Covid-19 briefing: pandemics, natural disasters and banks’ balance sheets

Neeltje van Horen

The Covid-19 (Covid) pandemic is a major shock to the economy but unlike traditional crises or credit crunches, its origin is exogenous to the financial sector. The economy’s ability to recover from the impact of the pandemic will however depend in part on the availability of credit. This raises the question how banks absorb a large shock which originates from outside the financial sector. To answer this question this post reviews the literature on how previous pandemics and natural disasters in the developed world affected banks’ balance sheets. One key message stands out: banks that are more rooted in their market are much more likely to continue lending when faced with the economic fallout from such shock.

Balance sheet adjustments

There is very little literature on the impact of a pandemic on banks’ balance sheets. This is not surprising as pandemics (fortunately) are very rare events. Furthermore, detailed balance sheet data that would be needed to examine this question are not readily available for historic episodes which occurred decades ago.

However, we do know that in 1918, during the height of the Spanish flu, banks active in the state of New York experienced a sharp outflow of deposits (Anderson, Chang and Copeland (2020)). Banks with access to central bank liquidity continued or even expanded their lending. Banks without access did not shore up their finances by borrowing on the interbank market, but curtailed their lending instead. These effects were quite short-lived, however: the authors find that by the end of 1919, banks were able to restore their balance sheets.

Studies on natural disasters are more common and can give some further insights. Banks active in the disaster area after hurricane Katrina hit, increased their capital ratios by reducing risk-weighted assets. Specifically, by buying government bonds and reducing total loan exposure to non-financial firms (Schüwer, Lambert and Noth (2019)). Only stand-alone banks, especially those with already high capital ratios, behaved this way. Those part of a bank holding company did not. These stand-alone banks apparently had a strong incentive to strengthen their buffers against future income shocks.

The Kobe earthquake of 1995 damaged headquarters and branch networks of several banks. These banks reduced lending to firms outside the disaster area, which negatively affected their investment (Hosono et al (2016)). The impact was only short-lived and lasted for about a year. Furthermore, over time the negative trend reversed due to catching up effects. Not only firm lending was negatively affected, but also household credit was tightened during this period. This especially affecting households with limited collateralisable assets, such as housing equity (Sawada and Shimizutani (2008)).

The importance of local lenders

A pandemic or natural disaster can impact lending in several ways. On the supply side it can create operational / business continuity issues which can damage a bank’s capacity to originate loans. For example, by hindering its ability to screen and process loan applications. Similar to the aftermath of the Kobe earthquake, in the wake of hurricane Katrina a significant number of bank branches did not reopen due to infrastructure damages (Brown (2005)). In addition, the damage to borrowing firms’ productive capacity leads to a deterioration of a bank’s portfolio and therefore risk-taking capacity. Finally, heightened uncertainty about borrowers’ health increases the need to actively monitor and screen borrowers and raises doubts about the profitability of future lending.

At the same time, demand for credit tends to go up because households and firms need to rebuild destroyed or damaged physical capital and/or need to deal with a sudden cash flow shortage. Furthermore, banks are often encouraged by regulators to extend loans to borrowers affected by a natural disaster.

This raises the question: who continues to lend? Judging from the literature, being firmly rooted in the local economy seems to matter a lot. In the wake of hurricane Katrina local lenders (ie banks with a large local presence) were originating a higher share of new mortgages and small business loans in affected areas compared to more diversified banks (Chavaz (2016)). This seems to have had a positive effect on the recovery after Katrina: job creation and retention at young and small firms was higher in affected areas where local banks were more dominant (Cortés (2014)). These changes in lending patterns seem to be quite persistent over time. Two years after Katrina, local mortgage lenders reverted back to lending at pre-Katrina levels in New Orleans, while non-local lenders remained largely absent from the market (Gallagher and Hartley (2017)).

Local lenders can be lenders that are only active in one market, but often they are active in multiple markets. Do these financially integrated banks reallocate funds between those markets? Following natural disasters in the US (hurricanes, earthquakes, tornadoes and floods), small multi-market banks indeed reallocated mortgage lending towards markets affected by the disasters (ie markets with increased demand for credit). They specifically reallocated away from unaffected markets in which they had only limited presence, ie their non-core markets (Cortés and Strahan (2017)). In general terms, these banks tend to shield their core markets and cut lending in areas where their ability to generate profits is lower. Larger multi-market banks did not reduce credit in non-affected markets, subduing the aggregate effect on credit supply in these markets

This is not just a phenomenon specific to the US or to mortgage lending. After the flooding of the river Elbe in Germany in 2013, multi-regional banks with a larger share of firms exposed to the flooding increased lending to these firms, ie they engaged in recovery lending (Koetter, Noth and Rehbein (2020)). The authors find that this recovery lending did not induce excessive risk-taking and did not incentivise banks to engage in rent-seeking at the expense of flooded SMEs.

These studies are all in line with earlier findings that local lenders are better able to access information about borrower quality and the value of collateral (eg Berger et al (2005); Degryse and Ongena (2005); Loutskina and Strahan (2009)) and that this advantage is magnified during adverse economic conditions (eg Berger and Bouwman (2017)).(1)

Financing credit demand shocks

How do banks finance these credit demand shocks in the wake of a natural disaster? In the case of hurricane Katrina, local banks circumvented potential capital constraints through the sale of new mortgages into the secondary market (Chavaz (2016)). Banks that were active in multiple markets in addition bid up deposit rates in markets that were not directly affected by the disaster. As a result credit contraction remained subdued in these banks’ unaffected, non-core markets (Cortés and Strahan (2017)).

In Germany, exposed banks financed their lending expansion to flooded firms by activating local savings, not by increasing their wholesale funding (Koetter, Noth and Rehbein (2020)). Importantly, exposed banks with access to geographically more diversified intra-group markets to collect deposits were better able to provide recovery lending. This suggests that the presence of regionally diversified banking groups enhances the ability of local group members to mitigate shocks.


Learning from the past is not so straightforward in the current scenario. Covid is a global pandemic which has proven to be rather difficult to contain and eradicate. The closest historical analogy is the 1918 Spanish flu, but empirical studies focusing on financial stability implications from that pandemic are few and far between. Furthermore, one has to be careful to infer too much from this episode. The financial system was significantly different and the world was recovering from the aftermath of World War One.

However, some lessons can be learned from the empirical literature studying how banks react to a natural disaster. The most important one seems to be that local banks are much more willing to continue lending when disaster hits. These banks seem to be an effective mechanism to mitigate rare disaster shocks faced especially by more opaque borrowers such as SMEs. This could be an important factor to take into account when targeting policy responses to Covid.

(1) None of these studies differentiate between domestic and foreign banks. However, insights from the global financial crisis suggest that when faced with increased uncertainty foreign banks tend to rebalance their lending towards their home country (Giannetti and Laeven (2012)) and towards their core foreign markets (De Haas and Van Horen (2013)), with global banks exploiting their internal capital markets to readjust their loan portfolio (Cetorelli and Goldberg (2012)).

Neeltje van Horen works in the Bank’s Research Hub.

If you want to get in touch, please email us at 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.

2 thoughts on “Covid-19 briefing: pandemics, natural disasters and banks’ balance sheets

  1. Is there a next step to the conclusion that local lenders were more supportive of recovery credit provision? How did the evolution of their NPLs look and NIM net of all charges, after the more time had elapsed for complete recovery compared to the less local institutions?

  2. You say that “local banks are much more willing to continue lending when disaster hits” and that banks provided in the past an “effective mechanism to mitigate rare disaster shocks faced especially by more opaque borrowers such as SMEs”. Thus it is implied that they acted purely out of benevolence, which contradicts “entrepreneurial striving for profit”. The claim that policymakers should consider banks’ selfless support for SMEs when designing policies in response to the Covid can be upheld, only if there is evidence, that any bad debts were written off on a large scale or at least that banks have not benefited from such “careless lending” in any form ex-post. If your claim is true, then I wonder why banks are not showing the same gesture to developing countries, whose suffering is worse and prolonged than experienced by the modern world. It would be wonderful, if there was evidence that banks responded to calamities in other parts of the world, took part in risk sharing and accepted, as the borrowers do, the possibility of losing.

Comments are closed.