The risk of national and global contagion during the Global Crisis forced governments to bail out financial institutions, and this had the potential to severely disrupt the global financial system. To the extent that such interventions can be anticipated, they may induce moral hazard. Implicit guarantees of government support have, in fact, been shown to amplify risk-taking, undermine market discipline, create competitive distortions, and generally increase the likelihood of future distress (Dam and Koetter 2012).
In light of this, in November 2011 – as part of a modified supervisory framework on institutions deemed ‘too big to fail’ – the Financial Stability Board (FSB) explicitly identified 29 financial institutions as ‘global systemically important banks’ (or GSIBs). To reduce moral hazard and promote a stronger financial market infrastructure, the framework subjected these institutions to higher capital requirements, closer supervision, and more effective resolution mechanisms. Yet, a global systematically important bank designation also certifies banks as being too big to fail (which may lead to more moral hazard). Such a designation also increases regulatory costs, for which banks might try to compensate with additional risk-taking. The ultimate effect of a global systemically important bank designation is therefore an open empirical question.
This is particularly true given that the known effects of higher capital requirements – even for large banks – do not necessarily extrapolate to institutions identified as global systemically important banks. First, this is because these institutions tend to have more market power, political influence, and more opportunities for cross-jurisdictional arbitrage. And second, this is because GSIB designation implies a competitive disadvantage vis-à-vis similar non-designated competitors (that do not face the additional requirements). In the specific case of the FSB’s framework, capital requirements are further complemented by additional supervisory scrutiny. This may or may not be successful in reigning in risk-taking incentives.
One particularly important dimension along which designated banks may respond is in their corporate lending. The main question that arises in this context is whether (and how) banks adjust their credit supply, and if these adjustments have consequences for the real economy.
The FSB recently described its own lessons from the evaluation of the too-big-to-fail reform (Buch 2020) and – like in existing research – generally finds economically small effects. Corporate borrowers from the US, for example, experience reduced credit supply from designated institutions, but appear to offset the effect with funding from less-affected banks (Favara et al. 2020). Other existing research has studied the longer-term impact of banks’ global systemically important designation (Behn and Schramm 2020), as well as the isolated effect of capital surcharges (Favara et al. 2020), or has relied on bank-level data (Violon et al. 2017, Behn et al. 2019, Goel et al. 2019).
In our recent work (Degryse et al. 2020), we instead analysed the short-term effects of the initial designation announcement on credit supply to the global syndicated loan market. The initial global systemically important banks list we rely on did not detail actual capital surcharges and this improves comparability of the effects across all affected banks. Focusing on the first announcement – and thus on an earlier date than existing work – allows us to capture anticipation effects.
Our analysis uses a difference-in-differences approach with a one-year window around the first designation on 4 November 2011 (see Figure 1) to investigate three research questions: How do designated banks adjust their loan terms around their designation? Do designated banks change the composition of their corporate lending? And do the adjustments in designated banks’ lending behaviour affect firm outcomes?
Figure 1 The first GSIB designation, relevant events, and the sample period
Note: The figure shows the timeline of the first GSIB designation and relevant events. It also illustrates the pre- and post-treatment periods used in our analysis.
To answer these questions, we use data on syndicated loans for 81 international bank holding companies that were all (at some point) considered for global systemically important bank designation by the Bank for International Settlements. As shown in Figure 2, we study bank lending at the bank level (panel a), bank-country-industry level (panel b), and bank-firm level (panel c). For identification, we not only rely on the fact that the FSB was deliberately cautious about its intentions to publish an official list of designations, but also exploit the leaks of only partially correct global systemically important bank lists by the Financial Times.
Figure 2 Bank lending in the syndicated loan market over time
Notes: The figure shows the evolution of the mean of the logarithm of outstanding loans calculated at the bank level (Panel A), the bank-country-industry level (Panel B), and the bank-firm level (Panel C). In addition, we distinguish between Old GSIBs (solid blue line), New GSIBs (solid red line), FT Non-GSIBs (solid green line), and Pure Non-GSIBs (solid grey line). Two dashed vertical lines mark the year right before (November 2010 – November 2011) and after (November 2011 – November 2012) the FSB’s first GSIB designation.
Our main findings can be summarised as follows:
- At the bank level, designated institutions (our treatment group) cut their lending in the syndicated loan market by 9.1% compared to non-designated institutions (the control group) in the year following their designation. This result should be treated cautiously, however, since we do not control for demand factors that might drive the differential responses by each set of banks.
- At the bank-country-industry level and at the bank-firm level, we find that designated institutions reduce their credit supply primarily at the intensive margin (e.g. by 6% to the same firm compared to non-designated institutions). Importantly, this lending cut seems to be driven by the additional supervisory scrutiny and not by the expectation of capital surcharges. Designated insitutions that can expect moderate surcharges cut lending to a specific firm by 6.5% compared to non-designated institutions. We find no significant effect for designated institutions that can expect higher surcharges.
- Although the FSB list in November 2011 was the first official designation, a potential concern could be that the Financial Times leaked an early list in November 2009. We find that designated institutions that were not on the leaked list reduce lending at the intensive margin to a specific country-industry or firm by 10% to 11%, whereas banks that were correctly predicted reduce lending by approximately 3% to 5%.
- Designated institutions reduce their lending to high-risk borrowers but not for low-risk ones. Specifically, designated institutions that were not on the Financial Times list cut their credit supply at the intensive margin to all borrowers and are more likely to stop (and less likely to start) lending to risky borrowers. In contrast, designated institutions that were correctly predicted by the Financial Times reduce their lending to high-risk borrowers at the intensive margin (by 5.7%), but not to less risky firms. At the extensive margin, they do not appear to adjust their lending.
- In a final step, we link the lending adjustment of designated institutions to outcomes at the firm-level, in order to investigate whether the introduction of the new global systemically important bank framework has any effect on the real economy. Using firm-level data – and by comparing firms that are dependent on designated institutions’ credit supply (the treatment group) with firms that are not dependent on GSIBs’ credit supply (the control group) – we show that designated institution-dependent risky borrowers experience lower asset growth (by 2.2%) and lower investment growth (by 5.4%) compared to similarly risky firms that are not dependent on such banks.
We exploit the first global systemically important bank designation on 4 November 2011, as well as the publication of a preliminary designation list by the Financial Times, to examine how designated institutions adjust their lending behaviour and whether this adjustment has any effect on the real economy. Overall, we find that designation causes an economically relevant decrease in corporate lending in the syndicated loan market at the intensive margin, and that it induces designated institutions to stop lending to some borrowers at the extensive margin. The lending cut seems to occur across industries but is concentrated among risky corporate borrowers. This implies a lower risk profile under the FSB framework, resulting primarily from stricter supervision. Our findings therefore suggest a decrease of designated institutions’ risk-taking in the corporate loan market, which is in line with the intended effects of the policy – namely, to reduce ex ante moral hazard among systemically important banks.
The success of the policy, with respect to stabilising the financial system, comes at the cost of lower asset growth, investment growth, and sales growth among those riskier firms that experience reduced credit supply from designated institutions and are seemingly unable to substitute their borrowing with funding from other sources.
Behn, M and A Schramm (2020), “The impact of G-SIB identification on bank lending: evidence from syndicated loans”, ECB working paper.
Behn, M, G Mangiante, L Parisi and M Wedow (2019), “Does the G-SIB framework incentivize window-dressing behaviour? Evidence of GSIBs and reporting banks”, European Central Bank, working paper.
Buch, C (2020), “Evaluation of too-big-to-fail reforms: Lessons for the COVID-19 pandemic”, VoxEU.org, 25 September.
Dam, L and M Koetter (2012), “Bank bailouts and moral hazard: Evidence from Germany”, The Review of Financial Studies 25 (8): 2343-2380.
Degryse, H, M Mariathasan and T H Tang (2020), “GSIB status and corporate lending: An international analysis”, CEPR Discussion paper 15564.
Favara, G, I Ivanov and M Rezende (2020), “GSIB Surcharges and Bank Lending: Evidence from U.S. Corporate Loan Data”, Journal of Financial Economics, forthcoming.
Goel, T, U Lewrick and A Mathur (2019), “Playing it safe: Global systemically important banks after the crisis”, BIS Quarterly Review.
Violon, A, D Duranty and O Toaderz (2017), “The impact of the identification of GSIBs on their business model”, Banque de France, working paper.