3. Central Banks as Balance Sheets of Last Resort: The ECB’s Monetary Policy in a Flow-of-Funds Perspective

The term ‘non-standard’ (before the introduction of quantitative easing), rather than ‘unconventional’, also reflects this specificity.

The questions whether the ability of the central bank to reduce risk in the economy depends on its mandate and whether the mandate should be changed are much debated. See for instance the volume coordinated by Bordes and Raymond (2014). Aglietta (2014) argues in this volume that monetary policy must be overhauled to recognise its multiple objectives and should use multiple instruments accordingly. By contrast, Cobham (2015) shows that whether the central bank pursues a single objective of inflation or constrained multiple objectives does not make any real difference for their economic performance (in terms of inflation and economic growth).

The ECB tends to call ‘non-standard’ its monetary policy measures aimed at supporting the proper transmission of its standard, policy rate, measures, reserving the wording ‘unconventional’ for active balance sheet expansion measures taken once the lower bound of policy interest rates was reached.

This was acknowledged by the ECB President in September 2014 with a further cut in policy rates bringing the deposit facility rate to -0.20 basis points.

To recall, credit easing—in distinction to quantitative easing—is a way to offer support to the economy that involves a change in the central bank balance sheet composition rather than in its size. The literature focuses on the asset (or liability) composition of the central bank balance sheet, or the composition of the assets eligible as collateral in monetary policy operations. For a review see, for instance, Durré and Pill (2012). At the time this chapter was finalised, the ECB (2015) published an article that considers quantitative easing and credit easing as active balance sheet policies and further discusses the notion of ‘contingent balance sheet policies’.

Even in normal times, the operational framework of central banks entails tools pertaining to the lender of last resort function: the standard liquidity-providing standing facility and emergency facilities such as the Emergency Liquidity Assistance (ELA) in the case of the ECB. During the crisis, the function of lender of last resort was additionally realised in the case of the Federal Reserve through special lending facilities in 2007–2008, and in the case of the ECB through the change to a fixed rate full allotment procedure in its lending operations, by which the Eurosystem central banks would supply the liquidity demand of banks in full, provided they had adequate collateral.

There are many ways to define the leverage ratio. The indicator in this paper takes a macroeconomic perspective based on national account conventions. It should not be confounded with regulatory leverage ratios used in Basel III based on business accounting, which involve Equity/Assets. Recognising that assets are equal to liabilities and liabilities are the sum of debt plus equity, the leverage ratio used in this paper is conceptually equal to 1 minus the Basel III leverage ratio.

The figure extends, with a slightly different presentation, Fig. 3.10 of Cour-Thimann and Winkler (2012).

This comparison also illustrates how the practice of marking to market (favoured by international accounting rules) can hinder the proper assessment in real time of the mispricing of risks at the origin of the financial crisis.

The increase in total assets in Fig. 3.1 also corresponds to a more than proportional increase in debt, and thus to an increase in the debt-to-assets ratio, since assets are equal to liabilities, which as mentioned are themselves the sum of debt plus equity, and central bank equity was broadly unchanged in the crisis.

See, in particular, Girón and Rodríguez-Vives (2015a, 2015b) for a discussion on the interaction between leverage in the public sector and in the private non-financial sector.

The targeted lending operations were aimed at providing incentives for banks to use the central bank funds for lending to the real economy (with lending for house purchases excluded from the calculation of allowances).

See ECB (2013) for a discussion on the link between the central bank’s credibility and its financial strength. In addition, fulfilling the mandate can only be done with an objective of sustainability, otherwise it would foster adverse side effects. For instance, too large a balance sheet expansion through central bank liquidity support to banks could foster moral hazard in the financial sector, including a renewed mispricing of risks and asset price bubbles, and prepare the ground for financial and price instability further down the road.

Alternatively, net assets (A-D) can be increased by replacing debt with equity as source of financing.

A simplified balance sheet presentation could easily show the difference in the impact of the central bank measures on the sector of financial institutions in the Euro Area and the USA.

A parallel can be made with the case of the policy rates instrument, whereby a cut (hike) does not necessarily have an expansionary (contractionary) effect on the economy, depending on whether this cut (hike) takes place in a situation where the natural rate of interest increases, declines or remain stable.

The US Federal Reserve’s various backstop measures to the financial system and specific financial institutions undertaken in late 2008 and early 2009 also entailed some contingent elements.

Furthermore, the indexation of the interest rate in longer-term refinancing operations on the future main refinancing rate over the lifetime of the operations adds to the contingent character of the fixed rate full allotment measure. The indexation feature was introduced in December 2009 and kept in most of the subsequent refinancing operations of a maturity above 3 months until its waiving in June 2014 with the introduction of the targeted longer-term refinancing operations. Indexation implies in particular that a reduction in the policy interest rate is immediately translated to reduced costs for the borrowing bank over the remainder of the outstanding operations. It thus constitutes a form of contingent easing (while operating as well in the other direction of a contingent tightening in case of an increase in the policy interest rate).

The announcement of the OMTs followed the statement of President Draghi (on 26 July 2012) that the ECB would do ‘whatever it takes, within its mandate’ to preserve the Euro. The OMTs were designed to address unwarranted redenomination risk premia that had appeared in bond markets.

The period during which fixed-rate full allotment would apply was extended many times during the crisis: this extended the promise given to banks that they could borrow unlimited amounts of central bank liquidity, limited only by their capacity to pledge assets as collateral in exchange. Nevertheless, the endogeneity nature of the measure still has an indirect exogenous component: the central bank sets the criteria on which banks can obtain liquidity, such as the collateral eligibility, and therefore can influence the volume of liquidity demanded by banks.

The ECB’s targeted longer-term refinancing operations introduced in June 2014 depart from the full allotment mode but also include some endogenous elements (the provision of central bank liquidity to individual banks is made dependent on their actual amounts of lending to corporates).

The amount of assets purchased is determined exogenously and assets are held for an indefinite period of time. That period is only limited by the maturity of the assets purchased, which was also dramatically lengthened compared to pre-crisis levels (e.g., to as much as 30 years for certain mortgage-backed securities). It would take a decision from the central bank to reduce its leverage in this case, contrary to the automatic reversibility attached to the lending operations in the case of the Eurosystem.

In the same way as the assessment of public finances needs to account for the government’s contingent liabilities (e.g., related to social security systems or to the outcome of certain multilateral arrangements), the assessment of the central bank’s balance sheet size can a priori be enhanced through taking account of the central bank’s contingent liabilities.

TARGET2 is the payment system operated by the Eurosystem. It is an acronym for Trans-European Automated Real-time Gross settlement Express Transfer system (in its second generation).

For the link between Target balances and the crisis in the Euro Area, see for instance Cour-Thimann (2013).

This point could be illustrated using a reproduction of Fig. 3.2 at the national level for the individual Euro Area countries, or for the groups of countries with a positive Target balance and those with a negative Target balance on their respective central bank balance sheets.

The delimitation of national banking systems and the relevance of national borders for assessing cross-border payments is blurred in a financially integrated area: banks participating in monetary policy operations (and in TARGET2) at different Euro Area NCBs may be part of the same banking group. In addition, the cross-border flows may not reflect the geographical location of the activities underlying the transactions.

Indeed, in the case of a central bank having its own currency, a massive injection of liquidity in a situation of large payment outflows would lead to a depreciation of the currency. This would in turn trigger a rebound in competitiveness and thus reduce the outflows and the very need for central bank liquidity, thereby limiting the extent of the leveraging up.

One could be tempted to infer that the larger leveraging-up of NCBs in crisis countries would imply a transfer of risks across borders, at the expense of the countries whose NCBs leveraged less. This is not an appropriate reading, as one also needs to consider the origin of the risks that were transferred to the central banks’ balance sheets (see Cour-Thimann, 2013, for an explanation).

See, for instance, the presentation of the transmission mechanism of monetary policy on the official websites of the ECB and the US Federal Reserve.

The financial risk attached to the monetary policy operations includes a credit risk (as in the case of lending operations), a market risk (as in the case of assets purchased or underlying collateralised operations, the value of the assets evolving with the market), and a liquidity risk (as in the case where the central bank would liquidate assets). The liquidity risk depends on whether the asset acquired is tradable/marketable or is an illiquid claim (such as in the case of the assets being individual bank loans).

The assets included mainly government bonds, government-guaranteed mortgaged backed securities, commercial paper, and securitised loans.

For an analysis of the link between bank leverage and the credit cycle in the Euro Area, see, for instance, Girón and Mongelluzzo (2013).

From a flow-of-funds perspective, in the Tobin-Brainard portfolio balance tradition, the demand-supply balance for different assets matters for monetary policy transmission. With an application to the Euro Area sectoral data, Jaccard (2013) shows, in a dynamic stochastic general equilibrium model, that the economic crisis in the aftermath of the financial crisis was primarily caused by liquidity factors, with strong non-linear effects on the corporate sector.

Operating in the short-term domain puts the focus on liquidity provision and liquefying less liquid claims in exchange for, and as a close substitute for, money (in a quantity perspective). The exchange of money (reserves) for longer-dated paper is typically assessed with reference to the so-called term premium (in a price perspective). Such exchanges would imply intervening in the longer-term allocation of savings and capital and thereby introduce market distortions (e.g., distortions are seen to be larger in the case of longer-dated corporate bonds than in the case of short-dated claims, such as commercial paper).

Covered bonds and collateralised lending are both de facto collateralised instruments, with a priori two differences: (1) in the regular lending operations the collateral is generally liquid, while the covered bonds, which are themselves liquid, contain illiquid pledges (also originated by the borrower); and (2) the maturity is a priori longer in the case of assets purchased outright than in lending operations. However, those differences were attenuated by the extension of the range of eligible collateral in the lending operations to less liquid assets and of their maturity.

Unlike covered bonds, ABSs do not remain on the banks’ balance sheets, and thus they do not benefit from double recourse protection whereby the claim would be both on the issuer and on the collateral.

The motivation behind the SMP, which was active during 2010–2011, differed from that of quantitative easing such as in the case of the US Federal Reserve’s Large Scale Asset Purchases (LSAP). The Federal Reserve purchased federal bonds with the intention of lowering term premia and long-term yields and, hence, providing additional monetary accommodation. By contrast, the ECB aimed at countering dysfunctional markets and intervened in selected government bond markets at the national (sub-federal) level that were regarded as important elements in monetary policy transmission. This motivation was also paramount for the OMTs launched in the summer of 2012 (see above). Unlike the SMP, potential purchases under the OMTs were ex ante unlimited, in a maturity bracket limited to up to 3 years, and made conditional on countries entering an adjustment programme in order to ensure solvency.

Even if the ABSs are often issued by banks or associated special purpose vehicles, they package loans or other claims (credit cards, leasing, and possibly trade receivables) that are vis-à-vis non-financial private agents. ABSs are thus an intermediate case in the matrix.

A trade receivable or payable (i.e., trade credit) is a bookkeeping entry for an intercompany sale not yet paid. A trade bill is the corresponding paper, which can be used as a payment instrument vis-à-vis a third party.

See, for instance, Winkler and de Rougemont (2013), for an exposition of the use of sectoral accounts in the Euro Area.

See Carpenter, Demiralp, Ihrig, and Klee (2013) for a flow-of-funds based assessment of the asset purchases of the US Federal Reserve, as well as Thornton (2012) for a sceptical view on the portfolio balance channel in the case of government bonds. Bertaut, DeMarco, Kamin, and Tryon (2011) estimate asset demand equations for bank deposits, treasury securities and corporate debt in a portfolio balance model.

This is also documented by the econometric literature. For instance, exploiting data on transactions among banks and between banks and the Eurosystem, Giannone, Lenza, Pill and Reichlin (2012) find a small but significant effect of the ECB’s increased role as a financial intermediary on bank lending.

Such operations have some similarity with the Bank of England’s ‘funding for lending’ scheme. The first targeted operation on 18 September 2014 faced a relatively low demand of 82.6 bn Euro, despite its unprecedented long maturity of four years.