There are many things that JPMorgan Chase
On traders' compensation and clawbacks
- "I don't believe that the compensation made this problem worse." (In response to the question of whether or not compensation played a role in allowing the trading losses to begin to mount. Dimon further justified this answer by saying that none of the employees were paid based on a formula and were instead paid for their performance as a whole.)
- "When the board finishes the review, you can expect we'll take appropriate action...I would say it is likely there will be clawbacks... [For senior people], we can clawback even for things like bad judgment. I was in favor of a clawback system."
Dimon also observed that everyone involved has had a long and stellar career at JPMorgan and that no one's pay has yet been taken back and that this will be a matter for the board to review.
Now, I understand that he was responding to a direct question, but whether or not the compensation system made things worse is irrelevant in the context of the wider problem, which is that traders' compensation structure remains fundamentally flawed: It's a case of heads I win, tails the bank, its shareholders (and taxpayers, potentially) lose.
First, it remains to be seen whether any clawbacks will be enforced in this instance. As far as I know, no bank has ever exacted clawbacks in relation to any high-profile trading loss (although it's not impossible that this has occurred in relation to losses that never made it into the press.)
Furthermore, even clawbacks won't do the trick unless a trader's entire net worth is at stake -- which would be impossible to implement. The problem with trading is that "a long and stellar career" can mean five or 10 years' of fat, consistent trading profits capped off by an "unlucky" year which wipes out all the accumulated profits -- and then some.
On value-at-risk
- "We don't as of today believe it was done for nefarious purposes." (Commenting on the change in the chief investment office's value-at-risk model, which reduced the value-at-risk from $129 million to $67 million.)
Again, Dimon was responding to a direct question, but whether the value-at-risk was $129 million or $67 million is completely irrelevant in the wider scheme of things. The problem isn't why anyone tinkered with the VAR model; the problem is the VAR model itself (this is clearly something that financial journalists haven't understood, either). Let me explain why.
VAR is a measure of a bank's maximum loss for a given level of likelihood. If the chief investment officer's 95% daily VAR was $129 million, that means the bank would expect the CIO's daily losses to be less than $129 million on 95 days out of a hundred. But what about the other five days during which one might expect the losses to exceed that figure? As Andrew Haldane, who is responsible for macro-prudential policy at the Bank of England, wrote in a recent paper (link opens PDF):
But VaR suffers a fatal flaw as a risk management and regulatory measure: it is essentially silent about risks in the tail beyond the confidence interval. For example, even if a trader's 99% VaR-based limit is $10 million, there is nothing to stop them constructing a portfolio which delivers a 1% chance of a $1 billion loss. VaR would be blind to that risk and regulatory capital requirements seriously understated.
This example is far from hypothetical. ... On 10 May, JP Morgan announced losses totalling $2 billion on a portfolio of corporate credit exposures. This took the world, and JP Morgan's management, by surprise. After all, prior to the announcement the 95% VaR on this portfolio in the first quarter of 2012 had been a mere $67 million. This VaR measure was revised upwards to $129 million on the day of the announcement. Whether this proves a more accurate measure of tail risk on this still-open position remains to be seen.
Unfortunately, it's difficult to imagine that JPMorgan will abandon VAR since it is the firm that unleashed this virus on the banking industry. A group within the bank called RiskMetrics developed the methodology in the early 1990s in response to a demand by JPMorgan's then-chairman Sir Dennis Weatherstone for a daily report that aggregated and summarized the bank's risk. (As if risk could be encapsulated in a single number!) The measure gained widespread acceptance with clients and competitors (and regulators). RiskMetrics was so successful it was eventually spun out of JPMorgan (in 2010, it was acquired by index and software provider MSCI).
On bank capital levels
- "We never argued about having more capital." (Though Dimon went on to say that he has complaints about how much more.)
Here, Dimon is referring to his opposition to the new Basel III capital requirements, which he has described as "anti-U.S." in the past. In a September 2011 interview with Financial Times, he objected to the additional capital buffer applied to so-called G-SIFIs -- Global Systemically Important Financial Institutions. The term refers to institutions who, by virtue of size and degree of interconnectedness pose a potential risk to the global financial system (naturally, JPMorgan falls in this category).
But what are the requirements under Basel III? As it applies to JPMorgan, the bank will need to achieve a 9.5% ratio of core Tier 1 capital to risk-weighted assets, of which 2.5% is a buffer due to the bank's size. That is certainly much higher than the Basel II requirements, but those were clearly inadequate. Furthermore, it's hard to see how 9.5% could be considered too onerous, as I argued here. If anything, that floor remains too low, particularly when one considers that it is based on risk-weighted assets -- another "sophistication" we could probably do without.
Not all banks are equally well run -- our free report identifies the stocks only the smartest investors are buying.
