Notes on JPM Whale Watch from Alphaville

Financial Times Alphaville had poured through the JPM Whale Watch Report.   Some new revelations

• At the end of 2011 the JPM higher ups wanted to reduce the Risk Weighted Assets (RWA) of the CIO from USD 43 BB to USD 25 BB in the upcoming year.  Management had a generally positive view of the economy so it was thought they could do without some default protection.
• Initially JPM's traders started reducing positions but after that undertaking proved expensive JPM Chief Investment Officer Ina Drew told the traders to be more sensitive to profits.  The head trader told his underlings to focus on PnL.
• Mid January the head trader was told to make sure they were well positioned for future credit events.  Alphaville speculates that this directive may have been the result of Kodak's January filing for bankruptcy.
• The traders decided to hedge their purchases of insurance on high yield names by selling protection on investment grade names (that was previously speculated here and here).  Throughout February the JPM traders sold protection on investment grade and increased their purchases of protection on high yield names.
• There may have also been a curve trade in which they sold USD 20 BB in investment grade (IG-9) protection for 10 years and purchased USD 12 BB of investment grade (IG-9) protection for 5 years.  This trade may have been unbalanced for two reasons (see below).
• They hoped this combination of transactions would 
• reduce RWA (the two trades are somewhat correlated)
• position them well for further defaults
• earn profits (ok that didn't quite work)

• By late March / early April management became concerned about both negative PnL and the integrity of the marks being used to calculate that PnL.

The report refers to a curve trade in which they sold USD 20 BB in investment grade (IG-9) protection for 10 years and purchased USD 12 BB of investment grade (IG-9) protection for 5 years.  This trade may be unbalanced for two reasons.  If the instantaneous probability of default were to increase equally across the forward curve (ie a parallel shift in the forward curve) then the price of longer dated protection would increase more than the price of shorter dated protection.  A balanced hedge would then require more of the short dated protection to balance less of the long dated protection.  JPM's trade had more long dated protection against less short dated protection.  Secondly if the instantaneous probability of default were to increase by 1% in the front of the forward curve how much would it increase in the back of the forward curve (ie does the forward curve move in a non-parallel fashion?)?  In physical commodity markets the front of the curve tends to move more than the back.  However interest rate forwards move more in the middle of the curve.  I am not sure about credit markets.