Learning lessons from government guarantee programmes for bank lending to firms

Government guarantee programmes for bank loans to businesses adopted during the COVID-19 crisis were broad in scope and sometimes quite different in their design and terms. A new CEPR Policy Insight uses a survey of 17 central banks to explore policy lessons by examining the design parameters of the guarantee programmes implemented in their jurisdictions and whether these led to different outcomes across countries. While survey respondents indicated that the interventions were ‘effective’ in enabling bank lending to firms during the pandemic, the findings highlight the importance of focusing on the specifics of programme designs that shape the extent of effectiveness and prevented moral hazard concerns.

Governments across the globe adopted emergency government guarantee programmes for bank loans to businesses during the COVID-19 crisis. Such interventions were broad in scope, sometimes quite different in their design and terms, and often viewed as successful from the vantage point of maintaining credit supply to private borrowers. Policy evaluation studies have been conducted in some countries (Arseneau, Fillat, Mahar, Morgan and Heuvel 2022 for the US; De Mitri, Socio, Nigro and Pastorelli 2021 for Italy; Fatouh, Giansante and Ongena 2021 for the UK; Mateus and Neugebauer 2022 for Portugal, among others). Yet, comparative analysis of these heterogeneous programmes for policy lessons and future programme design – if ever needed – has not been sufficient, given the broad range of potential fiscal and economic ramifications.

Additionally, government guarantees of bank lending have potential impacts on financial stability. An important issue is the problem of moral hazard. Banks may have less incentive to scrutinise borrowers if the loans are guaranteed (Holmström and Tirole 1997); and banks may take excessive credit risks by lending to poor quality borrowers, thereby transferring risks to the public (Boyd and Hakenes 2014). Guarantee schemes may also create moral hazard by encouraging riskier lending at the margins (Kelly, Lustig, and Nieuwerburgh 2016; Gropp, Grendl and Guettler 2014), which can be more severe for riskier banks (Freixas Parigi and Rochet 2004; Bruche and Llobet 2014).

In a recent CEPR Policy Insight (Cao, Goldberg, Sinha, and Ungaro 2024), we explore conceptual issues, some of the heterogeneity across programmes, and some of the policy questions. We start with a comparative analysis of the lessons learnt from the COVID-19 experience to inform critical government guarantee programme design and implementation in the event of unexpected crisis situations. We provide evidence for a survey of 17 central banks related to programmes implemented in their jurisdictions. These loan guarantee programmes varied in size, scale, and structure across countries. Our objective is to draw policy lessons by examining the design parameters of the guarantee programmes and whether these led to different outcomes across countries. While survey respondents indicated that these interventions were ‘effective’ in enabling bank lending to firms during the pandemic, we highlight the importance of focusing on the specifics of programme designs that shape the extent of effectiveness and prevented moral hazard concerns.

Survey and initial evidence

The survey was conducted in 2023 by members of the International Bank Research Network (IBRN) ¹ asking central banks respondents about the primary bank guarantee programmes in their jurisdictions. The survey included an exhaustive list of questions about the design, impact, and country-specific evidence on the effectiveness of the programmes. Seventeen central banks responded to the survey, giving us reasonable sample size to draw inferences. ²

Most central banks participants reported that the guaranteed loan programmes were ‘effective’, i.e. they enabled supply of credit to businesses and aided credit intermediation during uncertain times. Specifically, the interventions enabled credit disbursement to healthy companies that were otherwise constrained for alternative financing.

To ensure that these lending programmes do not increase potential for default or moral hazard among borrowers, many countries adopted several important design elements in their guarantee programme. These included restrictions on borrowers’ quality, loan coverage limits, general lending caps, target sectors, and end-use restrictions on usage of programme loans. While these features differed across countries, most countries report that loan guarantee programmes proved effective in helping small and medium-sized enterprises. Some countries reported that they relied on the banking sector to oversee bank risks and potential for moral hazard.

The significant heterogeneity across countries in implementing the guarantee programmes is visualised in Figure 1. Fourteen (out of 17) countries reported to have introduced restrictions on end-use of guaranteed loans. Three countries in particular – Finland, Sweden and the US – explicitly disallowed using government-guaranteed loans for repayment of existing loans. In countries where such end-use monitoring was not specified, there was potential for partial credit substitution (Altavilla, Pagano, Polo and Vlassopoulos 2022; Boyd and Hakenes 2014).

Figure 1 Restrictions on end-use of funds lent through guarantee programmes

Source: Cao, Goldberg, Sinha, and Ungaro (2024).

Programme designs also adopted different approaches in loan coverage restrictions, for example the maximum percentage of loan capital guaranteed by select governments that participated in the survey. In more than half of the countries surveyed (9 out of 16), governments guaranteed different percentages of loan principal depending on the size of the firm, where higher loan coverage was associated with smaller firms and vice versa. Only in four countries (Germany, Italy, Portugal, and the UK) was the loan coverage 100% of the principal, although limited to special cases for small enterprises and/or small-sized loans. In sum, banks across countries retained a fair degree of ‘skin in the game’ even while extending guaranteed loans, as they either maintained regular due diligence criteria while screenings borrowers or extended smaller-sized loans (Nicolas, Ungaro and Vansteenberghe 2022).

Takeaways and a call for future work

Our survey of central banks also had a focus on whether there was a moral hazard issue faced by banks in extending guaranteed loans. In general, the loan guarantees were viewed as associated with lower levels of risk-taking by banks. In many countries, the loan guarantees were extended with restrictions on borrowers’ quality, loan coverage limits, general lending caps, target sectors, and end-usage of loans – with a view to prevent excessive risk exposure for the banking sector, by ensuring either partial loan coverage or by extending smaller-sized loans. Our survey-based results come with caveats. First, there was higher participation of advanced economies in our survey, and it also is important to study programmes implemented in emerging market and developing economies in greater detail. Second, our survey focused only on guarantee programs launched during COVID-19, which presented a specific set of challenges for economies.

More research could be conducted on optimal and comparative programme designs to see how specific features are optimally calibrated and spell out the near-term and longer-term economic, fiscal, and financial stability features associated with different combinations. More research is necessary to fully understand implications for banks, borrowers, and fiscal cost implications for future. Moreover, it would be valuable to study other emergency contexts wherein extending guarantees may be necessary, with a comprehensive study of metadata related to guaranteed loan programmes. Conducting such research during stable times could help preparedness for future crises.

Source : VOXeu