Taking Firms and Markets Seriously A Study on

Taking Firms and Markets Seriously: A Study on Bank Behavior, Market Discipline, and Regulatory Policy Thomas Bernauer and Vally Koubi

Objectives § Accounting systematically for the behavior of firms and markets is important for understanding regulatory policy, particularly in the economic realm. We examine the US banking sector and its regulation in the 1990 s to gain insights on how studies along these lines could be constructed. § Examine whether market discipline could fully substitute for governmental regulation of banks’ minimum capital requirements => empirical puzzle of overcompliance § Examine complementarities among markets and regulation Datum Autor/Institution/E-Mail

Average Capital-Asset-Ratios of Banking Sectors in Selected G-10 Countries Average Capital-Asset Ratios in the U. S. Banking Sector Datum Autor/Institution/E-Mail

Bank Capital § It is a direct source of funds for loans. § It affects the perceived riskiness of bank deposits. § Banks with higher capital-asset ratios are perceived as safer managers of borrowed funds and are able to attract deposits at more favorable terms (lower rates) than low capital-asset ratio banks. § A bank then can use its capital ratio as a means of establishing a cost advantage against its competitors (substitution effect). § A highly capitalized banking industry makes bank deposits relatively safer in comparison to other investments (from the investordepositor’s point of view) benefiting all banks (the general effect: an externality) Datum Autor/Institution/E-Mail

Theoretical Model § A bank may manipulate its capital-asset ratio (CAR) to influence its borrowing costs. This behavior has implications for bank capitalization and hence for the necessity - and form - of bank regulation. § The relationship between borrowing costs and CAR is R = R(k, k*) R = interest rate on borrowing funds K = own bank capital k* = average capital of entire banking system Datum Autor/Institution/E-Mail

§ If bank creditors (depositors) care about individual bank risk (bankruptcy) then § If bank creditors care about bank systemic risk then § Depositors may care about both idiosyncratic and systemic bank risk. Which one do they care about more? The answer depends on the sign of This sign determines whether the banking system is undercapitalized or not. Datum Autor/Institution/E-Mail

CASE A Free riding An unregulated banking sector is undercapitalized. CASE B No free riding Market forces work against undercapitalization We show that, in the absence of deposit insurance, banks are not undercapitalized. Datum Autor/Institution/E-Mail

Empirical Analysis Does competition make banks care about CAR? Rit = w 1*kit + w 3*xit + uit H: w 1 < 0 Data: yearly observations on the entire population of US banks in the period 1990 -2000 (approximately 130, 000 bank-years). Method: pooled cross-section times series regressions with a fixed effects procedure. Rit = average cost of deposits = ratio of deposit interest expense to value of total deposits kit = tier 1 capital-asset ratio (tier 1 capital divided by total risk weighted assets) xit = bank size and idiosyncratic characteristics (number of employees, total value of assets, return on equity, non-performing loans) Datum Autor/Institution/E-Mail

§ Well-capitalized banks face lower average interest expenses on their deposits (ki). § While additional capital lowers borrowing costs, it does so at a decreasing rate (ki 2). Datum Autor/Institution/E-Mail

Datum Autor/Institution/E-Mail

§ Well-capitalized banks face lower average interest expenses on their deposits (ki) at a decreasing rate (ki 2). Competition among banks should contribute to higher levels of bank capitalization. § Large banks – either in terms of number of employees (EMPL), or value of assets (ASSET) – face lower borrowing costs, and so do banks that are more successful in terms of higher returns on equity (ROE). § Banks with a larger share of non-performing loans (NONP) face higher interest expense ratios. Datum Autor/Institution/E-Mail

If banks care about CAR, do they care enough to make them eliminate the free riding problem? Rit = w 1 (kit –k*) + w 2 kt* + w 3 xit + uit w 1 < 0 (competition effect) w 2 < 0 free riding w 2 > 0 no free riding kt* is the average capital-asset ratio at time t in the US state where the bank is located. Datum Autor/Institution/E-Mail

Datum Autor/Institution/E-Mail

§ Both w 1 and w 2 (the coefficients for the variables ki - k* and k*) are statistically significant and negative) § if a bank increases its capital-asset ratio relative to the U. S. state average, it reduces its borrowing costs; § an increase in the industry-wide capital-asset ratio also reduces the bank’s borrowing costs. § Yet, given that the coefficient for the average capital-asset ratio, k*, is much larger than that for the individual capitalasset ratio (ki –k*), the latter effect dominates. § Competition among banks is not sufficient to eliminate the free-rider problem. Datum Autor/Institution/E-Mail

Conclusions § In the US, better capitalized banks indeed face lower borrowing costs. § Bank competition could not have perfectly substituted for capital adequacy regulation because of substantial systemic effects (free riding). § The main value of some bank regulations, such as parts of the Basle Accord, may be found in their ability to strengthen bank competition by generating simple benchmarks for comparing the riskiness of banks. § Policy Implications: an integral - if not the key - part of domestic and international regulatory efforts should involve transparency-creating regulation. International standards of this type can encourage economic actors in various countries to signal superior quality to their respective markets by complying or even over-complying with environmental, corporate governance, accounting, or other standards. Datum Autor/Institution/E-Mail