Bill Black: Embedded Examiners always married the Natives, but now their Bosses Do Hook Ups

Posted: 28 May 2012 08:59 AM PDT

Bill Black, the author of The Best Way to Rob a Bank is to Own One and an associate professor of economics and law at the University of Missouri-Kansas City. Cross posted from New Economic Perspectives.
Jessica Silver-Greenberg and Ben Protess have written an extraordinarily important column for the New York Times about embedded examiners at JPMorgan.
Embedded examiners’ are federal regulators whose normal work station is a desk at the bank.  We only embed examiners for systemically dangerous institutions (SDIs) – banks so large that they pose a systemic risk to global economy.

Embedded examiners do not work.  They get too close to the bank officers and employees.  In the regulatory ranks we called this “marrying the natives.”  Nothing works with SDIs – they are too big to manage, too big to fail, and too big to regulate.  A conventional bank examination, scaled up to size to fit an SDI the size of JPMorgan would have 500 examiners and take 18 months to “complete.”  (Obviously, when it takes that long to complete an examination it is impossible to “complete” an examination in any meaningful sense – by the time you’ve spent 18 months examining an SDI it can be a radically different bank.)  One cannot conduct an effective conventional bank examination of even a medium-sized bank on a “real time” basis because of the amount of new information pouring in every minute.  Conventional examinations examine a bank’s records and operations “as of” some date (typically the last quarter-end for which reports have been filed).  Embedding examiners is an effort at achieving an “early warning” system.  It has one virtue – it indicates that some senior regulator(s) recognized that they cannot rely on the bank’s own reports to determine whether it is steering toward trouble.
In fairness, twenty-five years ago the proponents of embedding recognize the severity of the “marrying the natives” problem.  They simply viewed embedding as the least bad manner of attempting the impossible – effectively regulating SDIs.  Here is the key passage of the NYT column.

Roughly 40 examiners from the Federal Reserve Bank of New York and 70 staff members from the Office of the Comptroller of the Currency are embedded in the nation’s largest bank. They are typically assigned to the departments undertaking the greatest risks, like the structured products trading desk. Even as the chief investment office swelled in size and made increasingly large bets, regulators did not put any examiners in the unit’s offices in London or New York, according to current and former regulators who spoke only on condition of anonymity.
Senior JPMorgan executives assured the bank’s watchdogs after the financial crisis that the chief investment office, with hundreds of billions in investments, was not taking risks that would be a cause for concern, people briefed on the matter said. Just weeks before the trading losses became public, bank officials also dismissed the worry of a senior New York Fed examiner about the mounting size of the bets, according to current Fed officials.

The authors of the article frame the issue as whether Jamie Dimon’s role as Director of the Federal Reserve Bank of New York poses a conflict of interest and could have led to the regulatory failure to place any examiners in the chief investment office (CIO).  The CIO appears to be the largest de facto hedge fund in the world.  (Note: “hedge fund” is a deliberately misleading term.  Entities called hedge funds typically speculate rather than hedge.  When I call the CIO a “hedge fund” I mean that it largely speculates and disingenuously calls its bets “hedges.”)
I have often expressed my view that Congress answered the policy question about such conflicts of interest with the passage of FIRREA in 1989 – which decided that the analogous conflicts of interest in the structure of the Federal Home Loan Bank System were intolerable and mandated that the governmental functions of examination and supervision be conducted by federal officials.  The regional Federal Reserve banks should be stripped of any involvement in examination and supervision.  That conflict, however, is not my focus in this column.  The point I emphasize is that even at the OCC where that conflict of interest did not exist the “marry the natives” syndrome posed an inherent problem.  Embedded examination did not work even in an era a quarter-century ago when examiners were considerably more willing to say “no” to banks.
I also write to explain why the remainder of the NYT article illustrates how embedded examiners come to see bank propaganda as fact.  I focus on the CIO’s propaganda that it “hedges” and does not gamble.  (The examiners and supervisors also come to believe the SDIs’ proprietary models and to pore over the output of those models rather than the perverse incentive structures that cause the models to err so disastrously and the core, false, assumption that there is some exogenous distribution of financial risk that can be modeled using statistical techniques that require an exogenous distribution.  Example, if we create a criminogenic environment encouraging massive amounts of fraudulent “liar’s” loans then the probability of catastrophic failure becomes 1.0.  Models are not my focus here, though I note that models that assume that bets are “hedges” must produce disaster.)
Let us start with one of the essential attributes of successful bank examination and supervision – professional skepticism.  Our job is to kick the tires.  Our job is to figure out when banks and their officers have perverse incentives, typically arising from compensation.  Our most important task is to detect accounting control fraud – the “weapon of choice” in finance.  We have long known that hedge accounting abuses are a fertile area for fraud.  Fannie and Freddie were the most infamous SDIs to have recently abused hedge accounting – causing the SEC to take action against what it explicitly charged was an effort by Fannie’s senior managers to maximize their bonuses by manipulating supposed hedges.  If JPMorgan’s senior officers were using the CIO to gamble instead of hedge, then they were violation the purpose of the Volcker rule and posing a grave threat to the nation.  Phony hedges designed to hide doubling-down on losing bets are such a common problem that skeptical examiners and supervisors would have made examining the CIO a high priority.  When you’ve married the natives, however, skepticism is the first and primary regulatory casualty.
Here’s how the NYT reporters (inconsistently) describe the CIO.

“Regulators are not typically stationed at divisions like JPMorgan’s chief investment office, which are known as Treasury units. The units hedge risk and invest extra money on hand, and tend to make short-term investments. But JPMorgan’s office, with a portfolio of nearly $400 billion, had become a profit center that made large bets and recorded $5 billion in profit over the three years through 2011.”

It is not clear that the reporters understand that the paragraph contradicts itself.  It states, as if it were an indisputable fact, that the CIO is a “Treasury unit” and such “units hedge risk and invest extra money on hand.”  The next sentence contradicts the first.  It admits that the CIO actually made “large bets” and “recorded $5 billion in profit.”  It gambled on derivatives rather than hedged.  It may have won these bets in the first three years.
Skepticism about the “$5 billion in profit” is essential.  It is easy to abuse investments in financial derivatives in a manner that creates fictional income and hides real losses in the early years.  AIG’s sale of credit default swaps (CDS) protection provide a classic example – book the income now, pay the bonuses now, create no reserves to pay for the massive liability taken on by AIG, and make the officers wealthy while destroying AIG.  By selling CDS protection, AIG was agreeing to guarantee other entities against loss for their investments in “green slime,” e.g., the toxic collateralized debt obligations (CDOs) “backed” largely by endemically fraudulent “liar’s” loans.  It is apparent that the OCC and NY Fed have not examined the CIO sufficiently vigorously to draw any conclusion as to whether the CIO actually made $5 billion in “profit” on its “large bets” in the early years.  The fact that the CIO “recorded” $5 billion in “profit” does suffice to show that they were making bets, not hedges.
Unfortunately, the anonymous regulators quoted by the reporters display even weaker analytics on this point.  JPMorgan has followed an aggressive strategy to keep the regulators on their (round) heels.  When the examiners married Jamie Dimon they married a shrill harpy convinced of his innate superiority over the examiners.  He also has trust issues.  Dimon views examiners who are skeptical and kick the tires as disloyal.  The president of the United States, after Dimon got it very badly wrong, sang his praises.  Obama will stand by his man (donor).  Dimon responds badly to anything less than unreserved praise.  He is a traditional type, he wants the regulator he marries to be a submissive help mate.

“Long before the recent trading blunder, JPMorgan had a pattern of pushing back on regulators, according to more than a dozen current and former regulators interviewed for this article. That resistance increased after Mr. Dimon steered JPMorgan through the financial crisis in better shape than virtually all its rivals.
‘JPMorgan has been screaming bloody murder about not needing regulators hovering, especially in their London office,” said a former examiner embedded at the bank, adding, in reference to Mr. Dimon, “But he was trusted because he had done so well through the turmoil.’”

There are two ways an agency leadership can respond to such a prima donna.  They can be professional but skeptical.  Whenever Dimon “screams bloody murder” they can demonstrate their support for the troops asking the tough questions.  Alternatively, they can send the message that they do not want to upset Dimon.  This will undercut the professional examiners who have resisted marrying the natives.  Over time, the best examiners will tend to leave or wangle transfers to other assignments.  The OCC and New York Fed have historically followed the second management approach.  The reporters cite a specific example involving access to JPMorgan’s capital plan that the examiner believed represented a deliberate effort by JPMorgan management to undercut the examiner.
But here is a vital point – even at their weakest the regulators who marry the natives are better than the natives when it comes to evaluating risk.  The most recent President Bush (in sharp distinction to his father) chose as his regulatory leaders the some of the nation’s leading opponents of regulation.  These regulatory leaders were exceptionally anti-regulatory and pro-industry, but they still were years ahead of most of the industry (and virtually every SDI – including JPMorgan) in warning about liar’s loans, CDOs, and over concentration in commercial real estate.  The NYT authors make the point that the NY Fed examiners want the CIO gamble on a derivative of derivatives unwound “yesterday” while the CIO has continued the gamble.
The tendency of embedded examiners to “marrying the natives” at the SDIs is a serious problem, but the most severe weaknesses in regulation are at the senior levels.  The examiners remain the strongest part of the regulatory chain at the Office of the Comptroller of the Currency (OCC) and the Federal Reserve System.  The reporters provide an excellent example of the this point in their discussion of the OCC’s role at JPMorgan.

“At JPMorgan, when media reports surfaced that the bank was making aggressive bets on credit derivatives, comptroller officials began taking a closer look, people briefed on the matter said. After thumbing through the bank’s own projections for the related risks in early April, the people said, the examiners pushed for more answers but saw no immediate need to change course. The agency notes that it does not bless specific trades.
In a briefing on Capitol Hill last week, two comptroller officials told a room of Congressional staff members that it was ‘common’ and ‘appropriate’ for banks in general to hedge their exposure to various risks, according to people who attended.
‘I know in college they teach you everything is black and white,’ one official said in response to hypothetical questions about creating the perfect hedge.  ‘But it’s not that way in the real world.’”

This brief passage shows why regulators who lack professional skepticism are abject failures.  First, one cannot evaluate adequately a purported hedge by “thumbing through the bank’s own projections for the related risks….”  By the time the OCC was looking, those projections had been shown to have relationship to reality.  Second, of course, the agency does not “bless specific trades.”  No one said it did.  The OCC leaders created a straw man to deflect criticism.  Third, yes it is “common” and “appropriate” to hedge risks, but that is another straw man.  One of the few common elements to the four contradictory major stories that JPMorgan’s press flacks have put out is that their own descriptions of the specifics of the transactions demonstrates that they were bets, not hedges.
Fourth, no, they don’t teach in college that hedging is simple or has no gray areas.  Fifth, the relevant issue has nothing to do with “the perfect hedge.”  A perfect hedge exhibits a negative correlation of -1.0.  JPMorgan engaged in “hedginess” – it made subsequent bets in the same direction as the original bets (positive correlation) – it “doubled down” and lost the gamble.  It delayed informing investors and regulators that it had lost and falsely stated that nothing meaningful had gone wrong.  Dimon then declared his earlier declarations about CIO losses inoperable.
The OCC officials, however, gave Congress the opposite impression that JPMorgan was engaged in “appropriate” “hedging” and should, if anything, be applauded for doing so.  OCC is a bureau within the Treasury Department.  Treasury Secretary Geithner is a virulent opponent of the Volcker rule.  The current draft of the regulation that will eventually implement the Volcker rule was, at the behest of Dimon, crafted by Treasury and the Federal Reserve (another fierce opponent of the Volcker rule) to embrace “hedginess.”  If an SDI claims that a bet on financial derivatives is a hedge (and with portfolios the size of the SDIs one can always claim that “X” is a hedge to “Y”), then Geithner and Bernanke want them to be able to evade the Volcker rule.  This will, of course, destroy the rule.  It was SDIs’ investments in “green slime” financial derivatives that drove much of the ongoing financial crisis and caused eight SDIs to fail.  There is no evidence that the SDIs or their regulators have learned this core lesson.  That is understandable because SDIs inherently have perverse incentives.
As always, I urge that conservatives, libertarians, and progressives join to end the SDIs.  SDIs that are banks receive an explicit federal subsidy through deposit insurance and a far larger implicit subsidy because of the “too big to fail” doctrine.  Congress has never approved this implicit subsidy.  The SDIs are, as they proved during the ongoing crisis, capable of causing global systemic damage.  A nation cannot, therefore, credibly claim that it will not bail out the SDIs’ general creditors.  SDIs are, implicitly, government sponsored enterprises (GSEs) most akin to Fannie and Freddie.
(For those readers who think Fannie and Freddie failed due to government-imposed “affordable housing goals,” please see my articles on their failure.  The short version is that no entity ever required Fannie and Freddie to purchase “liar’s” loans – which did not count towards their affordable housing goals.  Consider why they both, eventually, purchased massive amounts of fraudulent liar’s loans.  The truth is that Fannie and Freddie eventually emulated the (then) investment banks’ massive purchases of liar’s loans and the creation of CDOs for the same reason that the investment banks did – it created guaranteed, massive (albeit fictional) “income” in the near term and made the officers wealthy.  Remember also that Fannie and Freddie did not have any explicit federal guarantee and that their bonds explicitly stated on their face that they were not federally guaranteed.)
The implicit subsidy of FDIC paying the SDIs’ creditors in full even if there is a receivership means that the SDIs can borrow money more cheaply than smaller competitors.  SDIs, therefore, make a mockery of “free markets.”  They are so large that they also make a mockery of democracy.  SDIs are the face of American crony capitalism.
SDIs are not simply dangerous, they are also inefficient.  Shrinking the SDIs to the point where they no longer posed a systemic risk would also increase their efficiency, make them small enough to regulate, and help recover our democracy.
SDIs that function as banks pose intolerable risks to the global economy.  SDIs that function as (thinly disguised) hedge funds should be far beyond the pale.  Conservative and libertarian philosophy rightly condemn providing enormous federal subsidies to a private entity whose senior officers claim any wins and socialize any severe losses.