Deposit insurance and the stock market: evidence from Denmark

Abstract

Previous studies of the relationship between deposit insurance and bank market values have usually been limited to consideration of minor changes in bank regulations, but the 1987 initiation of deposit insurance in Denmark permits examination of a potentially major policy shift. It is found that the market values of large Danish banks exhibited a modest positive reaction to the announcement of insurance, but that small risky banks responded negatively. These results partially contrast with those previously found for the USA, an outcome that seems likely to reflect the interaction of deposit insurance with the particular characteristics of the pre-existing Danish regulatory system.

Keywords:

• deposit insurance
• bank regulation
• Denmark

Acknowledgements

We are particularly grateful to Carsten Tanggård for providing us with the data used in this study. We also appreciate the helpful comments we received from Louis Ederington, David Mayes, Vivien Pullar, and two anonymous EJF referees.

Notes

For example, Duan et al. (1992) are restricted to looking at the 1981 change in capital requirements in their assessment of the role of deposit insurance in bank investment decisions.

Stover (1997) argues that the Early Resolution Program followed by the US Resolution Trust Corporation in the early 1990s accomplished a similar result, but only for a very limited number of institutions. As in Denmark, the RTC attempted to intervene while the failing institution still had positive net worth.

For examples of the US evidence, see O'Hara and Shaw (1990) and Cornett and Tehranian (1990).

Studies that document these incentives include Duan et al. (1992), Billett et al. (1998), Pyle (1986), Ronn and Verma (1986), Kane (1989), and Pennachi (1987).

By contrast, financial institutions in the USA were principally required to produce historical-cost book value statements. Bernard et al. (1995) detail how the two systems compare in the context of their overall regulatory structures.

Equity capital only is employed in the numerator of these ratios. They should not be confused with the total capital/assets ratio used in the Danish regulatory insolvency determination. Banks are permitted to use other liabilities in achieving the minimum regulatory capital requirements.

Indeed, discussions with market participants suggest that such fears were on the rise at this time due to increased licensing of risky banks during the 1980s.

For other examples of this approach in similar contexts, see Binder (1985), Cornett and Tehranian (1990), and Wagster (1996).

These results are for one-day event dates. Two-day returns yield similar findings.

A simple χ² test confirms that the large-bank reaction was significantly greater than those of medium and small banks.

As the first debate of the legislation on 3 November 1987 indicated a clear majority in favour, it seems possible that the market viewed approval as a foregone conclusion after this date. We therefore re-estimate equations (1) and (2) excluding the last four events in Table 1, and obtain similar qualitative results. However, the difference between large and small banks becomes significant at the 5% level. Similarly, the distinction between high-risk banks and medium/low-risk banks in Table 4 becomes statistically more robust, but is otherwise unaffected.

Again, a χ² test indicates that the reaction among high-risk banks on each of these dates is significantly more negative than those of medium-risk and low-risk banks.

Indeed, the original scheme explicitly prohibited the deposit insurance fund from rescuing troubled banks. A subsequent law change allowed regulators to undertake rescue in circumstances where payment of insurance to depositors would be more expensive.