Eric Rasmuse, erasmuse@indiana.edu, Septmeber 12, 2011

Pay for Regulator Performance M. Todd Henderson and Frederick Tung THE LAW SCHOOL THE UNIVERSITY OF CHICAGO September 2011

http://truthonthemarket.com/2011/09/09/incentive-pay-for-bank-regulators/

 This is a good paper, but incomplete. Its big flaw is that it does not distinguish between bank examiners and agency policy.  Even if bank examiners knew exactly how risky banks were in 2007, could they have done anything, given their instructions?  I read the paper hoping for an answer to that question.  It is crucial to the  first part of the paper, which discusses what went wrong.  It is also crucial to the second part, which discusses pay for performance, because if bank examiners just did what they were told, changing their pay structure wouldn't help.

      Setting the right pay plan  based on upside and downside risk is a very interesting problem and does need all the attention you give it.  In the context of bank examiners, tho, you need to think a little about how that interacts with  existing non-pay  incentives. I'm thinking of bribes and future jobs, in particular. Bribes are illegal already,  but--- thinking back to the Becker-Stigler article on crooked cops-- how will they interact with your scheme?  Also, is the revolving door employment a big problem here, and can it be contracted against? (that is, do examiner contracts prohibit employment in banking for 3 years after leaving governmetn service?).

"In the Department of Transportation, for example, only one DOT employee earned a salary exceeding $170,000 at the start of the recession. Eighteen months later, that number had ballooned to 1,690 employees. The number of civilian employees in the Defense Department earning $150,000 or more jumped from 1,868 in December 2007 to 10,100 in the succeeding eighteen months. A plausible impetus for this trend is the belief that greater pay will bring better talent." http://www.usatoday.com/news/washington/2009-12-10-federal-paysalaries_N.htm.

 Rephrase this. The Administration is rewarding its supporters with these pay raises, or perhaps buying off potential opposition--- the Defense Dept. raises are very interesting.  Incidentally, and in the long run, the effect may be to attract more talent, and that may be efficient, but is that the motivation now?

"Exact individual figures are not available. Although one examiner was reported to have received a bonus of about $40,000 on a salary of about $180,000, the large number of recipients means the average bonus per year was likely more on the order of a few thousand dollars." You've got all the necessary numbers; tell the reader the bonus budget per worker, noting that the  average bonus per worker who received a bonus is more relevant.  I'm surprised that government salaries aren't public info, though.  In any case, phone up some officials or submit a FOIA request. If you're refused, that will be a footnote worth including in the paper.

"All of these fact-based judgments go into determining the bank's CAMELS rating, which is the single metric used by regulators to capture safety and soundness."

  How much of the bank inspector's report is released to the public, if any?  Is the CAMELS rating released? Is there any reason the report should NOT be released? Simply releasing reports would act as an important incentive device if the regulator team names are in it, since the quality (and honesty) of their work would become public knowledge.

"For instance, regarding the failure of IndyMac in 2008, the inspector general of the Treasury Department concluded: "examiners did not identify or sufficiently address the core weaknesses that ultimately caused the thrift to fail until it was too late."60"

 It sounds like the Inspector-general did not name names, which is important to getting incentives right. A good incentive reform would be that whenever a bank fails, the regulatory agencies must release the names of the last several examiner teams, the name of the chief examiner's boss, and the promotion history of the chief examiners  *after* examination of the failed bank.

"At the time, banks effectively chose their regulators and paid regulators assessment revenue. From 2003 to 2008, WaMu represented 12 to 15 percent of OTS's total assessment revenue."

 This needs expansion! The implication is that there indeed exist monetary incentives,  extremely big ones, but the incentives are to regulate laxly.  And this is just the tip of the iceberg, perhaps. What about illegal bribes to examiners? We know that it would be worth some banks' bribing an examiner $5,000,000. Does it happen?

And you must make the comparison with rating agencies, which also bought ratings after shopping around with different raters.

"When asked by the inspector general why they did not act in the face of these numerous deficiencies, examiners responded, "even though underwriting and risk management practices were less than satisfactory, WaMu was making money and loans were performing [and] [a]ccordingly, the examiners thought it would have been difficult to lower WaMu's asset quality rating."81...

it contravenes explicit regulatory policy, which instructs examiners against taking comfort in loan and market performance in the face of underwriting or concentration risks. If a bank "has a high exposure to credit risk, it is not sufficient to demonstrate that the loans are profitable or that the association has not experienced significant losses in the near term."82 As in the case of underwriting standards, the inspector general found it puzzling that examiners did not downgrade WaMu's CAMELS rating in the face of these longstanding shortcomings.83"

  Here, a crucial question is what the examiners could do. It seems that they did know that the assets were risky and overvalued in the accounting  (in ways permitted by GAAP).  Did the examiners have the authority and discretion to give the assets a low rating, or did they have to follow a protocol formula that said they had to follow the accounting and not their better judgement or market valuations?  The IG seems to say they had the discretion, but you should go into more detail  in quoting  the internal guidelines.  It's good that you quote them a little. Also, have examiners ever actually followed the quote you give and condemned a bank's assets when those assets were still performing well?  That is, is Washington Mutual exceptional in anything but size?

 This is   crucial to your paper because on the answer to it turns whether the 2007 problem was bad incentives for bank examiners or bad incentives for bank examiners' bosses back in Washington.

"around 2006, federal regulators noticed banks were lowering underwriting standards and amassing large concentrations of commercial and residential mortgage loans.86 Regulators issued very mild warnings to reduce the concentration risk and raise capital to act as a cushion against losses. The response was aggressive:

"Though far from a crackdown, even that mild guidance was too much for banks. Thousands of industry comments poured in objecting to the regulators' intrusion, and the FDIC and other agencies backed off, clarifying that they didn't intend to impose limits." 87"

  That indicates a problem in Washington with the agency, not a problem out in the field with the bank examiners.

"Consider an examiner with a base salary of $100,000 and a variable pay component with three possible outcomes: plus $50,000, minus $50,000, and zero. If she regulates aggressively,..."

 This aggressively  feminist style of writing is conventional in law, now, but it's still bad writing. It jars the reader with a political statement that distracts from the substance.  To be sure, a feminist reader will also be distracted by seeing "he" in a paper, but that reader will also be distracted by seeing "she" since he---or, more likely, she---  will be reading with an eye to political correctness as much as with an eye to the paper's substance and so will be paying a lot of attention to how pronouns are used. The distraction might be a grimace or might be a contented smile, but it will be a distraction either way for the feminist reader.  Thus, you should write for the normal reader.