The Recourse Rule and Its Role in the Last Crisis: An Update

Last year, I wrote a blog post in which I discussed my findings from a working paper that showed how the so-called “Recourse Rule,” a regulation that lowered bank holding company capital requirements for the financial products at the heart of the last crisis, could have contributed to the last crisis. The Journal of Regulatory Economics recently published that working paper. And while the overall conclusions have not changed, in response to the journal’s peer review process I dug deeper into the origins of the rule, so here I’ll summarize the new findings.

New Finding #1: Which Banks Commented on the Regulation

In the initial working paper, I examined how bank size and/or securitization activities may have related to the holdings of the financial products that were affected by the rule. Yet, even with claims that some banks are “Too Big to Fail” banks, bank size alone should not necessarily be a “causal” factor, and banks vary their securitization activities over time. To identify how the regulatory process could have factored into the run up to the last crisis, I looked into the details of which banks might have been involved in influencing the regulation based on information in the Federal Register noticefor the final rulemaking.

First, the Federal Register notice on p. 59629 says that in response to the Regulatory Flexibility Act, the Fed, FDIC and OCC each stated that the rule would only apply to “large” firms. However, the Federal Register notice did not specify which firms. Second, the Federal Register notice on p. 59616 pointed out that the regulators received 34 comments concerning the 1997 notice of proposed rule-making and 32 comments concerning the 2000 notice of proposed rule-making. However, once again the Federal Register notice did not specify which firms commented.

In general, you can track down comment letters on more recent regulations through However, none are available for the Recourse Rule. I therefore, turned to the Electronic Freedom of Information Act facilities. An initial request at the FDIC turned up no comment letters. I next tried the Federal Reserve, and got 58 comment letters (the remaining ones were apparently destroyed). Finally, I turned to the OCC, and got 53 comment letters.

The list of commenting banks includes Bank of America/Fleet Bank, BNY/Mellon, Bank One/First USA/JP Morgan Chase, Capital One, Citibank, Comerica, First Union/Wachovia, MBNA, National City Bank, PNC, State Street, SunTrust, Union Bank California, United States National Bank Association and Wells Fargo. This list includes all of the U.S. Global Systemically Important Banks (U.S. G-SIBs).

New Finding #2: Bank Holding Companies with Commenting Subsidiaries Increased Estimated Holdings of the Assets Affected by the Regulation

Using this list of bank holding company subsidiaries, I found that holding companies with subsidiaries that submitted comments increased their holdings from about 1 percent of their total portfolio to about 7 percent in early 2007. For holding companies with no commenting subsidiaries, their holdings of the highly rated assets remained constant at about 1 percent on average throughout the sample. Since banks typically argue that capital provides a “more expensive” form of funding, this finding suggests that on average only holding companies with commenting subsidiaries took advantage of the rule.

New Finding #3: Bank Holding Companies with Commenting Subsidiaries Reduced Holdings of Lower Rated Assets Affected by the Regulation

Because the Recourse Rule also proposed increasing capital requirements well beyond the traditional 8 percent minimum for similar assets with lower ratings, I also examine how average holdings for these lower rated holdings varied over time. I find that holding companies with and without subsidiaries that submitted comments each on average decreased their holdings. That suggests that all holding companies moved away from the asset classes with even “more expensive” funding requirements than the highly rated assets.


The notice and comment period, summarized effectively in steps 5–7 of this graphic, offers a chance for regulators to hear feedback from the public. Typically, however, the parties most affected by the regulation respond. The affected parties’ interests may sometimes coincide with the broader public interest, but sometimes that may not be the case. Economic research suggests that higher bank capital requirements are associated with a lower likelihood of a crisis. In the case of the Recourse Rule, the findings in the now published version of the paper suggest that on average bank holding companies with commenting subsidiaries contributed to the crisis, first through their comments to influence the regulation, and then through their subsequent risky actions that took advantage of the regulatory changes. The operative word here is “on average”, as bank holding companies did not all take advantage of the rule changes in the same way. That’s important to keep in mind, when it comes to the “Too Big to Fail” issue, since size offers a very rough indicator of bank risk taking, while actual activities offer a precise indication of bank risk taking.