central banks as a funding backstop – Bank Underground

Matthieu Chavaz, David Elliott and Win Monroe

Large-scale provision of long-term funding to banks has become a central bank tool to support credit supply during downturns. However, scholars have worried that allowing banks to rely on public funding could create moral hazard and crowd out private funding. In a recent paper, we address these concerns by showing that central bank and private funding can be complements rather than substitutes. The mere availability of central bank funding improves private wholesale funding conditions, thus supporting lending without central bank funding being used. This ‘equilibrium’ effect makes central bank funding more powerful than previously thought. Finally, the fact that central bank funding comes with strings attached can help to explain why it is an imperfect substitute for private funding.

Public and private funding: substitutes or complements?

To test the effect of central bank funding on lending, we exploit the announcement of the Bank of England’s (BoE) Funding for Lending Scheme (FLS), which was launched in 2012 in response to stress in wholesale funding markets during the eurozone debt crisis. We study the FLS rather than more recent schemes because it was not launched alongside other monetary policy measures, which facilitates empirical identification. A subsequent amendment to the FLS also provides a clean laboratory to test the effect of strings attached to central bank funding.

Under the FLS, UK banks could get four-year loans from the BoE. To incentivise banks to use this funding to lend to the real economy, the quantity and price of funding available were conditional on banks’ lending to households and firms. The idea to make central bank funding conditional on real economy lending was subsequently adopted by the ECB’s targeted longer-term refinancing operations (TLTROs), and by several central banks during the Covid crisis.

We start by examining the drivers of participation in the FLS. If FLS funding is primarily a substitute for wholesale funding, then banks more exposed to stressed wholesale funding markets should borrow more from the scheme. We observe the exact opposite: specifically, banks who use more wholesale funding (as a share of total assets) borrow less from the scheme relative to their size. This raises the possibility of a complementarity between central bank funding and private funding: the availability of central bank funding rejuvenates wholesale funding markets, and this disproportionately benefits banks more reliant on wholesale funding. These banks then need to borrow less from the FLS. Consistent with the idea that the availability of central bank funding helps to rejuvenate private markets, measures of banks’ wholesale funding costs fell sharply after the FLS was announced (Chart 1).

Chart 1: Wholesale funding costs for UK banks

Note: The chart shows measures of long-term wholesale funding costs for UK banks.

Sources: Bank of England and Barclays.

These patterns point to a potential ‘equilibrium effect’, whereby central bank funding stimulates bank lending even if the funding is not actually used. To investigate this effect, we use the Product Sales Database (PSD) of residential mortgage originations. Using loan-level regressions covering the period January 2012 to June 2013, we estimate how the interest rates charged by different banks after the announcement of the FLS in June 2012 vary with the bank’s exposure to wholesale funding (measured before the announcement). We control for the effect of direct participation in the FLS using a measure of the bank’s FLS borrowing allowance based on its pre-FLS balance sheet. We control for developments in the euro-area crisis using euro-area bank and sovereign CDS spreads. And we control for changes in credit demand using product-time fixed effects, which imply that we compare how different banks change their interest rates over time for a given mortgage product category.

We find that after the FLS announcement, banks with a higher exposure to wholesale funding charge lower mortgage rates – regardless of whether they draw on the FLS itself. In fact, we find that the aggregate impact of this ‘equilibrium effect’ on mortgage rates is significantly larger than the impact from banks’ direct participation in the FLS. This is partly because the large banks that dominate the UK mortgage market tend to have larger wholesale funding exposures, and smaller participation in the FLS. This suggests that the overall impact of central bank funding schemes can be significantly more powerful than previously thought.

Why does the FLS announcement reduce private wholesale funding costs? One possibility is that having access to an outside option to borrow from the central bank increases banks’ bargaining power in private funding markets. Alternatively, this central bank funding ‘backstop’ could lower banks’ funding liquidity risk, leading to a reduction in the risk premia charged by private wholesale lenders. Our results are consistent with this second hypothesis. We find that the equilibrium effect is driven by banks’ exposure to short-term wholesale funding (which creates more funding risk) rather than long-term wholesale funding (which is a closer substitute for FLS funding). Also in line with this idea, the equilibrium effect weakens when a second FLS funding scheme is announced, at a time when wholesale funding stress has subsided.

Untying strings attached

In addition to indirectly benefiting from central bank funding, the equilibrium effect could also allow banks to avoid any non-pecuniary costs associated with using this funding directly. One such cost is that the conditions attached to central bank funding might constrain banks’ ability to deploy it towards the most profitable uses. Our setting provides an ideal laboratory to test this idea because conditionality was a central innovation behind the FLS, and because subsequent changes to the program created two important shocks to the reach of this conditionality. 

First, in April 2013, the BoE announced a second wave of FLS funding (‘FLS2’). As for the original ‘FLS1’, the amount a bank could borrow from FLS2 was conditioned on its real economy lending. During the transition period between the two schemes, new mortgages could still be funded with FLS1 drawings, but would also generate ‘initial allowances’ for future FLS2 drawings. These future drawings could be used to fund any asset, thus constituting unconditional funding. In contrast, once FLS2 begins, borrowing allowances could only be unlocked by originating new loans to households or firms, thereby constituting conditional funding. Therefore, if banks find conditionality costly, they should have an incentive to unlock future unconditional funding by lowering rates and originating more mortgages during the transition period. In line with this idea, we find that during the transition period, banks more reliant on FLS funding reduced rates more on new mortgages.

Second, in November 2013, the BoE amended the terms of FLS2. In order to incentivise corporate lending, mortgage lending during 2014 would no longer increase FLS2 borrowing allowances. We find that this amendment reduces the impact of FLS participation on mortgage rates, consistent with the increased conditionality of FLS2 funding reducing its impact on lending. In addition, during the short time window before the amendment becomes binding, banks more reliant on FLS funding reduce mortgage rates further, consistent with an attempt to secure future FLS borrowing allowances before conditionality becomes tighter.

Taken together, our results suggest that the FLS achieved its goal of improving credit conditions – and that the increase in credit supply was larger than would be suggested based on direct participation alone. Our results also suggest that conditionality matters, and that banks prefer public liquidity with fewer strings attached. This suggests a trade-off in the design of central bank funding schemes. While conditionality may help the central bank to achieve its policy objectives by targeting particular sectors, this may also weaken the degree to which central bank funding is a close substitute for private funding, which may weaken the equilibrium effect.

Matthieu Chavaz works at the Bank for International Settlements, David Elliott works in the Bank’s Monetary Policy Strategy Division and Win Monroe works at Copenhagen Business School.

If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.