A Reconsideration of the Risk Sensitivity of U.S. Banking Organization Subordinated Debt Spreads: A Sample Selection Approach

1. INTRODUCTION Since the mid-1980s, a growing number of proposals have been set forth that would require large banking organizations to publicly issue subordinated debt on a regular basis. (1) The earliest mandatory subordinated debt proposals were aimed at increasing the size of the financial cushion that could be used by the deposit insurer in the event of a bank failure. Subordinated debt was viewed as a relatively inexpensive substitute for equity capital, because subordinated investors receive their funds only after the deposit insurer is fully compensated and because the tax code permits corporations to deduct interest payments on debt instruments but not dividend payments on equity. Subsequent proposals were aimed at reducing regulatory forbearance. While such proposals vary in their specific details, some would use a bank's ability to issue new subordinated debt as a market signal of its viability, (2) others would use yields on subordinated debt to trigger supervisory actions, (3) and still others would require banking organizations to shrink their assets when they could not issue subordinated debt instruments at a rate below a specified cap. (4) The evolution of mandatory subordinated debt proposals has reflected deposit insurance reforms that were implemented since the mid-1980s as well as empirical information on the risk sensitivity of banking organization subordinated debt spreads. The 1980s began with considerable tic facto too big-to-fail (TBTF) protection provided to subordinated debtholders. (5) Studies using subordinated debt market data from the 1983-84 period (for example, Avery, Belton, and Goldberg [1988] and Gorton and Santomero [1990]) and from the 1985-88 period (for example, Flannery and Sorescu [1996]) were unable to detect a significant correlation between bank-specific accounting risk measures and secondary-market subordinated debt spreads over comparable-maturity Treasury securities. (6) Because subordinated debt investors were not found to receive a premium for default risk, it was reasonable to infer that these investors could not substitute, even partially, for government regulatory oversight of large banking firms. By the mid-1980s, however, the FDIC had put in place mechanisms, such as purchase and assumption transactions, by which it could rescue an insured bank subsidiary without protecting the holding company, or even all of the creditors of the insured bank. (7) As these reforms were implemented, bank-specific risks were found to be increasingly correlated with secondary-market subordinated debt spreads. (8) With this increased risk sensitivity, mandatory subordinated debt proposals shifted toward using market signals to trigger supervisory actions (in some proposals, such actions would even include the revocation of a bank's charter). In 1991, the Federal Deposit Insurance Corporation Improvement Act (FDICIA): 1) required least-cost resolutions of failed depositories and 2) established the system of prompt corrective action (PCA). The PCA system imposes increasingly severe actions on undercapitalized banking organizations as their capital ratios decline, including restrictions on deposit interest rates; elimination of brokered deposits; restrictions on asset growth; restrictions on interaffiliate transactions; and required approvals for acquisitions, branching, and new activities. (9) Under PCA criteria, critically undercapitalized banks, defined as those with tangible equity capital less than 2 percent of total assets, must be placed in receivership within ninety days, unless such actions would not achieve the purposes of PCA, or within one year, unless specific statutory requirements are met. (10) Also under PCA, sixty days after a bank is determined to be critically undercapitalized, the bank cannot make payments on subordinated debt without regulatory approval. Although these reforms could potentially reduce the expected default losses borne by subordinated debt investors, because of more timely and lower cost resolutions, numerous studies have consistently found evidence that post-FDICIA subordinated debt spreads remained closely correlated with various indicators of bank risk, including nonaccruing loans to assets, past due loans to assets, "other real estate owned" to assets, the ratio of (book) equity to assets, the ratio of total (book) liabilities to the market value of common stock plus the book value of preferred stock, cardinalized Standard and Poor's or Moody's bond ratings, supervisory ratings, and portfolio shares for lending and for trading activities. …