The Federal Reserve has eliminated certain hybrid instruments from bank holding companies’ Tier 1 capital with its recent approval of the final regulatory capital rule. However, the exclusion of all hybrid instruments is unnecessary and potentially damaging. Given its performance under stress, cumulative preferred stock should be part of Tier 1 capital for both BHC and bank operating companies.

Beginning in 1996, the Fed allowed BHCs to treat certain hybrid instruments as part of Tier 1 capital and regularly published the list of eligible securities in the Code of Federal Regulations. Tax advantages and lower costs made hybrid instruments a popular means of meeting regulatory requirements for BHCs. On the other hand, BOCs have never been allowed to use hybrid instruments in Tier 1 capital.

In 2010, Section 171 of the Dodd–Frank Act, commonly known as the Collins amendment, required regulators to harmonize BHC regulatory capital with the more stringent BOC capital requirements. The Dodd-Frank Act also required that the Government Accountability Office study the possible effects of the Collins amendment. The resultant GAO study concentrated on the exclusion of trust-preferred securities, which comprised 82% of all hybrid securities in BHC Tier 1 capital as of December 2010. The GAO study argued that troubled institutions would be unwilling to defer dividend payments on cumulative preferred stock, but failed to examine the relative abilities of cumulative and non-cumulative preferred stock to absorb losses in times of distress.

In coordination with the Federal Deposit Insurance Corp. and the Office of the Comptroller of the Currency, the Fed has promulgated its final and comprehensive regulatory capital rule. Among many issues it addresses, the rule imposes strict eligibility requirements for regulatory capital instruments. The rule excludes cumulative perpetual preferred stock, in addition to trust-preferred securities, from Tier 1 capital for BHCs. The rule is supported with the assertion that “the instruments that allow for the accumulation of interest payable are not likely to absorb losses to the degree appropriate for inclusion in Tier 1 capital”. In the absence of empirical verification of this argument, which should have been included in the GAO study, the rule indiscriminately applies the “non-cumulative” criterion for capital instruments eligible for Tier 1 capital in the Basel Committee on Banking Supervision’s 1998 guidance and the dividend/coupon discretion criterion in paragraph 55 of the Basel III framework.

In an earlier article we wrote in Moody’s Analytics, Praveen Varma and I addressed the issue of losses experienced by holders of cumulative and non-cumulative preferred stock when issuers deferred dividends. Importantly, we looked at the ability of issuers to continue operating as going-concerns when their companies did not default on their debt at the time of dividend deferral. Dividend deferrals were not followed by debt defaults in 66 cases and they were followed by debt defaults within a five-year period in 52 cases. The five-year period is crucial because the outcome of the discretionary dividend deferral decision by distressed issuers would be clear within that period.

When issuers deferred dividends on non-cumulative preferred stock, the losses were high. This was reflected in post-deferral valuations of around 15% of face value irrespective of whether a debt default followed the dividend deferral or not.

When dividends were deferred on cumulative preferred stock, the story was quite different. In cases when the dividend deferral on a cumulative preferred stock was followed by a default, post-deferral valuations of around 25% indicated losses comparable to those experienced by holders of non-cumulative preferred stock. For the cases of interest, when the dividend deferral on a cumulative preferred stock was not followed by a default, the losses were much lower: post-deferral valuations were around 38%.

This analysis has important implications for retaining cumulative preferred stock as part of Tier 1 capital. First, preferred stock investors are knowledgeable about the survival and recovery prospects of cumulative preferred stock as shown by valuation of cumulative preferred stock following a dividend deferral. The post-deferral losses of non-cumulative preferred stock are comparable only to losses of cumulative preferred stock deferrals followed by debt defaults.

In cases, where a default did not follow dividend deferrals of cumulative preferred stock, the higher valuations revealed the preferred stock investors’ ability to assess the relative financial health of deferring issuers. These comparative valuations indicate the ability of cumulative preferred stock to absorb losses at times of distress. Second, the retention of cumulative preferred stock as part of Tier 1 capital does not contradict the Fed’s and the Basel Committee’s criterion of “absorb[ing] losses … on a going-concern basis” for inclusion of eligible instruments in Tier 1 capital.

I believe that, while the exclusion of trust-preferred stock from Tier 1 capital is understandable, cumulative preferred stock should be part of Tier 1 capital for both BHCs and BOCs when the regulators revisit the implementation of the regulatory capital rule in the U.S.

Ozgur B. Kan is a director at Berkeley Research Group. He advises financial institutions on credit risk analytics, modeling, regulatory and compliance issues.