In response to the global financial crisis, EU governments increased the coverage of their national deposit guarantee schemes in a coordinated manner. The stated aim was to stop withdrawals by depositors and, ultimately, preserve financial stability. Previous studies have shown that the introduction of deposit insurance and broadening of its coverage have the potential to preserve/restore confidence in the banking sector and in this way increase the volume of deposits held at insured banks (Diamond and Dybvig 1983, Chernykh and Cole 2011, Karas et al. 2013, Boyle et al. 2015, Nys et al. 2015). However, increased deposit insurance coverage could be accompanied by increased awareness about the details of the deposit guarantee scheme, in effect ‘waking up’ depositors. A reasonable reaction of depositors to the news about the expansion in coverage would be to take the newly announced threshold into account in their future decision-making, which can be viewed as a manifestation of the ‘availability heuristic’ (Tversky and Kahneman (973). The availability heuristic entails that people attach higher subjective probabilities to information that is more readily available in their mind.
The implication is that the expansion in deposit insurance coverage, although it may in the short run protect individual banks and the banking system as a whole by immediately diminishing the likelihood of a bank run, may also entail an increase in the volatility of individual banks’ deposit funding, because depositors reshuffle their depositors to more optimally benefit from the increased coverage. This may occur in particular when deposit insurance limits apply per customer-bank relation.
Deposit insurance and bank ownership
In a recent study, we employ proprietary micro data of more than 300,000 Belgian customers of a large European bank during 2008 and 2009 to analyse how – in times of crisis – depositors react to an increase in deposit insurance and the concurrent nationalisation and re-privatization of the bank (Atmaca et al. 2020). In November 2008, the coverage of the deposit insurance in Belgium was increased from €20,000 to €100,000 per customer-bank relation. The bank under study was, like several other EU banks, first nationalised and then re-privatised during our observation period. We can therefore compare depositors’ behaviour in response to the increased insurance coverage during state and private ownership. Deposit insurance should in principle only play a decisive role in depositor behaviour of private banks because deposits at state-owned banks are generally assumed to be implicitly, though fully, insured by the state.
Our results suggest that the deposit insurance reform was effective in restoring trust in banks during the crisis. Once the deposit insurance limit is raised to €100,000, depositors slow down their withdrawals. This happens across the board during the period of nationalisation, which seems to calm down all depositors. After the bank’s re-privatisation, however, depositors with a larger increase in deposit insurance coverage increase their deposit growth more than depositors who were already covered before the reform, demonstrating that the ultimate effects of higher deposit insurance only kicked in after bank re-privatisation.
The role of bank relationships and trust in the government
The previous literature has shown that strong bank-customer relationships mitigate deposit withdrawal risk during crises (e.g. Brown et al. 2020). There are several reasons this might be the case during times of bank distress. Transaction costs may be high when switching to another bank, the benefits from private information built up in a close bank-customer relationship may be lost or the range of products offered by the relationship bank may be broader than that offered by other banks. We therefore not only expected that captive depositors would respond less strongly to the crisis by withdrawing less, but also that they would react less strongly and less specifically to increased deposit insurance. We indeed find that the effect of the increased deposit insurance limit on individual depositor behaviour is mitigated by strong bank-depositor relations.
Our results further show that trust in the government is one of the underlying mechanisms explaining the placating effect of increased deposit insurance coverage during the financial crisis. We find accordingly that, during the nationalisation period, the general calming effect of increased deposit insurance on depositor behaviour is more pronounced in cantons with higher trust in the federal government. Once the bank is re-privatised and the new €100,000 limit becomes binding, the stabilising effect of the increase in deposit insurance on treated depositors is more pronounced in high-trust canton. Trust in the government, that is, boosts the effectiveness of federal initiatives such as nationalisation or an increase in the coverage limit of the deposit insurance.
Depositor bunching
To understand to what extent individual depositors really take the (changes in) deposit insurance coverage into account for their deposit choices, we examine the degree to which they sort below the old 20K and the new €100,000 insurance limits to ensure their deposits are fully covered. In the run up to the crisis, we find that more and more depositors limit their deposits to €20,000 for full coverage. Once coverage is increased to €100,000, however, €100,000 bunching largely substitutes this €20,000 bunching. In fact, depositors even start bunching at €100,000 just before the coverage was raised to €100,000, suggesting there were some anticipation effects. This €100,000 bunching behaviour fades away during the period of nationalization, when implicit blanket guarantees apply, but returns in full force once the bank is re-privatized and the new €100,000 coverage limit becomes binding for the bank's depositors. Together, these findings provide clear evidence that depositors are well aware of changes in bank ownership and deposit insurance and react rationally to these changes.
Conclusions
In November 2008 the Belgian government revised its deposit insurance scheme by increasing the coverage from €20,000 to €100,000 per depositor and per bank and also nationalised banks. Our results suggest that these reforms were effective in restoring trust in banks during the crisis. The full effect of the increased deposit insurance only kicked in after a bank was reprivatised. The calming effects of nationalisation and increased deposit insurance are largely conditional on the strength of bank-customer relationships and, more importantly, trust in the federal government. This finding should caution governments against squandering this trust too easily, lest not to dent the effectiveness of two crucial stabilization policies in times of banking crises.
References
Atmaca, S, K Kirschenmann, S Ongena and K Schoors (2020), “Deposit insurance, bank ownership and depositor behavior”, CEPR Discussion Paper 15547.
Boyle, G, R Stover, A Tiwana and O Zhylyevskyy (2015), “The impact of deposit insurance on depos- itor behavior during a crisis: A conjoint analysis approach”, Journal of Financial Intermediation 24: 590–601.
Brown, M, B Guin and S Morkoetter 2020), “Deposit withdrawals from distressed banks: Client relationships matter”, Journal of Financial Stability 46: 100707.
Chernykh, L and R A Cole (2011), “Does deposit insurance improve financial intermediation? Evidence from the Russian experiment”, Journal of Banking & Finance 35: 388–402.
Diamond, D and P Dybvig (1983), “Bank runs, deposit insurance, and liquidity”, The Journal of Political Economy 91: 401–419.
Karas, A, W Pyle and K Schoors (2013), “Deposit insurance, banking crises, and market discipline: Evidence from a natural experiment on deposit flows and rates”, Journal of Money, Credit and Banking 45: 179–200.
Nys, E, A Tarazi and I Trinugroho (2015), “Political connections, bank deposits, and formal deposit insurance”, Journal of Financial Stability 19: 83–104.
Tversky, A and D Kahneman (1973), “Availability: A heuristic for judging frequency and probability”, Cognitive Psychology 5: 207–232.
