TLTRO III and bank lending conditions

Prepared by Francesca Barbiero, Miguel Boucinha and Lorenzo Burlon

Published as part of the ECB Economic Bulletin, Issue 6/2021.

1 Introduction

Targeted longer-term refinancing operations (TLTROs) play a key role in preserving favourable bank financing conditions for households and firms, thereby contributing to inflation reaching the ECB’s target of 2% in the medium term. The operations are part of a broad set of complementary policy instruments, which include asset purchases, negative interest rates and forward guidance.^[1] Since their inception in 2014, TLTROs have supported the transmission of monetary policy by incentivising lending through their targeting feature and by providing a reduction in bank funding cost, which has been instrumental in avoiding a deterioration in lending conditions that would have otherwise occurred. The third series of the TLTROs (TLTRO III) was introduced in early 2019. The initial announcement of TLTRO III in March 2019 reassured markets about the extension of the pre-existing TLTRO II. The operations were intended to stave off “congestion effects” in bank funding markets that would have otherwise materialised because of the need to replace expiring TLTRO II funds. The operations were recalibrated in September 2019 to preserve favourable bank lending conditions, ensure the smooth functioning of the monetary policy transmission mechanism and therefore further support the accommodative stance of monetary policy. From the start of the coronavirus (COVID-19) crisis, the recalibration of this tool was, thanks to its design and the role of the euro area banking system in the monetary policy transmission mechanism, an integral part of the ECB’s policy response to ensure favourable borrowing conditions for firms and households during the pandemic.

TLTRO III provided ample liquidity at attractive rates to address the emergency liquidity needs of households and firms induced by the pandemic. The ECB’s monetary policy response to the COVID-19 crisis involved two main tools. First, asset purchases supported favourable financing conditions for the real economy in times of heightened uncertainty, both through an additional envelope under the regular asset purchase programme (APP) and via the launch of the pandemic emergency purchase programme (PEPP). Second, the recalibration of the existing TLTRO III operations helped banks secure funding at favourable terms to support access to credit for firms and households.^[2] The Governing Council’s decisions of 12 March^[3] and 30 April^[4] 2020 have secured the transmission of monetary policy via banks at times of elevated uncertainty and high liquidity needs by expanding banks’ borrowing allowance under TLTRO III from 30% to 50% of the eligible loan book (providing an additional leeway of approximately €1.2 trillion) and reducing the interest rate applied on these operations to a rate as low as -1% until June 2021 for banks fulfilling the lending requirements. These decisions also enlarged the set of assets eligible to collateralise the borrowing under TLTRO III and enhanced banks’ flexibility of repayment options and participation modalities across operations. The Governing Council’s decisions of 10 December 2020^[5] further widened the borrowing allowance to 55% and prolonged the period in which banks could secure a rate as low as -1% to June 2022, subject to additional lending requirements until the end of 2021. This served to shelter borrowing conditions from the ripple effects of the pandemic.

The magnitude of the pandemic shock, the broad-based policy response and the attractive design of TLTROs (after the various recalibrations) resulted in one of the largest liquidity injections by the ECB directly into the euro area banking sector, bringing the total uptake to €2.2 trillion as of June 2021, thereby providing substantial support to the euro area throughout the entire pandemic period. The monetary policy response to buffer the impact of the pandemic on borrowing was complemented by policy support from other policy domains, ranging from microprudential and macroprudential policy via capital relief measures, to fiscal policy via extensive use of government guarantees and moratoria. The favourability of TLTRO conditions, together with the broadened eligibility of assets that could be pledged as collateral (see Box 1), the capital space and loan demand reinforced by other policies, enabled euro area banks to participate widely in the TLTRO III programme, leading to the largest participation in Eurosystem refinancing operations so far. The overall take-up exceeded €1.5 trillion after the June 2020 operation and subsequent operations brought it up to €2.2 trillion as of June 2021 (Chart 1). This article studies how, and by how much, this targeted longer-term central bank funding has affected bank lending conditions.

This article is organised as follows. Box 1 sheds light on the role of the collateral easing measures. Section 2 explains how the stimulus from TLTRO III is transmitted via banks to the final borrowers. Box 2 focuses on the programme’s impact on money market rates. Section 3 documents the extent to which TLTRO III has eased bank lending conditions, considering its efficiency vis-à-vis other policy measures and the scope for potential side effects. Box 3 discusses the impact of TLTRO III on excess liquidity. Section 4 provides this article’s overall conclusions.

2 The transmission of TLTRO III to borrowers via banks

The transmission of TLTRO III to bank lending conditions operates through a variety of channels. TLTROs target bank lending to non-financial corporations and to households for purposes other than housing. The operations also provide a reduction in the funding cost of euro area banks, which activates a bank lending channel that can lead to lower lending rates and higher lending volumes. Moreover, as detailed below, additional channels complement the transmission of TLTROs, stimulating lending and supporting a decrease in borrowing costs for firms and households.

The incentive scheme embedded in the TLTROs stimulates bank lending to specific sectors, leading to increased competition in lending markets. In contrast with standard non-targeted operations, TLTRO III offers more advantageous pricing on borrowed funds; this pricing is conditional on participants achieving lending targets. This increases participants’ propensity to lend, especially in situations where uncertainty leads to a risk of tightening in access to funding for banks, which would normally induce them to tighten credit standards. As participants aim to lend more, competitive pressures in lending markets increase, which also induces non-participants to ease lending criteria to protect their market share.^[9] The increase in competition is also one of the reasons why lending targets are carefully calibrated based on lending volume projections. Too high a bar to clear in order to achieve favourable TLTRO pricing could induce disproportionally competitive behaviour and potentially encourage excessive risk-taking. At the same time, setting a lending target that is too demanding could discourage some banks from participating and consequently reduce the effectiveness of the programme. At the same time, an insufficiently ambitious lending benchmark would compromise the effectiveness of the targeting feature of the operations.

TLTROs provide both direct and indirect reductions in the cost of funding of euro area banks. Bank funding costs have moderated following TLTRO announcements (see Chart 2). TLTROs compress banks’ funding costs by offering long-term borrowing from the central bank at attractive rates. This can be used to replace more expensive sources of market funding. The resulting compression in the cost of funding of banks has a direct and an indirect component. For each participating bank, the direct cost reduction stems from the substitution of more expensive funds. For banks not directly participating in the programme, the reduction in the cost of funding is indirect and originates from a positive externality: since banks participating in TLTROs are likely to cancel or postpone their bond issuance, the resulting decrease in bond supply generates a reduction in the external funding cost even for those banks that do not directly borrow under the operations. Since banks operate in a competitive market, this leads to lower lending rates and increased credit volumes, as borne out by experience of previous TLTROs.^[10] For this indirect channel to become active, participation in the programme at the aggregate level must be large, suggesting there is scope for important non-linearities to come into play, especially after the large take-up in the June 2020 operation. In this regard, the collateral policy of the Eurosystem and the decision to increase the borrowing allowance are key to enabling a large and broad-based participation, as discussed in Box 1.

TLTROs ultimately support borrowing conditions faced by households and firms by mitigating potential increases in lending rates and by providing a substantial compression in the cost of bank funding. TLTROs are normally transmitted to interest rates via their aggregate effects on the loan market. This is due to aggregate loan supply expanding enough to exert downward pressure on lending rates, which eventually decrease if loan demand is not strong enough. The recent experience with TLTROs after March 2020, when the demand for loans registered unprecedented levels, indicates the heightened relevance of at least two additional mechanisms of propagation are specific to the pandemic. First, the availability of TLTRO funds contributed to mitigating a potential increase in lending rates due to the surge in credit risk in the context of the economic disruptions brought forth by the pandemic. Second, the sharp and large increase in uncertainty about the macroeconomic outlook induced strong precautionary behaviour on the part of firms and households, which contributed to a large increase in deposit volumes. Given the relative difficulty in imposing negative rates on retail deposits, the reduction in bank funding cost provided by TLTROs contributes to preserving the smooth transmission of monetary policy by enabling banks to lower lending rates while preserving lending margins and avoiding excessive credit risk-taking.

By reducing liquidity constraints, TLTROs mitigate procyclicality that is driven by medium-term to long-term financing conditions. Since TLTROs have a longer maturity than standard refinancing operations, they contribute to easing regulatory constraints related to liquidity requirements, especially when the operations are initially implemented and their residual maturity is considerable. This allows banks to structure their liquidity composition in a less procyclical manner, issuing longer-term bonds when they are not overly expensive due to the transitory financial distress brought forth by the crisis to which the TLTROs are a response.

In addition, TLTROs inject central bank liquidity into the financial system, putting downward pressure on money market rates. TLTROs stimulate demand for central bank funding and effectively increase the quantity of excess liquidity in the system. The more attractive the TLTRO terms, the more broad-based the participation and resulting distribution of excess liquidity in the system. This reduces reliance on interbank markets for short-term liquidity needs, and thus compresses short-term rates, as shown in Box 2.

TLTROs also offer a backstop against escalating funding stress because banks can increase their recourse to TLTROs if faced with adverse scenarios. Like standard Eurosystem refinancing operations, TLTROs are demand-driven and allow potential participants to use the facility when access to market-based funding sources becomes challenging. Hence, the current price constellation for bank funding conditions is likely to reflect the effect of this option, including with regard to future refinancing conditions for banks, compressing the risk premia on bond yields and contributing to preserving favourable financing conditions.

Finally, TLTROs operate as a powerful credit easing measure in conjunction with the other active monetary policy instruments. The negative interest rate policy reinforces the incentive for loan origination by lowering the returns on risk-free assets, and at the same time maintaining the availability of TLTRO funding at very low rates, while the two-tier system for excess reserve remuneration mitigates potential frictions associated with the downward rigidity of retail deposit rates. Forward guidance keeps the intermediate segments of the risk-free curve, used by banks to price loans, at low levels. This stimulates credit demand and is complemented by the long term of TLTRO funds which helps compress the term premium of bank funding. Purchases of government bonds lower the returns on alternative uses of TLTRO funds, which helps to convey TLTRO liquidity towards the targeted sector. Purchases of corporate bonds narrow corporate bond spreads, thereby providing a direct channel of transmission from monetary policy to the real economy and creating space on banks’ balance sheets for extending more loans to firms not directly exposed to the asset purchases, such as small and medium-sized enterprises, via the use of TLTRO funds.^[11]

3 Impact on bank lending conditions

Evidence from bank balance sheet data

TLTRO III allowed banks to accommodate the large-scale increase in credit demand triggered by the pandemic. The growth in bank credit to the private sector has been substantial among TLTRO III participants since the start of the pandemic in the 13 months covered by the lending performance benchmark introduced in March 2020 (see panel (a) of Chart 3).^[12] Loan growth has continued for participants in TLTRO III since the start of the additional lending performance evaluation period in October 2020^[13], especially if compared with the figure for non-participants over the same period (see panel (b) of Chart 3). Together with an increase in the maturity of loans, which was also favoured by the availability of public guarantee schemes, this suggests that the operations helped banks to meet the increased credit demand in a sustainable way, allowing for a rotation from the initial emergency credit demand towards lending for longer-term purposes, including investment.^[14] At the same time, acquisitions of government securities increased initially, reflecting increased issuance and liquidity demand by governments in order to finance the public support measures. Following this initial period, the net flows into government securities since October 2020 were negative, which is consistent with banks favouring origination of loans to the private sector over potential acquisition of government securities and also reflects the large absorption of these securities by asset purchases.

The temporary build-up of liquidity by individual TLTRO borrowers may also point to the credit expansion that these lenders may provide in the future. There has been a major increase in Eurosystem deposits since the start of the pandemic. This increase at the aggregate level is entirely mechanical, as the liquidity injected through the TLTROs (and the asset purchases that were conducted in parallel) circulates within the closed system of banks that have reserve accounts with the Eurosystem and is not reduced when banks expand credit to firms and households, which are not part of this closed system.^[15] Moreover, the still sizeable build-up of liquidity deposited with the Eurosystem since the introduction of TLTRO III is the result of a range of factors, as discussed in Box 3.^[16] At the bank level, it is important to note that euro area banks were actively engaged in meeting the urgent liquidity needs of the corporate and household sector. This supports the post-crisis recovery under challenging conditions in wholesale funding markets. At the onset of the crisis, banks responded to the emergency liquidity needs of firms by relying on their liquidity buffers. Considering the overall uncertainty and investor risk aversion, tapping markets to obtain the funds necessary to operate such a process could have generated strains in wholesale funding markets or outright rationing episodes. Individual banks have been accumulating on-balance sheet liquidity to be better able to buffer shocks as they have expanded lending. The reduction in money market rates observed after the June TLTRO III, and documented in Box 2, is a further signal that TLTRO funds are pivotal in applying the necessary downward pressure on the cost of the various funding options for banks. Currently, banks are still in the process of accompanying the exit from impairments in supply chains brought about by the pandemic shock, while also sheltering the corporate sectors from pockets of liquidity needs that isolated lockdowns might still entail as the vaccine roll-out normalises the functioning of the economy.

Evidence from survey data

Soft information coming from surveys confirms that most banks participated in TLTRO III operations with the intention of using the funds for lending purposes. Hard data on bank balance sheets available up to March 2021 suggests that TLTRO funds have underpinned the ability of the euro area banking system to answer the initial unprecedented demand for liquidity coming from the non-financial private sector (NFPS) since the onset of the pandemic and to help with the rotation of firms’ exposures from emergency liquidity to term loans. On the one hand, the operations supported euro area banks’ ability and incentives to provide loans. On the other hand, developments in loan demand have played a crucial role. Soft information coming from surveys, such as the euro area bank lending survey, offers insight into the uses of these funds up to now and into the future. Comparing the responses over time also sheds some light on how the banking system reacted to the evolving features of TLTRO III.

The evolution of banks’ replies to the euro area bank lending survey illustrates how banks' intended uses of TLTRO funds adapted to the changing circumstances (Chart 4). Before the tensions associated with the COVID-19 pandemic, banks mainly expected to use TLTRO III funds either to grant loans to the NFPS or to roll over expiring funds from TLTRO II. As the pandemic crisis unfolded, some banks started to consider using TLTRO funds as a substitute for their market funding, although granting loans has remained the main expected destination of the additional liquidity. The TLTRO recalibration on 30 April 2020 has introduced stronger incentives to devote at least part of the funds to various forms of balance sheet expansion including, at least temporarily, holding deposits with the Eurosystem and purchasing government debt securities. At the same time, holding excess liquidity from TLTRO III participation was also motivated by banks’ precautionary attitudes amid unprecedented levels of uncertainty.^[17] As lending to the private sector materialised, remaining TLTRO funds featured an increased role of mixed uses. In any case, banks’ expected uses of funds continue to suggest a high effectiveness of TLTROs in supporting lending conditions.

Estimated impacts of the TLTROs

TLTRO III provides substantial support to bank lending conditions (Chart 5). A large cross-section of studies encompassing a range of econometric methods, sample periods and jurisdictions, reveals a strong easing impact on bank lending conditions. These studies cover the wide spectrum of the transmission channels mentioned above. A holistic approach combining all these studies is therefore likely to average out biases introduced by the absence of specific channels within the same study. The results of each paper are classified in terms of their period of reference and the corresponding TLTRO uptake, rescaling the impacts to account for differences in data, samples and methodologies.^[18] This allows for a recasting of the respective elasticities in terms of the percentage increase of loan volumes per annum and basis points of impact on lending rates for each unit of TLTRO take-up. This results in an impact on loan volumes of above 2 percentage points each year, and on lending rates of over 60 basis points after the last operation in December 2021, against a counterfactual of nil participation. Importantly, this can be considered a conservative assessment of the actual impact of the operations, in particular with regard to their role as a backstop against a sharp deterioration in borrowing costs. This reflects that the sample period of the bulk of the studies does not feature episodes of financial and real economic stress of the magnitude experienced since the start of the COVID-19 crisis. Moreover, the absence of the TLTRO programme could have triggered a sharp deterioration in banks’ funding conditions, leading to a massive deleveraging episode. The effects of such a scenario are very difficult to quantify. Lastly, other pandemic response policies aimed at supporting bank lending, such as broadening collateral eligibility, capital relief measures and government guarantee schemes, are likely to have boosted the effectiveness of TLTROs even further.^[19]

TLTROs have staved off a sizeable part of the uncertainty about bank funding conditions and continue to provide a backstop function, conditional on the availability of spare borrowing allowances. The presence of TLTROs alone ensures that banks have access to ample funding at attractive conditions without stigma from participation into the programme, thereby increasing investor confidence and reducing the likelihood of adverse shocks in bank bond markets. Considering the outstanding amount of bank bonds by yield to maturity suggests that, if sizeable shocks to banks’ funding costs were nevertheless to materialise, this could prompt banks to replace a large fraction of their market funding with TLTROs. This highlights the importance of the TLTRO backstop function. Concretely, currently about 20% of banks’ outstanding bank bonds are priced below the deposit facility rate and more than half these bonds carry yields below the entry rate of TLTRO III. Some banks can thus still find it convenient to resort to market-based financing, if only to maintain their market access and for regulatory compliance purposes. However, a sudden shock to the cost of bond financing could make market-based financing considerably more expensive than central bank funding. At that point, banks could gradually replace the stock of their maturing debt with TLTRO III borrowing to the maximum possible extent given regulatory constraints and the residual spare capacity in terms of borrowing allowance. The current programme therefore serves as a backstop in case of the materialisation of severe distress in banks’ funding markets, but its effectiveness would in any case be constrained by the residual borrowing capacity of banks. Indeed, the availability of spare TLTRO borrowing capacity eases the decision for banks to venture into loan origination even in uncertain times by offering a comfortable level of liquidity and funding buffers to be accessed on demand. Simulations of a micro-structural model of the euro area banking system also show that the availability of attractive TLTROs helped to avoid adverse equilibria for banks’ riskiness, contributing to preserving accommodative funding conditions, above and beyond the actual reduction in the funding cost associated with TLTRO take-up.^[20]

Moreover, the pass-through of TLTROs to lending conditions is complemented by other policy measures. The TLTRO incentive scheme motivates the channelling of borrowed funds to eligible lending. Yet, once the lending benchmarks are achieved, banks are likely to base their investment choices on the relative risk-adjusted return of alternative investment options. If this return is higher for lending to the private sector than for alternative uses of TLTRO funds, then banks have incentives to devote the necessary risk-bearing capacity to loan origination. Conversely, if the risk-adjusted return on loans falls below that of other assets, or if banks face de-facto quantitative capital constraints, they may find it more profitable to divert at least part of the funds into acquiring sovereign bonds or risk-free arbitrage with the ECB’s deposit facility. This has not been the case so far, partly because of complementary support from other policy areas. Asset purchase programmes have contained upward pressures on risk-free and sovereign yields, while microprudential and macroprudential actions have provided capital leeway for an expansion of credit to the private sector and government guarantees have substantially reduced banks’ exposure to the associated credit risk. From a longer-term perspective, the common fiscal instrument of the Next Generation EU programme has supported market expectations around a scenario of substantial fiscal support over the coming years. However, this balance could change for certain segments of the euro area banking system if faced with changes in the constellation of policies in place, such as a premature tightening of regulatory, supervisory or macroprudential policies, or delays in the deployment of the common fiscal instrument.

Soft information and econometric analysis of the impact of TLTROs on lending volumes and terms and conditions confirm that, alongside the stimulus to loan creation, there is no substantial increase in risk-taking by banks benefiting from the programme (Chart 6). While indicating a predominant profitability motive for participation and a recurring intended future use of TLTRO funds for granting loans to the private sector, euro area banks have consistently reported that TLTROs spur an increase in their lending volumes and a decrease in their lending rates, especially for loans to firms. Yet, the impact on credit standards, though it points towards an overall easing of criteria, has been moderating over time and especially relative to the early stages of the pandemic, when TLTRO incentives operated in strong interaction with government guarantee schemes and collateral and capital relief measures. The results of the euro area bank lending survey illustrated in Chart 6 show a negligible contribution to the easing of lending standards. This suggests relatively prudent lending behaviour on the part of participants. The results of this survey also point to a continued tightening of margins on riskier loans since the start of the pandemic, in contrast with broadly stable margins on average loans. Moreover, econometric evidence based on the euro area-wide credit register confirms that banks’ exposure to TLTROs was associated with an increase in lending concentrated towards ex ante safer borrowers, and that the registered increase was not associated with an increase in arrears ex post.^[21]

Together with the stimulus to real economic activity provided by the easing of financing conditions, the impact of TLTROs on employment is likely to have been sizeable. It is generally challenging to analyse the effects of TLTRO III on the real economy, especially at such an early stage after the actual participation in TLTRO III operations. However, previous experience with TLTRO programmes, in situations characterised by a less pivotal role of the operations for bank lending conditions, shows the potential for significant real economic effects of TLTRO III, especially in conjunction with other policy areas activated during the pandemic, such as government guarantees and capital relief measures. A study finds that firms more exposed to TLTROs and capital relief measures tend to increase their employment levels significantly. The impacts predicted for the current juncture based on historical regularities are economically meaningful. Considering the reduction in bank funding cost brought forth by the TLTROs and the capital relief observed over the period between March and April 2020, the overall impact of the policy measures implemented in response to the pandemic has the potential to forestall an employment decline in the corporate sector of more than one million workers over the period 2020-2022.^[22]

4 Conclusions

TLTRO III has been effective in protecting bank-based transmission during the pandemic. Previous TLTROs already supported borrowing conditions for households and firms long before the COVID-19 pandemic hit the euro area. The recalibrations of TLTRO III in the first half of 2020 propelled the programme into a much wider role, as a central bulwark against the impairment of the bank-based transmission mechanism of monetary policy in a context of unprecedented financial stress for the euro area banking system. The largest liquidity injection in the history of the ECB, in June 2020, followed by robust participation in the subsequent operations, has provided central bank liquidity at attractive rates under the condition of complying with demanding and yet achievable lending targets for euro area banks. The stimulus coming from the enhanced operations was transmitted in various ways to lending conditions above and beyond the explicit lending criteria ingrained in the programme. This has helped to secure favourable financing conditions to households and firms throughout the pandemic. Since the stimulus to lending conditions via banks is generally transmitted with a lag, the overall impact of TLTROs has not yet fully materialised.

There is no evidence of substantial side effects or dilution of the stimulus coming from TLTROs so far, with the programme interacting positively with the broader policy package. By supporting lending margins, the programme granted banks enough leeway to extend credit to the private sector without necessarily scaling up the risk spectrum, especially against the background of the unprecedented demand for emergency liquidity needs expressed by corporates in the early stages of the pandemic. There is also evidence that the design of TLTROs ensured that the stimulus reached households and firms as intended, without being excessively diluted by unwarranted uses of funds such as lending to governments. The effectiveness of TLTROs in transmitting the accommodative conditions to the targeted sector was substantially supported by the contribution of other monetary policy instruments such as asset purchases, forward guidance and negative interest rate policy as well as collateral easing measures, which worked in unison with TLTROs, as well as contributions from other policy domains like microprudential and macroprudential policies or fiscal policy via, for example, the provision of government guarantees.

TLTROs offering conditional funding at rates below the deposit facility rate are an effective and flexible tool of monetary policy accommodation in the vicinity of the effective lower bound, especially in an environment characterised by a high level of uncertainty. The targeting element is a key feature of the operations, increasing banks’ incentives for lending. The conditional pricing of TLTROs below the deposit facility rate has created additional room for easing funding conditions for banks in a negative interest rate environment and offers an effective backstop against strains in banks’ access to market-based funding. Moreover, the close link with the deposit facility rate has allowed the negative interest rate policy to continue spurring loan origination, even as deposit rates fell further.