly reliant on ELA from the CBC. On March 21, 2013, the ECB withdrew its approval for CBC ELA, requiring Laiki to repay its borrowings in five days and conditioned further approval for ELA on Cyprus implementing a joint EU/International Monetary Fund (IMF) program to ensure banking system solvency. On March 25, the CBC resolved Laiki through a purchase and assumption transaction involving another Cypriot bank, the Bank of Cyprus, and the sale of Laiki’s Greek operations. In 2017, the Bank of Cyprus—having undergone its own restructuring and recapitalization—repaid its outstanding ELA (the majority inherited from Laiki) to the CBC. On the Laiki case generally, see, e.g., Schaefer-Brown (2024); Central Bank of Cyprus (2013).
The CBC in 2013 issued a public statement saying that prior to the ECB’s March 25 prohibition, the ECB Governing Council had not objected to the CBC’s provision on ELA to Laiki, citing the CBC’s own view that Laiki was solvent despite capital shortages as a result of “the prospect of recapitalisation through the financial assistance programme.” The CBC then said that the ECB’s eventual decision to stop approving the ELA was a result of Cyprus’s failure to implement intended assistance packages; once the recapitalization plan was rejected by Cypriot parliament on March 15, 2013, that impending recapitalization was no longer reliable and Laiki became insolvent as a result (and the ECB withdrew its consent for CBC ELA eight days later). This statement contained no explicit reference to the monetary financing prohibition.
Documents leaked to the New York Times, however, show that monetary financing concerns—both with respect to ELA to an insolvent institution and to the use of an unfunded sovereign bond as collateral for ELA—were top-of-mind for ECB officials as early as 2012. A then-confidential document of meeting minutes from the ECB Governing Council in March 2012 on Cypriot ELA requests cautioned continued ELA on the basis of solvency concerns: “information is lacking in order to verify the solvency and financial soundness of [Laiki] and hence its status as a counterpart for ELA . . .”
In an email dated July 31, 2012, just months after the recapitalization, with the subject line “ELA collateral,” an ECB official wrote to a CBC official on the question of using the unfunded sovereign bond as collateral for CBC ELA (emphasis added):
"As illustrated by the Irish case, using special purpose government bonds as collateral, whether under monetary policy operations or under ELA, may raise monetary financing concerns and I would therefore advise to first make sure that all alternatives have been exhausted."
A few months later, in September 2012, a letter from the CBC to the president of the ECB shows that the ECB had requested alternatives to CBC ELA against the unfunded sovereign bond (emphasis added):
"It will be recalled that in the discussion on ELA at the Governing Council meeting on 6 September 2012 it was decided that the Central Bank of Cyprus would report as soon as possible on the potential contingency liquidity measures that could increase the funding capacities of [Laiki] excluding the unfunded government bond issued to recapitalise it."
Here again, similar to the Anglo case, two major ELA-related monetary financing concerns were present: first, the straightforward Article 123 prohibition on lending to insolvent firms; second, the “traditional” monetary financing concern related to unfunded sovereign bonds being discounted, without intermediation, by the central bank.
United Kingdom 2007
In early fall of 2007, the large UK-based bank Northern Rock was facing funding difficulties as the markets for short-term debt began to freeze against the backdrop of the beginning of the Global Financial Crisis (GFC). On September 14, the Bank of England (BoE) announced that it would provide Northern Rock with emergency liquidity. When Northern Rock requested expanded liquidity facilities, the BoE amended the liquidity facilities on October 9, providing unlimited liquidity against all of Northern Rock’s assets. Notably, given the risk this posed to the BoE and the potential insolvency of Northern Rock, the BoE obtained a fiscal guarantee on its lending. The government later nationalized Northern Rock, and the Treasury assumed the lending facilities from the BoE, at that point fully eliminating any risk of monetary financing. On the Northern Rock case generally, see Plenderleith (2012).
After the October 9 amendments and fiscal guarantee, the governor of the BoE noted (pp. 127–8) that the fiscal authority was the entity bearing the risk in the lending. While the BoE did not explicitly note that the bank had, by October 9, become insolvent (or that its solvency was ambiguous), BoE leadership did not feel sufficiently guaranteed in their lending alone, and thus requested the fiscal guarantee (which applied only to lending on the facilities after October 9). Notably, this was not a case of ELA, given that the BoE was not a member of the Eurosystem. The EC raised no monetary financing concerns.
Spain 2009
In Spain in 2009, the regional savings bank, or caja, Caja de Ajorros Castilla–La Mancha (CCM) faced severe challenges as the Spanish real estate market, to which it was highly exposed, took losses in the context of the GFC. When CCM exhausted its ECB-eligible collateral, it turned to the Bank of Spain (BoS) for ELA. At first, the BoS advanced EUR 900 million in liquidity secured against CCM’s assets. When a proposed merger later fell through, the BoS expanded its ELA with a credit line of up to EUR 9 billion.
Although the BoS insisted that CCM was solvent at the time, concurrent reporting suggests that one of the precipitating events in the failure of CCM’s merger with another caja was the determination by CCM’s auditors that the bank was insolvent and their refusal to certify CCM’s year-end accounts for submittal to the BoS. The audit report showed that CCM had negative equity of EUR 3 billion. The BoS, before providing the EUR 9 billion credit line, requested a fiscal guarantee. The government passed Royal Decree-Law 4/2009, which gave the fiscal authority the authorization to guarantee up to EUR 9 billion on the BoS’s lending. On the CCM case generally, see European Commission (2010).
The BoS maintained publicly that it had lent to a solvent institution. Nonetheless, in testimony to parliament days after the emergency credit line extension, the governor of the BoS said (emphasis added),
"The [TFEU] prevents central banks from extending an urgent provision of liquidity to an institution in difficulties, such as Caja Castilla La Mancha, without adequate collateral. Given the volumes that could be committed, the Spanish Treasury had to provide explicit guarantees to the issuing bank, so as not to infringe the Treaty's prohibition on central banks extending "monetary financing" to Treasuries."
This refers to the Brussels View stated earlier: monetary financing in relation to ELA is viewed through the lens of using monetary resources (i.e., creating reserves) in place of fiscal resources for “public undertakings.” Here, though, as the BoS’s governor points out, the fiscal authority’s guarantee of the lending ensured that any risk from advances to an insolvent institution was borne by the state. A review of the BoS’s handling of the CCM failure by the EC shed some light on how the EC thought about the lending and guarantees: Notably, the EC reviewed the fiscal guarantee of the BoS lending as restructuring aid, not liquidity aid, consistent with the view that liquidity advancement to insolvent firms is a form of restructuring and/or recapitalization (see EC State Aid report, paras. 141-143). Nonetheless, that restructuring aid, as it were, was guaranteed by the state (a “public undertaking”). It is no surprise then that, consistent with the BoS’s lending risk residing with the fiscal authority, the EC’s State Aid report found no violations of the prohibition on monetary financing.
The Case for Cooperation: The Role of Guarantees and the Fiscal Authority
Fiscal Authority Guarantees: Opportunities
As surveyed in a number of these recent case studies, one apparent solution to the monetary financing problem is fiscal authority guarantees. Although numerous tests and guidelines of varying complexity and abstraction attempt to answer the question of when, in difficult cases of judgment, a central bank ought to lend, central bankers themselves often do away with the head-scratching and obtain a fiscal guarantee.
Indeed, the IMF has written about the use of central bank liquidity in resolution and notes that the proper practical safeguard against monetary financing when the line between ELA and resolution financing becomes blurry is the involvement of the fiscal authority (emphasis added):
"ELA to a bank in resolution may need to be backed by an indemnity or a guarantee from the government to protect the central bank balance sheet if, for example, there is significant uncertainty over the (new) bank’s ability to repay. The central bank balance sheet should not be used for purposes other than for providing liquidity on a prudently collateralized basis, i.e., the central bank should not provide capital or unsecured loans to a bank in resolution."
Although the concern of this policy note is not liquidity in resolution, that state-guaranteed ELA is a common and accepted pathway to achieve liquidity in resolution in the EU validates various EU countries’ use of state guarantees of ELA when unsure if the bank is a going or gone concern. As researchers at the economic think tank Bruegel have pointed out, the EU’s resolution framework leaves unclear which entity should provide liquidity to banks in resolution if the bank in question is sufficiently large such that its “liquidity needs . . . exceed the size of the [Single Resolution Fund] and its backstop.” As they point out, in such cases, in practice, a state guarantee of ELA is the usual path. Indeed, the researchers conclude that “since the ECB cannot provide liquidity to banks without collateral, public guarantees need to be established.”[12] In their view, ultimately European banking union should be completed and such guarantees would then be provided fully by a euro-area fiscal body, but until then (emphasis added):
"As long as banking union remains incomplete and liquidity provisioning is still to a significant part done by national central banks through ELA, it might be appropriate to continue relying on some guarantee from the respective national treasury(ies) combined with a larger guarantee from the ESM."
This is, of course, not to say that the only use of state-guaranteed ELA is for liquidity in resolution, but it is to say that attaching a fiscal guarantee to ELA allows the tool to be used equally as a standard-fare emergency liquidity tool for solvent banks and, just as acceptably, as a wind-down tool. It is therefore an attractive intervention for EU countries dealing with banks that are difficult to categorize as going or gone concern.
As discussed, monetary financing is only a concern for independent central banks. As long as they have fiscal guarantees, central banks can be more agnostic to the recipient bank’s solvency status. Once the fiscal authority agrees to underwrite the lending, any resulting taxpayer losses would not be a matter of monetary financing, but rather of fiscal policy. Should a government decide that it wishes to use the fiscal resources at its disposal to recapitalize a bank, that is its prerogative.
However, a number of problems remain.
Fiscal Guarantees as Insufficient
There are two distinct reasons that fiscal guarantees might not be sufficient for avoiding legal infringement allegations in the EU.
First, in the EU, a fiscal guarantee per se may be insufficient to avoid charges of monetary financing (or illegal State Aid, depending on the narrowness of one’s legal interpretation). A speech by ECB executive board member Yves Mersch in 2018 provides the logic (emphasis added):
"In no circumstances would a guarantee “cure” the lack of financial soundness of a given counterparty, or the lack of collateral or a combination of the two, which is often the case in a resolution scenario. If a public guarantee can prevent insolvency, it is only after having tested its value in the markets. If a public guarantee serves to cut out the markets and replace them with central bank liquidity it would contravene the principle of an open market economy with free competition and could be seen as a circumvention of the monetary financing prohibition."
Again, this is a State Aid question, not a monetary financing question.[13] The underlying logic is that whether truly monetary financing or fiscal financing, if the government—in any of its permutations—is supplying necessary credit to an institution but not to others for an extended period, then this distorts competition and is effectively industrial policy à la finance. Industrial policy is straightforwardly outside the scope of this note. This is a State Aid question that EU policymakers must navigate separately.
Second, and potentially more broadly applicable outside the EU, is the lesson learned from the Anglo Irish and Laiki Bank cases discussed in this note. Should the fiscal authority lack the means to credibly provide the full guarantee, and should it then rely on the monetary authority to buy its debt, it would be attempting to resolve one form of monetary financing with another. In other words, capitalizing a bank through reserve-creation is one form of monetary financing but backstopping that reserve-creation with what amounts to simply more reserve-creation (through direct central bank financing of unmarketable sovereign debt) does not resolve the inherent tension. Assuming the nation in question arrived at this point by virtue of exhausted fiscal space, resources would likely need to be provided by an international lender of last resort.
Fiscal Guarantees as Impractical
Second, fiscal guarantees are of little use when they’re not available, and they’re not always de facto available for one of two common reasons: (1) the fiscal authority refuses to guarantee the lending (in which case the game is probably up); or (2) the fiscal authority in principle agrees directionally, but with some limitations that make it impractical for it to be able to issue the guarantee timely enough to avoid the failure of the institution at hand (e.g., requiring approval from the legislature).
Although of course jurisdiction dependent, there may be workarounds available for this latter challenge.
While not a case of bank insolvency and not in the EU, the Swiss National Bank’s (SNB's) recent experience with Credit Suisse is illustrative of the challenges of fiscal guarantees writ large. In that case, the SNB’s proposed ad hoc liquidity facility featured a fiscal guarantee, but that required parliamentary approval. As regulators had no time to convene the legislature, the executive used an emergency authority power to bypass the legislature (such an executive action required ex post parliamentary approval, which it received). But while in Switzerland an emergency executive ordinance allowed the executive to waive the parliamentary approval process and order a public guarantee (and in Spain we saw similar authority), such standing executive authority may not exist in all legal systems.
The particular powers vested in the executive necessary for enjoining a fiscal authority guarantee will vary widely by jurisdiction and are matters of political and legal norms. The practical upshot for authorities in crisis-fighting roles is to know those legal constraints and authorities ahead of time and to build emergency plans around them, so as to avoid unnecessary and potentially protracted legal and political debates.
Fiscal Guarantees as Lacking Credibility
Fiscal guarantees are of little use if they lack credibility. Owing to conditions commonly associated with lending to potentially insolvent firms, adverse selection with respect to credible fiscal guarantees is likely. A common reason for a central bank to lend to a potentially insolvent firm is that the firm is systemically important and therefore often—though not always—very large relative to the financial system and/or economy. In such cases, the larger the size of the recipient institution relative to the size of the fiscal authority’s resources, the less credible a fiscal authority guarantee is. As a result, in the most systemic of situations, fiscal guarantees’ credibility may be at their weakest. While I am unaware of any instances in which a fiscal guarantee of central bank lending was deemed to lack credibility, we know from existing research that fiscal guarantees in general (usually in the form of bank liability guarantees) often suffer from this flaw.[14]
Use of Nonfiscal Authority Guarantees
Last, while fiscal authority guarantees are the most common form of guarantee observed in modern financial crisis interventions, they are not the only ones. Guarantees can be issued by government agencies, sovereign wealth funds, deposit insurers, state-sponsored firms, or private firms or consortia. For legal, political economy, and administrative reasons, such nonfiscal authority guarantees may be attractive. For example, they can avoid lengthy political processes, provide positive market signaling effects, and keep risk off the fiscal authority’s balance sheet. However, such guarantees may also face an uphill credibility battle. Further, particularly in times of systemic stress, those guarantees can be procyclical. For example, if a deposit insurer guarantee were called upon, the deposit insurer could be forced to levy an ad hoc fee on its members to fund the payout, resulting in contagion and systemically reducing the capital of the financial system.
Private-sector guarantees of central bank lending are most likely to take the form of guarantees issued by private deposit insurance groups or banking industry groups.[15] In the case of Denmark's Roskilde, the first tranche of the central bank emergency lending was guaranteed by a private deposit insurer.
Considerations for Policymakers
In general, central bankers and financial policymakers should be aware of the risks involved in the extension of emergency lending with respect to monetary financing—and should be mindful of solvency regulations as expressed in their own guidance (and, in the Eurosystem, to applicable ECB guidance). While not a substitute for collateral, and insufficient by itself to ensure safe harbor, fiscal guarantees may provide some protection from monetary financing allegations specifically. When solvency is in question and the intervention may straddle the line between the standard LOLR function and restructuring and resolution liquidity, involvement of the fiscal authority may be prudent. In the case of depleted fiscal resources, recourse to a supranational organization (e.g., ESM, IMF) may be necessary.
It is perhaps noteworthy that while the legal codification described above of monetary financing with respect to the LOLR role is particular to the EU, the principles apply broadly: It is often the case, even in regimes not explicitly prohibiting the extension of central bank advances to insolvent entities, that those central banks often take recourse to the fiscal authority for guarantees (variously for political insurance or balance sheet protection).[16] Fiscal guarantees are obtained (or, at the very least, requested) with frequency, both within and outside the EU. In other words, while the legal motivation for fiscal guarantees of emergency lending is peculiar to the EU, the resulting policy intervention is a regular feature of emergency lending and something for which central bankers and their fiscal counterparts should have a “break the glass” plan.
