The debate on the lender of last resort has been particularly topical in recent times. In Europe, for example, governments intervened to rescue their banks, acting as lenders of last resort, and subsequently found themselves in difficulty, paradoxically needing a lender of last resort themselves. Aid from the International Monetary Fund (IMF) and the creation of the European Stability Mechanism (ESM) have also moved in this direction. This contribution takes up the points developed by Goodhart in his article « Myths about the Lender of Last Resort, « with the aim of clarifying this concept and the debates surrounding it.
The concept
The lender of last resort (hereinafter referred to as LoL) is an institution to which an entity in difficulty (bank, state, etc.) turns when it can no longer obtain help elsewhere. This concept is generally associated with banks or states, as the failure of these entities would have very damaging economic and social consequences.
The concept of lender of last resort is often attributed (wrongly?[1]) to Bagehot, a19th-century British economist, who devoted an entire section to this subject in his book Lombard Street (1873). Goodhart emphasizes three of the proposals in Bagehot’s work on the LDR: lending without restriction, at a high interest rate, against high-quality collateral. These three principles are, in a way, those that best summarize the ideal/initial terms of action of the lender of last resort.
What is a lender of last resort operation?
It seems common sense to say that a lender of last resort, when it comes to rescuing a bank, is not supposed to bail out an institution that is experiencing solvency problems, but rather a bank that is experiencing liquidity problems. The distinction is important in theory, but very complicated to make in practice. The real value of a bank’s trading book, i.e., the value of its balance sheet if all its items are recorded at market value, is very difficult to obtain. Complex products on banks’ balance sheets are difficult to value at market value, and the volatility of certain other products makes the resulting estimate unreliable. In addition, the lender of last resort must generally act quickly to avoid triggering contagion, which makes it very difficult, if not impossible, to assess the real balance sheet of banks.
According to Goodhart, the specificity of a lender of last resort lies in its bilateral nature, which contrasts with large-scale operations such as open market operations. Quantitative easing, credit easing, or exceptional refinancing operations such as LTROs (Long Term Refinancing Operations) therefore do not fall within Goodhart’s definition of a lender of last resort operation. The reason for this is that it cannot be said that all the banks concerned are turning to the central bank in its role as lender of last resort. On the other hand, when a bank goes to the discount window (the term used for the Fed) or borrows from the marginal lending facility (the term used by the ECB for the same operation conceptually), it can be said with certainty that it is turning to the lender of last resort, since it can no longer obtain funds on the market at reasonable rates, as the market is too suspicious of it (Goodhart refers to « suspicion of insolvency »). Reputational and financial costs generally force banks to resort to this last resort only when it is impossible to go through the interbank market.
The IMF’s role as lender of last resort
So far, we have discussed the case of banks, not that of governments. If governments encounter serious budgetary problems, several scenarios are possible.
In theory, the central bank could intervene by creating money. This solution, which is prohibited in Europe, is synonymous with inflation and, in the most difficult cases, can even lead to hyperinflation. This is why countries such as Greece, deprived of their monetary sovereignty and unable to increase their internal resources or reduce their spending without causing further panic in the markets, have had to seek external assistance. The IMF, assisted by other ad hoc institutions, has therefore intervened.
In other cases, the central bank cannot intervene, since the currency in which the debts are denominated is not the local currency. It could, at best, create new money to exchange for foreign currency, but these operations would be largely offset by the resulting depreciation of the local currency and would therefore ultimately be ineffective. This is where the IMF logically comes in.
IMF – Central banks: what are the differences in terms of PDR?
The differences between these two institutions can be summarized as follows:
– Money creation power: central banks have this power, but the IMF does not, since the amount of « local » currency available to the latter (SDRs) depends on the contributions of its members.
– Seniority status: this is discretionary in the case of central banks, but formal in the case of the IMF. The risk of losses is therefore lower for the IMF.
– Speed of action: The difference in the speed of reaction between these two institutions can be explained by their shareholding structure. For Goodhart, when it comes to the solvency of central banks , « what stands behind the liabilities of the CB is not the capital of the CB but the strength and taxing power of the State, » which means that the latter can generally expect to be recapitalized quickly by the State in the event of losses. For the IMF, things are different in that all shareholder representatives must agree. It is this slowness in the decision-making process that has led some, such as Keleher, to declare that « the IMF cannot act quickly enough to serve as a LOLR. »
What other models of LPR exist?
The main cost associated with lender of last resort interventions is the moral hazard created by these operations: bailing out a bank or a state that has not been responsible in the past sets a damaging precedent for the future. Other models of LLR interventions can therefore be considered. The debate on these different models is very broad when it comes to banks, and it would be difficult to cover it in full here. Suffice it to say thatmutual insurance mechanisms between banks and monetary/governmental institutions, by requiring banks to set aside funds to pay in the event of failure, encourage these actors to take greater responsibility and thus reduce moral hazard.
Is it possible to envisage a PDR model focused solely on central banks and without the IMF? It could be argued that countries could borrow the necessary currencies directly from their neighbors. However, there is good reason to believe that such aid would be political and geostrategic rather than based on economic factors that are more objective in this context. This problem, highlighted by Goodhart, therefore raises the issue of the IMF’s indispensable economic role as a (theoretically) neutral actor free from geopolitical influences. Without the IMF, the Fed would certainly be in a position to assume these responsibilities, but its role would inevitably be much more politicized.
Conclusion
Goodhart’s article has the merit of clarifying a debate that often lacks clarity. The concepts of lender of last resort and the categorization of an intervention as a « lender of last resort intervention » have given rise to numerous discussions from which no single rule can be drawn.
While the distinction between illiquidity and insolvency may need to be made in theory, it is difficult to establish in practice, both because of the short time frame available to the PDR to intervene and because of the complexity and high volatility of the products on banks’ balance sheets. The IMF performs this role differently from central banks or governments, both because of the nature of its interventions (different target institutions) and because of the rules governing this institution.
While the debate on the existence of lenders of last resort is essentially semantic, the debate on how they should be organized is much more profound. A changing economic and political environment calls for constant re-examination of the optimal system for organizing lenders of last resort. The creation of the European Stability Mechanism, developments in financial regulation, and the quest for greater independence for central banks are thus going hand in hand with a new organization of the LRS system.
Reference:
Goodhart, « Myths about the Lender of Last Resort, » 1999, International Finance 2:3
Notes
[1] Henry Thornton, in his work « The Paper Credit of Great Britain » (1802), had already developed most of the points appearing in Bagehot’s work.
[2] « Lend freely » in the original version can be interpreted in several ways. Personally, I would interpret it as « do not restrict the loans granted, » given that at the time restrictions could come into play due to a currency backed by gold.
[3] In practice, as we saw with Dexia, the French and Belgian governments dismantled the bank’s activities in order to keep only the most profitable ones, rather than « saving » the institution itself, as the French government had done with its main banks immediately after the crisis. In theory, Bagehot does not say this: « advances should be made on all good banking securities and as largely as the public ask for them, » meaning that as long as the collateral provided is of good quality, the central bank should lend. It should be noted that one could argue that there are reasons why certain banks, particularly those that are « too big to fail, » should be bailed out even if they are insolvent, for political or economic reasons (risk of contagion). However, it could be countered that these cases are perhaps marginal, as in most cases clearly insolvent institutions are at best dismantled and at worst allowed to go bankrupt.
[4] It should be remembered that a bank will simply be in a situation of illiquidity when it is unable to meet its loan maturities at a given moment, without necessarily being insolvent (the problem of illiquidity, if resolved, is therefore a temporary problem).
[5] For this to happen, the state would have to be structurally insolvent, resorting only to money creation to cover its deficit, and inflation would have to be at a level that maximizes seigniorage revenues.
[6] This statement is open to debate (it is beyond the scope of this article), and is of course simplistic and intended only to illustrate the point.
[7] Some have also described the ECB’s SMP as a PDR intervention, which the ECB obviously denies. Insofar as this operation is not carried out bilaterally between Greece and the ECB, it cannot be described as a lender of last resort according to Goodhart’s definition, even if it has certain virtues. Nevertheless, this points to a gap in Goodhart’s thinking here, namely that he considers only central bank interventions with commercial banks and does not extend the semantic debate to central bank interventions with their governments. The ECB’s recent OMT, which in my view could be likened to a PDR operation (insofar as it would help to relieve states of their short-term debt in an exceptionally poor economic climate), would not, for example, be considered as such by Goodhart if we stick to his definition.
[8] Seniority defines the priority in the repayment of a debt vis-à-vis other creditors.
[9] This is often enshrined in law; see Stella, P., Lönnberg, A. (2008), « Issues in Central Bank Finance and Independence » for a study on this subject.
