How Citi sank itself on the Fed’s watch
By Nicholas Dunbar
The opinions expressed are his own.
Much of the financial crisis can be blamed on bankers who created complex products that allowed them to exploit and monetize less sophisticated investors, borrowers and bank shareholders. However, no account of the financial crisis is complete without an account of the inept regulators who permitted these activities to flourish, causing the crisis to become much worse than it might have been. Among these regulators, most surprising is the story of the New York Fed, supposedly the most sophisticated in its approach to risk. As I recount in this excerpt from my book, The Devil’s Derivatives and as staff at the Federal Reserve Board in Washington DC discovered, the New York Fed was in thrall to what in 2007 was the largest US bank – Citigroup – with disastrous results. -Nicholas Dunbar
The Federal Reserve may have been at the top of the U.S. regulatory pecking order, but within the Fed itself, the New York branch was top dog when it came to regulating banks. This was hardly surprising given the dual importance of Wall Street as the engine room of the bond markets and as the base for the largest multinational U.S. banks. It was only natural that industry risk-management innovations like VAR were first identified by staff in the New York Fed’s markets divi- sion, such as Peter Fisher, who transmitted the ideas to the rest of the regulatory community.
Ever since the regulatory blessing of VAR in the mid-1990s, the New York–based multinational banks had been growing rapidly. By 2003, when William McDonough retired as New York Fed president and was replaced by Timothy Geithner, an ambitious former Treasury and International Monetary Fund bureaucrat, bank supervision was equally important to markets.
If any U.S. commercial bank needed to be challenged, it was Citigroup. In 1999, when then-chief executive Sandy Weill had needed an act of Congress in order to fuse the SEC-regulated Salomon Brothers with Fed- and OCC-regulated blue-chip lender Citibank, he had taken care to reassure his new shareholders and supervisors about the importance of governance. A veteran ex-AIG and Chemical Bank executive, Petros Sabatacakis, was appointed chief risk officer of the new conglomerate and ordered to rein in the freewheeling Salomon traders. Sabatacakis was so tough in applying position limits that on the trading floor he was known as “Dr. No.”
Then came Enron and the dot-com bust. Sabatacakis may have ensured that the bank (unlike Chase Manhattan) avoided significant losses in the shakeout, but Citigroup’s conflicted role in bond underwriting, derivatives, and investment research left it open to the charge of having facilitated massive fraud at Enron and WorldCom. That led to the New York Fed and the OCC censuring the bank in July 2003, as part of a settlement in which it didn’t have to admit wrongdoing. Weill was forced to quit as chief executive (while remaining chairman).
The incoming CEO, former general counsel Chuck Prince, may have seemed like a steady pair of hands on the wheel, but it was Prince who undermined the risk governance mechanism that Weill had put in place. Prince allowed Tom Maheras, the head of fixed income, to appoint his own risk managers. Feeling that his independence had been compromised, Sabatacakis quit in 2004 and was replaced by a ￼Maheras crony, David Bushnell, whose first move was to abolish Sabatacakis’s trading book position limits.
Hidden from public view, this weakening of internal risk governance made it ever more essential that Citi, the largest bank in the United States, was supervised properly. It was essential that the bank’s day-to-day supervisors were not intimidated by the conglomerate and Weill, its charismatic chairman. A former senior New York Fed staffer recalls that the OCC seemed particularly cozy with Citigroup: “I remember being in a meeting in Citigroup, and Sandy Weill stopped by. He gave a big hug and a kiss to this lady examiner from the OCC . . . That didn’t give me the feeling of tough supervision.”
The names of two of the New York Fed’s key bank supervisors— head of bank supervision William Rutledge and head of risk management Brian Peters—appeared on the “written agreement” censuring Citigroup over Enron. Another key figure was Sarah Dahlgren, who from around 2003 onwards was the New York Fed’s chief relationship manager dealing with the conglomerate. Rutledge was widely respected but was also a graying career bureaucrat (he retired at the end of 2010) who defended the status quo of balkanized U.S. bank regulation because he believed it was conducive to financial innovation. In conversations, he was careful and precise about defining what was not his responsibility.
Today, the New York Fed is shy about allowing Rutledge’s key lieutenants, Peters and Dahlgren, to comment about their experiences (Dahlgren succeeded Rutledge after his retirement). Some idea of the challenges that these three and their staff faced during this time can be discerned from public documents and from the recollections of those who examined their work from the outside.
The Federal Reserve System has a governance mechanism intended to reinforce regulatory best practice. The Federal Reserve Board in Washington, D.C., provides centralized resources and sets standards for bank examiners based at the thirteen regional Federal Reserve Banks. In order to apply this governance, the Board has the authority to obtain information about the banks that were supervised within each region.
Setting the Fed’s centralized standards for market and liquidity risk supervision was the responsibility of a small D.C.-based team of former regional reserve bank examiners. Federal Reserve Board staff say that amid the smooth and politically attuned bureaucrats who thrive in large government bodies, the market and liquidity risk team stood out. The team had a reputation for talking about issues in a direct fashion and letting the best idea win, regardless of who supported it. They wanted to apply the same style—and integrity—to managing trading book risk across the Federal Reserve System.
Around 2003, the market and liquidity risk team began trying to collect the trading P&L and VAR data feeds that some of them had seen at regional Fed offices such as Richmond. “We thought that we would pull things together, look for trends, get ahead of systemic risks, and see where crowded trades are,” a member of the team recalls. The most important data would come out of New York: the daily P&L and VAR data for the biggest trading banks, J.P. Morgan and Citigroup.
It was a time when Fed staff recall being beset with distractions: the post–September 11 focus on money-laundering controls, or the vast exercise of preparing for Basel II. For Brian Peters, there was the additional burden of enforcing the post-Enron agreement with Citigroup. For whatever reason, Peters appeared too busy to talk to the market and liquidity risk team members when they approached him asking for daily trading data feeds from the giant New York banks. They returned to Washington, D.C., empty-handed.
Although the sources close to the New York Fed insist that the Board had the right to access whatever information it wanted, the market and liquidity risk team remembers things differently. To them, it was as if the New York Fed was on a different planet from its siblings. “It wasn’t that they wouldn’t provide information because we hadn’t asked for it,” a member of the team says. “They wouldn’t provide information because they weren’t forced to.”
Partial access eventually came from a committee on Large Financial Institutions (LFIs) set up in March 2005 in response to complaints from Fed governors that the Board was not getting enough information about regional Fed banking supervision. As a result, the market and liquidity risk team learned that the New York Fed did not receive electronic daily P&L, VAR, or other relevant trading book information from Citigroup. Instead it received three-month-old reports photocopied from originals provided to the OCC.
Was the New York Fed flying blind?
Even from the untimely trading reports that the New York Fed did receive, the staffers at the Federal Reserve Board became concerned that the New York Fed seemed to lack the expertise—and, just as crucially, the skepticism—to even ask the large banks the right questions. The team obtained information indicating that one major New York Fed–supervised bank had lost between $60 and $80 million trading in the nascent market for carbon emission credits. Up to the moment of the loss, the VAR loss estimate for this trading book had been approximately $1 million, on the basis that the bank’s long position in emission credits had been rising steadily by small increments for the previous year.
Although it was not a substantial or dangerous loss for the bank, this was the type of model methodology weakness that could be indicative of broader problems. Such a weakness could have been picked up by a fulsome trading book or regulatory capital inspection; however, such inspections did not appear to have been undertaken by the New York Fed, despite recommendations from the market and liquidity risk team. Questioned about the need for such inspections, New York Fed bank supervisors complained to their Washington, D.C., counterparts about a lack of resources. One person on the market and liquidity risk team vividly remembers a New York Fed bank examiner shrugging off the emission trading losses, arguing, “Don’t worry about that. We just have to respond to these things when they happen. We can’t get ahead of these problems. We don’t have enough people, and the bankers have a lot of smart people.”
In May 2005, some of the concerns being voiced internally by Federal Reserve Board staff found their way into a confidential operations review of the New York Fed that has since been made public by the Financial Crisis Inquiry Commission. Entitled “Draft Close-Out Report, Large Complex Banking Supervision,” the document highlighted Sarah Dahlgren’s direct management responsibility for the New York Fed examiners at Citigroup. The document then warned that “staffing at the relationship teams is under stress . . . as it pertains to the relationship teams’ ability to conduct continuous supervisory activity. This is of particular concern when evaluating the Citi team.”
Meanwhile, the market and liquidity risk team and others in the Federal Reserve Board supervision division had grown concerned that as large banks built up their trading businesses and accounting rules gravitated to fair value measurement, bank balance sheets were increasingly subject to short-term market moves that could lead to rapid falls in regulatory capital. A memo produced by the team pointed out the issues and risks involved in increased use of fair value and warned that a sudden freeze in certain markets might imperil bank solvency. But when the market and liquidity risk team tried to interest Dahlgren in their findings, she retorted, “I think our banks know how to manage to fair value,” ending the discussion.
In 2006, the market and liquidity risk team attended a Citigroup risk assessment presentation to a committee of Fed examiners. When asked for the rationale supporting the designated satisfactory rating for interest rate risk, the New York Fed team could not provide any information. At another Citi meeting the market and liquidity risk team attended, the New York Fed examiners had been asked to come up with a list of supervisory priorities for the bank. They identified approximately twenty items and patiently explained why each one was important. Near the end, Peters interrupted and told his staff to cut the number of priorities to five or six because twenty was “too many.” The Washington, D.C., team was stunned—twenty was too many things to check regarding the largest and most complex bank in the United States?
In the five years that the market and liquidity risk team struggled to improve the New York Fed’s supervision of Citigroup, the conglomerate added almost a trillion dollars to its balance sheet—visibly and “invisibly.” While sources close to the New York Fed might dispute the words and actions that the market and liquidity risk team attributes to its staff, they are curiously silent about its failings as a supervisor— failings that ultimately would hit U.S. taxpayers.
One of the sad ironies is that even the twenty priorities for Citigroup that Peters cut to the bone did not include the biggest problem of all: the way Citi was building up a $43 billion super-senior CDO exposure on its trading book. Both the New York Fed and its watchdogs in Washington, D.C., failed to spot a fundamental breach of the thin blue line they created: recording the super-senior CDOs as trading exposure and interrogating the bank’s VAR model. A senior Federal Reserve Board official who is still angry about that screwup says, “They didn’t put them in their VAR. And that is a complete violation of all the rules. I mean this is just basic. You do not need to be a quant to catch this. They were supposed to be mark to market. But the attitude seemed to be, ‘Why bother? They don’t change in value. They’re AAA.’ They didn’t put them in the VAR. You can stress-test your heart out. If it’s not in the VAR, you’re not going to get anything on it.”