From the enactment of the first commercial banking codes in the nineteenth century through the adoption of the Basel and Basel II accords in recent years to the anticipated adoption of Basel III, policymakers have argued that holding increased amounts of capital promotes bank “soundness and stability” (Basel Committee on Banking Supervision 1988, 2004). Even before governments began to mandate explicit minimum capital requirements, law, custom, and market forces led to alternative means for providing sufficient levels of capital, primarily though extended shareholder liability (Berger et al. 1995).
The oldest and most well-known system of extended liability is unlimited liability, under which partners bear liability for all the obligations of a failed firm. In England and Sweden, banks that issued currency during the nineteenth century were typically subject to unlimited liability. In the US, state law often mandated that banks chartered under their authority be subject to “double” liability. In case of the bank’s failure, shareholders would be liable for twice the amount they had originally paid for their shares; some states mandated “triple” liability. The theoretical argument underlying these arrangements is simple. With more “skin in the game,” shareholders will be less willing to undertake excessive risk.
A formal analysis of liability: Uncalled capital
Although the recent crisis has encouraged economists to consider the policy implications of adopting various forms of extended liability (see Kashyap et al. 2008 and Flannery 2009 on “contingent capital”), there have been relatively few systematic attempts to examine this historical experience with extended liability. Exceptions include Esty (1998) and Grossman (2001) who demonstrate that such strict liability rules played an important role in reducing moral hazard problem in the US banking sector during the half-century or so prior to the Great Depression.
In recent research (Grossman and Imai 2010), we consider another mechanism for imposing contingent liability upon shareholders which was common in Britain during the late nineteenth and early twentieth century: uncalled capital. Under this system, firms issued equity with a nominal value, all or part of which might have been required to be paid in by subscribers at the time of the initial offering. Shares which were only partly paid carried with them a contingent liability for the unpaid portion of the share and could be called in by the firm at the management’s discretion.
We investigate the consequences of uncalled liability by analysing bank balance sheet and share price data from British banks during 1878-1912. At the time of establishment, company promoters declared the nominal amount of the firm’s capital, the number of shares into which it would be divided, and the portion of each share that would be paid-in by subscribers. There were no statutory requirements for the amount of uncalled capital, which was determined by a variety of factors, including common practice within an industry and current opinion as to what share characteristics were conducive to promoting financial stability (Jefferys 1946).
Figure 1 demonstrates the extent of uncalled capital in a variety of sectors during 1870-1913. The figure reports data for only a few sectors, including banks, insurance companies, trusts, and land, mortgage, and financial companies. These sectors maintained a high amount of uncalled capital relative to the market as a whole. The high proportion of uncalled capital can be seen as a market-imposed requirement to engender confidence in sectors where leverage was high and the physical assets were either meagre or inaccessible to creditors.
Figure 1. Ratio of uncalled to market capital, selected
We measure bank risk in two ways:
- following Saunders et al. (1990) and Esty (1998), we use the volatility of share prices;
- the ratio of loans to total assets, under the assumption that loans are riskier and less liquid than other balance sheet assets.
We regress these measures against bank capital-to-asset ratios, a dummy variable for limited liability, a measure of the amount of uncalled liability, plus interaction terms between capital and unlimited liability and capital and contingent liability. The interaction terms are included to capture the possibility that extended liability is less relevant when banks are more adequately capitalised.
The results based on share price volatility are generally not significantly different from zero. Those based on the loan-to-asset ratio, however, are less ambiguous and more robust. The coefficient on contingent liability is negative and significant, meaning that higher levels of contingent liability are associated with less risk-taking. These effects are economically important. Comparing a bank with no contingent liability with a bank with contingent capital equal to the total amount of paid-in capital (i.e., the amount put up by shareholders in at the IPO), the latter’s loan-to-asset ratio is, on average, 16 percentage points less than that of the former. Further, the effects of contingent capital are stronger for banks operating with lower capital-to-asset ratios than with well-capitalised banks, suggesting that contingent capital may act as a brake on risk-taking for banks that are, in fact, in greatest danger of failure.
Do stricter liability rules encourage prudence?
Our results suggest that, at least prior to World War I, British banks which operated under more strict liability rules – particularly more highly leveraged banks – undertook less risk than those operating with lower levels of contingent liability. These results are consistent with both the predictions of economic theory as well as the findings of empirical literature that focuses on the consequences of double liability in the US (Esty 1998, Grossman 2001, Grossman 2010).
Of course, as shown in Grossman (2001), state-chartered banks in US states that imposed extended liability during the Great Depression fared worse than banks in limited liability states, suggesting that the benefits of extended liability may be of limited use in the face of a financial tsunami. Nonetheless, our results have an important implication for today’s policymakers. Extending bank shareholders’ liability can protect taxpayers by directly reducing the taxpayer’s share of bank resolution costs and, more importantly, by altering the risk-shifting incentives of banks.
Basel Committee on Banking Supervision (1988), "International convergence of capital measurement and capital standards”.
Basel Committee on Banking Supervision (2004), "Basel II: International Convergence of Capital Measurement and Capital Standards: a Revised Framework".
Berger, Allen N, Richard J Herring, and Giorgio Szegő (1995), "The Role of Capital in Financial Institutions", Journal of Banking and Finance 19(3-4):393-430.
Esty, Benjamin C (1998), "The Impact of Contingent Liability on Commercial Bank Risk Taking", Journal of Financial Economics 47(2):189-218.
Flannery, Mark J (2009), "Stabilizing Large Financial Institutions with Contingent Capital Certificates".
Grossman, Richard S (2001). "Double Liability and Bank Risk Taking", Journal of Money, Credit and Banking, 33(2):143-59.
Grossman, Richard S (2010), Unsettled Account: The Evolution of Banking in the Industrialized World since 1800, Princeton University Press.
Grossman, Richard S and Masami Imai (2010), “Contingent Capital and Bank Risk-Taking: Evidence from British Equity Markets before World War I”, Paper prepared for the Economic History Association Meetings, Evanston, IL, September 24-26.
Jefferys, JB (1946), "The Denomination and Character of Shares, 1855-1885", Economic History Review, 16(1):45-55.
Kashyap, Anil, Raghuram Rajan, and Jeremy Stein (2008), "Rethinking Capital Regulation". Federal Reserve Bank of Kansas City symposium on “Maintaining Stability in a Changing Financial System”, Jackson Hole, WY (August 21-23, 2008).
Saunders, Anthony, Elizabeth Strock, and Nickolaos G Travlos (1990), "Ownership Structure, Deregulation, and Bank Risk Taking", Journal of Finance, 45(2):643-54.