MAKING SENSE -- June 13, 2012 at 8:35 PM EDT
Breaking Down the Finer Points of Jamie Dimon's Testimony
President and CEO of JPMorgan Chase, Jamie Dimon testifies before a Senate Banking Committee hearing on Capitol Hill Wednesday. The committee is hearing testimony from Dimon on how JPMorgan Chase lost over $2 billion in stock market trades. Photo by Mark Wilson/Getty Images.
Since the JPMorgan scandal broke, I've been part of an email thread featuring a few of the country's leading finance economists -- all of whom have been featured here on Making Sen$e at one time or another.
This morning promised an appearance by JPMorgan CEO Jamie Dimon before the Senate Committee on Banking, introduced by the ever-acerbic "Tyler Durden" over on Zero Hedge as, "The crony capitalist show must go on: those bribed by Jamie Dimon are about to ask questions of the same person."
My correspondents are rather more measured, and so I wrote to ask them for comment.
A first key question: Was Dimon telling the truth, the whole truth and nothing but the truth? Doug Dachille, who used to run JPMorgan's Proprietary Trading and Derivative businesses, didn't think so.
First, a bit of back story. The nub of the scandal -- the mulit-billion dollar loss which Dimon acknowledged today, could reach $5 billion: the bank having first bought, and then later started to sell, the equivalent of catastrophic health insurance. Only in this case, the catastrophe would be another worldwide credit crash, a la 2008.
Dimon explained that JPMorgan has a huge passel of loans out there. Obviously, if the bank's borrowers can't repay, it's in trouble, as any bank would be. So JPMorgan hedged its credit risk by purchasing a form of insurance, a so-called "synthetic credit portfolio" that would pay off if a representative pool of credits (bonds) threatened to default. Hence, the term "credit default swaps." (The "swaps" was a term slapped on to avoid regulation by aggressive insurance regulators. In reality, it's just "insurance" by another name, which smells just as sweet if catastrophe strikes.)
If this isn't complicated enough, listen to what happened next, according to Dimon. International regulators wanted banks like JPMorgan to be safer by holding more capital. But those credit default swaps (or, CDS) were classified as risky, and he told his chief investment office "to reduce risk-weighted assets and associated risk."
Now comes a rather remarkable admission:
"To achieve this in the synthetic credit portfolio [that's the CDS insurance, remember], the chief investment office could have simply reduced its existing positions. Instead, starting in mid-January, it embarked on a complex strategy that entailed adding positions that it believed would offset the existing ones."
If you're keeping score at home, this means that instead of selling the credit default swaps it had already bought, JPMorgan starting selling credit default insurance. Which raises the question: why would it do that instead of just getting rid of the insurance it had already written? Unless, to raise one possibility: it thought it could make money on both transactions. Or -- another possibility -- because it had sold insurance in such a quirky deal, it couldn't get anyone else to buy it?
But the result -- having both written and bought credit default insurance -- caused Sen. Jack Reed, D-R.I., to ask Dimon this morning, "How can you be on both sides of the transaction?"
Dimon's answer? "I won't defend it."
But in his prepared testimony, Dimon had described the original purchase of credit default swaps as insurance as "designed to generate modest returns in a benign credit environment and more substantial returns in a stressed environment."
A "hedge" against catastrophe isn't supposed to generate any returns at all!
As former JPMorgan CIO head Doug Dachille put it: "If the synthetic credit portfolio was a true hedge, then in a 'benign' credit environment you would not make money. You would lose money. Insurance requires the payment of premiums and if the event you purchased insurance to protect does not occur, you lose your premium."
Sen. Tim Johnson, D-S.D., asked this morning:
"Or was the goal really to make money? Should any hedge result in billions of dollars of net gains or losses, or should it be focused solely on reducing a bank's risks? As the saying goes, you can't have your cake and eat it too."
(Dimon never answered this question, as it was in Sen. Johnson's opening remarks.)
A somewhat more technical version of the same point comes from Nobel laureate MIT professor of finance Robert Merton:
"Although I can accept the reality of prop trading in banks, I do have a problem if there is outright deception. The following is a JPM description of the function for C.I.O.: Our Chief Investment Office is responsible for making investments to hedge the structural risks of our balance sheet on a consolidated basis.' (JPMorgan Comment Letter on Section 619)
This description does not say anything about that unit being a major (or any other kind of) profit center for taking proprietary risks or even that the mission is a combination of hedging the structural risks and earning as much profit as possible. If it were the latter, then it doesn't work anyway since hedges that reduce risks are not expected to be profit-making. Instead, the positions taken elsewhere in the bank [such as loans] that are being hedged by CIO are supposed to make the profits."
Now for those who need a short break at this point, I recommend watching our dear friend and sometime collaborator Merle Hazard for a few minutes as he sings his country-and-Wall Street classic, H-E-D-G-E. It too describes hedges that weren't, but at a firm a good deal more vulnerable than Jamie Dimon's.
OK, still with me? Let's deal with a second bone of contention seized upon by our finance maven correspondents: compensation.
To Bob Merton, this is a key forensic issue in determining what went wrong. It has been widely reported, for long before the scandal hit, that an audacious JPMorgan trader in England named Bruno Iskil, aka "Voldemort" and "The London Whale," had been selling credit default swaps for the bank's account. How was he paid? Merton asks.
"If, for instance, the bonus system for the [London] Whale and his boss was primarily a percentage of gross profits of their trading book, then that is clearly not the incentive system for a hedging operation. If it turns out this was just proprietary trading dressed up in corporate hedging clothes, then I would support very stiff penalties, not for losing money but for the deception of regulators, shareholders and public policy intent."
Harvard's Bob Glauber, formerly number two at Treasury in the George W. Bush Administration and then head of the independent Wall Street regulator FINRA, suggested these questions for Dimon, none of which got asked this morning:
"If the purpose of the synthetic portfolio was to hedge, did you compensate those traders running it differently from those running the rest of the portfolio? Do traders whose job it is to make money get compensated differently from those whose job it is to hedge?"
Along these same lines, MIT's Andy Lo, asks:
"Were the individuals directly involved in managing the CIO compensated according to the profitability of the unit, or was their compensation based on some measure of their success in hedging the entire company's earnings or assets? The answer should tell us whether the unit was hedging or proprietary trading. Obviously, there's nothing wrong with prop trading, and a $2 billion loss on a $380 billion book is not a big deal from a prop-trading perspective, but that wasn't how the CIO was described.
Is the compensation for the new head of the CIO tied at all to the performance of the CIO assets. If so how? Same question for the chief risk officer."
So while reasonable minds might differ, in this case, they didn't. All our Making Sen$e experts, including Doug Dachille, agreed that compensation is key. If JPMorgan (or anyone else) is going to "pay for performance," the pay formula has to provide an incentive to do the right thing.
I have to admit, I was much taken this morning with Jamie Dimon's strong assertion that yes, there would be "clawbacks" -- taking back compensation from those who had taken undue and costly risks.
"I was glad to hear there are clawbacks," said Chuck Schumer, D-N.Y, himself sometimes derided as 'the Senator from Wall Street' Can you tell us a little bit about the policy?"
"We can claw back unvested stock and even bonuses," said Dimon.
Schumer: "Is there a limit to the time period for clawbacks and have you ever done any?"
Dimon: "It's limited to the past two years and no, it's a new policy. We haven't done any yet."
Both Schumer and I were satisfied. To listen to him, Dimon is a clear, forceful and believable guy. "Clawbacks" could make a real difference. So I emailed my group one last time. That's when Bob Merton really let me have it.
"Clawbacks are not a substitute. Clawbacks have some use if the basis for a bonus [incentive] is properly set up. My point is not whether they have some skin in the downside. That's necessary, but it is second order to the right bonus schedule. You could measure performance and give bonuses in buying insurance for the company. But it would be based on a number of factors."
A first measure, Merton suggests, might be the price of the insurance the trader buys. In other words, does she or he get a good deal on the insurance premiums?
Another measure: did your trader pick a reliable insurer who actually pays off in the event of disaster?
But, Merton insists:
"You would not want the bonus to depend on the gross profits of the desk: insurance payments for losses minus the premium paid. It's the wrong metric. Paying a percentage of profits (even with a clawback) works for proprietary trading but it does not work for hedgers because you want them to protect the risks of other bank businesses. So the CIO desk could have a very large gross profit but not from any kind of hedging positions at all. And it could have large gross losses when it did a perfect job of hedging the other businesses. Gross profit of the desk is neither necessary nor sufficient to judge whether that operation is being performed well or poorly. So it is just wrong. How could one possibly measure whether the performance of the hedging operation was doing its job without integrating the performance of the portfolio with the performance of the businesses it supposed to be hedging?
"If the bonus of the Whale and his boss were not directly linked to the accuracy of the hedging activity then it was a flawed compensation design. If the bonus was on gross profits it would give the incentive and signal to act like they would as proprietary traders. Clawback on the wrong bonus system is useless and a red herring.
"BTW, didn't the banks say post-2008 that they would have clawbacks for traders? [Yes, Bob, they did.] Did Dimon say that they had clawbacks in place and that the Whale and his boss lost substantial accumulated bonus from prior years as a result of this event? [No, Bob, he didn't.] Or is it a statement about what we will do in the future? [Er...yup.]"
Note: This post has been updated to correct the spelling of Sen. Johnson's name.