Saturday, January 19, 2013

J.P. Morgan "London Whale" Report: A Whale of a Tale

The 129 page internal report by J.P. Morgan on "2012 Chief Investment Office Losses" is alternately a dry and a comic unintentional indictment of the company, its management, and the academic concept of risk management for a modern investment bank.

The most widely read post on this blog is from September 2011, on the subject of AIG, financial risk management, and rogue traders.  Reading the current report absolutely convinces me about the merits of the arguments I made in the 2011 post. Further, it is also clear that the TBTF ("Too Big To Fail") argument is off the point.

Institutions like J.P. Morgan are TCTM ("Too Complex to Manage") The talk about the "London Whale" does nothing more than anthropomorphize the huge, systemic risk posed by institutions like J.P. Morgan Chase.  The 2012 CIO losses cannot legitimately be attributed solely to the behavior of one trader, or even to a group of traders.

To have outsiders look at the Synthetic Credit Portfolio ("SCP") would have been prohibitively expensive and time-consuming.  It would have also shed no more light than does the current report.  The internal report is poorly written, alternatively simplistic and opaque, and completely exonerates the cast of incompetents, charlatans and cynics that populated the CIO.  Any reader who understands Wall Street will understand where the issues lie after reading the report, although the reader may not understand the fine details of the SCP's portfolio implosion.

The whole premise of the SCP is unconvincing.  It was supposed to be a vehicle for deploying excess deposits in the system to inoculate J.P. Morgan's core credit portfolio from unanticipated changes in interest rates, inflation rates, and economic growth. SCP's overall strategy, trading tactics, and management truly describe a proprietary trading vehicle, a distinct profit center.  It was certainly not managed as a hedging vehicle.

Here's the arc of the story in numbers and events:

  • At month-end January 2011, cumulative mark-to-market (M2M) losses on the SCP portfolio were $100 million;
  • At month-end February, M2M losses are $169 million;
  • At month-end March 2011, M2M losses are calculated at $718 million.  
  • On April 5, 2011, CEO Jamie Dimon, the CFO and CRO are told the SCP's risk was "balanced," the market for credit index swaps"dislocated" and the M2M losses "temporary."  One trader, without any defensible basis, says that the losses would "mean revert."
  • On the basis of these dubious assurances, CEO Dimon utters the infamous quote on a 4/13/2011 conference call describing investor concerns about the SCP portfolio as "a tempest in a teapot." CFO Braustein says that he is "very comfortable" with the portfolio's position and risk profile. 
  • All during the month of April, losses in the SCP continue to mount.  Non-CIO personnel, primarily from the Investment Bank, take over management of the portfolio.  Taking a more rigorous and defensible look at the portfolio risk, they analyze correlations between the instruments and estimate risks across a range of different scenarios.  Their basic conclusion: SCP's risk was much higher than than was reported to senior management ahead of the April quarterly conference call.  Markets had become illiquid because other savvy investors like Boaz Weinstein had come to understand months before, the extent of SCP's predicament. So, the comments given to management about credit index swap markets being dislocated was disingenuous.  A noose had been slipped on JPM and tightening for months. 
  • By May 2011, M2M losses were more than $2 billion!  
  • Around this time, CEO Dimon says that the SCP investment strategy was "flawed, complex, poorly reviewed, poorly executed, and poorly monitored." 
  • JPM's external auditors, PwC had reviewed the portfolio marks in conjunction with the report of the first quarter's financials and found no issue with them.  Their subsequent May 2011 investigation found that the portfolio markets lacked "integrity."   Whereas PwC had expressed confidence in the process and outputs of the M2M, they now opined that the marks no longer represented "good faith estimates of fair market value," according to the report. 
  • By 6/30/2011, cumulative M2M losses on the SCP stood at $5.8 billion.  
  • on 7/13/2011, JPM restated the first quarter's net income based on the inadequate M2M marks, reducing it by $459 million.
Going back to December 2011, J.P. Morgan Chase had begun thinking about Basel III's capital requirements.  Management's directive was to reduce risk weighted assets ("RWA") both in the Investment Bank and in the CIO.  According to the internal report, the CIO was known to be a large consumer of total RWA.  

At this time, the SCP was generally short credit.  The directives to SCP management were to reduce the overall size of the portfolio, and to make it more "credit neutral", in order to reflect bank strategists' more optimistic views about the economy.  When the CIO traders tried to reduce significant short high yield positions, they found the market illiquid.  The traders then came back to management and said that the directive would be too costly.  

Now, the CIO traders came up with an alternative strategy to accomplish the corporate directives, namely to go long investment grade corporates.  One of the key findings of the report was that the SCP traders designed a "flawed strategy."  Chief Investment Officer Ina Drew was cited for failing to vet, understand, and manage the SCP investment strategy and its implementation.  In addition, the CIO's risk management functions and finance functions were found to be "ineffective."  That is a very bland word for pervasive incompetence and laughable processes and controls.

An appendix to the internal report describes the VaR model used by the CIO.  Post-crisis, everyone knows the limitations of these models.  At the time, the particular models used inside the CIO had been designed to fit Basel I, but they were not in line with the requirements of Basel II and II.5.  The corporate executive management were already looking ahead to being Basel III compliant.

Apparently, there was a large dichotomy between the risk management functions inside the Investment Bank and those inside the CIO.  The CIO's VaR models, for example, gave very limited treatment to asset correlation risks, even though the traders knew when they did their monthly M2M that correlations were increasing thereby raising the riskiness of the portfolio.  

In August of 2011, the CIO hired an outside expert, a British "mathematician," referred to in the report as the "modeler."  Apparently, this secret agent lingo is to protect the identity of this person so as to not jeopardize potential prosecution in Britain.  The modeler apparently sold his engagement by saying that a "better model" would both deal more effectively with asset correlation and simultaneously reduce RWA.  This is intuitively implausible, given what the traders were finding in their daily operations, but it was what everyone wanted to hear. 

The modeler built his model in an Excel workbook, requiring manual entry of daily data.  Meanwhile, because the CIO was top-heavy and understaffed with high-powered analysts, data entry for daily fair market valuations was done by a junior trader.  In mid-March 2011, according to the report, this junior trader was directed by his manager to choose an outmoded, simplistic formula for FMV which lowered the reported losses for the period.  Remember from earlier in the post that losses went from $169 million in February to $718 million in March.  

Meanwhile, traders inside the CIO approached their managers and CIO executives about their compensation being at risk for the growing chorus of monthly bad news!  They were assured that this was not going to an issue, and that the issue would be treated fairly.  Incredibly, the internal report absurdly asserts, with no evidence, that compensation issues did not affect trading behavior inside the CIO.  Some things never change on Wall Street. They never will, Dodd-Frank and any other Rube Goldberg regulations notwithstanding.  Paul Volcker, where art thou?

The modeler's output was implemented, even though it was widely understood that the model had limited back testing.  Executive management believed that on the day of implementation, that the model's data input would be automatic.  It wasn't.  There were two specific areas of remediation, related to the valuation of illiquid tranches, which were supposed to be fixed before implementation.  They weren't.  On April 10th, a data input error failed to record a $400 million daily fair market value loss in the portfolio, until the error was later uncovered.  What a comedy of errors in the most powerful investment bank in the world.

The J.P. Morgan Chase board of directors rightly concluded that the CEO should be able to rely on the competence and ethics of his direct reports.  However, it also said that "...he (Dimon) could have better tested his reliance on what he was told."  The board had no choice but to take the approach of "the buck stops here" with CEO compensation, and it did. This did show that the board was not brain dead, but little else changed inside the organization, as other key actors exited stage right in a genteel way.  Ina Drew's past incentive compensation was subject to some clawback, but again, the CIO organization was untouched by any real consequences.

This was the case despite the report's concluding that (1) CIO managers failed to escalate their concerns to executive management, corporate officers and even to the board of directors; (2) traders had hidden the full extent of their losses, and (3) the CIO report to the CEO and his senior executives ahead of the first quarter 2011 conference call was "inadequate."  

As of January 19, 2013, here's how the Huffington Post described CEO Dimon's conference call with analysts:
"On calls with reporters and analysts Wednesday, he (Dimon) was his usual swashbuckling self, intensely proud of the bank he runs and sometimes impatient with critical questions.He said the portfolio where the troubled bets were made is "very close to being a non-issue" as far as trading losses are concerned. .../When analyst Guy Moszkowski asked about the "exotic investment strategies" of the Chief Investment Office, where the loss occurred, he shot back, "It has got not a damned thing to do with exotic investment strategies – zero, nada, nothing. OK?"
Unfortunately, Mr. Dimon probably scored the debating point against Mr. Moszkowski.  The investment strategies at the SCP were not at all "exotic."  To mash up Mr. Dimon's words with the findings of the report, the investment strategies were hare-brained, incompetently designed for the wrong purposes, not vetted by CIO management, implemented with Tinker Toy tools, left in the hands of unsupervised individuals, and reported to the CEO through folks whose economic interests clashed with their duty of candor.  All the same cultural issues remain. But, perhaps we are closer to future trading losses being a non-issue.

Looking for a bridge to buy?  

No comments: