Tuesday, July 17, 2012

JP Morgan CIO Loss Update – The Beer Brewer and the Design of the Bottle

When I started writing this post, I had the intention to write about what JP Morgan's CIO had actually traded and how these mechanism work. As you will see, my topic has slightly changed while researching and writing about it.

Last week, JP Morgan has published its 2nd quarter earnings. As its top management had announced to publish details about JP Morgan's recent proprietary trading loss with its 2nd quarter results, I was curious to read the documentation published by the firm last Friday.

The press release seemed promising as it quotes Jamie Dimon saying:

"Since the end of the first quarter, we have significantly reduced the total synthetic credit risk in CIO - whether measured by notional amounts, stress testing or other statistical methods. The reduction in risk has brought the portfolio to a scale that allowed us to transfer substantially all remaining synthetic credit positions to the Investment Bank . The Investment Bank has the expertise, capacity, trading platforms and market franchise to effectively manage these positions and maximize economic value going forward. As a result of the transfer, the Investment Bank's Value-at-Risk and Risk Weighted Assets will increase, but we believe they will come down over time. […] "CIO will no longer trade a synthetic credit portfolio and will focus on its core mandate of conservatively investing excess deposits to earn a fair return. […] The Firm has been conducting an extensive review of what happened in CIO and we will be sharing our observations today.”

Before I continue writing, let me clarify 2 things:

  • First, I don't like power point. The software might be helpful, if it is used as a real support tool for a presentation. However, all too often, it is not used this way...
  • Second, I did not attend the presentations made by JP Morgan executives. Therefore, I must rely on the written material that has been published afterwards. It might be that the oral presentation was perfectly clear. By all means, the published information is not.

Among the supporting documents that the bank has published, my attention was primarily caught by a document called “Presentation Slides - CIO Task Force Update”.

This presentation starts off with a summary of the key takeaways:

  • CIO judgment, execution and escalation in 1Q12 were poor
  • Level of scrutiny did not evolve commensurate with increasing complexity of CIO activities
  • CIO risk management was ineffective in dealing with synthetic credit portfolio
  • Risk limits for CIO were not sufficiently granular
  • Approval and implementation of CIO synthetic credit VaR model were inadequate

Next comes the (much expected) description of the facts:

According to the presentation, the synthetic credit portfolio intended to provide a partial hedge to credit exposures, included long positions to reduce the costs of the credit protection, and was adjusted over time to reflect changes in macro views. During 2007 to 2011, the portfolio steadily generated revenues to JP Morgan's bottom line.

In late 2011, the firm's top management instructed its CIO to reduce the firm's risk weighted assets. My understanding of the presentation is that the firm's CIO did not follow this instruction entirely. It

  • increased net long positions on investment grade indices;
  • increased short positions in some junior tranches for further default protection;
  • continued to increase size in an attempt to balance the portfolio and deal with market and P&L pressures, including perceived vulnerability to other market participants.”

The result was that, by the end of March 2012, the portfolio was exposed to

  • the relationship of Investment Grade to High Yield indices;
  • a default correlation across the capital structure (e.g. super senior vs. mezzanine);
  • the basis risk between off-the-run indices and on-the-run indices.”

Unfortunately, if you had hoped for a more detailed description of what happened (or still happens), you will get disappointed.

From the context of the presentation, you can conclude that the above slide talks about December 30, 2011 and March 30, 2012. Besides the problem of scale in these graphs, we can learn that

  • The long net notional portfolio size tripled: As we know neither the original size nor the exposure, I am not sure that this information is helpful.
  • Notional long net tranche positions tripled: About what tranches are we talking here? What is the original size? What type of exposure are we talking about?
  • Long net notional exposure to off-the-run indices also tripled: What indices are we talking about? Again, what was the original exposure?

Looking and thinking about the above slide, I had a hard time linking the facts to the observations. I tend to think, that, most often, there is simply no relationship at all. My key takeaways from this slide are the following:

  • Instead of reducing the riskiness of the synthetic portfolio, the CIO increased it.
  • The risk limits for CIO were not sufficiently granular, meaning probably that the portfolio was not diversified enough.

At this stage, I still don't know what the famous synthetic portfolio is about. Unfortunately, I will not know it either once I will have finished reading the presentation slides. As a matter of fact, I can read a lot about what CIO has done or should have done, about the risk culture and its implementation, about VaR models and their unsuccessful modification, etc. As an example for this discussion please have a look at the following slide.

Here is a last example of CIO Task Force presentation:

Is there any relationship between the above items? What is the message of this slide, if any?

I must admit that reading the CIO Task Force presentation was not fun. It's actually like listening to a beer brewer talking about any detail of the design of the bottle instead of talking about its content.


  • JP Morgan Press Release dated July 13, 2012
  • JP Morgan - Presentation Slides - CIO Task Force Update - July 13, 2012