When the London Whale thing came out, JPMorgan made one sort of clever attempt to minimize it by saying this:
Since March 31, 2012, CIO has had significant mark-to-market losses in its synthetic credit portfolio, and this portfolio has proven to be riskier, more volatile and less effective as an economic hedge than the Firm previously believed. The losses in CIO’s synthetic credit portfolio have been partially offset by realized gains from sales, predominantly of credit-related positions, in CIO’s AFS [available-for-sale] securities portfolio. As of March 31, 2012, the value of CIO’s total AFS securities portfolio exceeded its cost by approximately $8 billion. Since then, this portfolio (inclusive of the realized gains in the second quarter to date) has appreciated in value.
What did this mean? Well, I think it roughly meant what it said, which is that as if March 31, JPMorgan's Chief Investment Office had about $375bn worth of bonds for which it had paid about $367bn, and that after March 31 (1) that portfolio of bonds increased in value to at least $375,000,000,001 and (2) JPMorgan had sold at least some of those bonds at a profit. But one nice thing about it is that, if you squinted, you could read it as "our hedge decreased in value, yes (and by $2bn), but that's because the underlying portfolio increased in value (by $8bn), so net-net we're way ahead, and it was a hedge, and whaddarya gonna do, hedges go down when things-hedged go up, that's life." That turned out to be an entirely wrong reading but hey they tried!
Reuters moved that story forward a bit with this kind of interesting parsing of Jamie Dimon's words, including particularly the statement that JPMorgan had realized $1bn of gains on the CIO portfolio of available-for-sale securities between the end of the first quarter and the beginning of that super-awkward whale-confession conference call. They concluded, based on realized profits on sales of AFS securities in previous quarters, that this probably meant JPM had sold about $25bn of those securities so far in Q2 of 2011. Which! Is a lot of securities. Slightly more than it sold last quarter (just under $20bn, with another $32.2bn maturing), though much less than it sold in Q2 of 2011 ($43bn, with another $39.9bn maturing). So, sort of an average amount, really.
Anyway, there's some hint that these possibly-slightly-higher-than-usual-though-also-maybe-not sales of AFS securities sacrificed shareholder value - in the form of paying taxes on realized gains and also, strangely, in the form of "eliminating future earnings from the securities" - to meet Wall Street's desire for a particular accounting EPS number: "rather than creating new value for investors, the transactions merely shift gains in securities from one part of the company's financial statements to another."
The eliminating future earnings thing ... puzzles me.* The taxes thing may also be a bit of a red herring - lots of JPMorgan's AFS securities mature each year, so the cost of accelerating taxes on $1bn of gains is not exactly $380mm** but whatever the present value is of paying those taxes now rather than in a couple of years - but it's hard to object to the basic point, which is basically "JPMorgan is sacrificing real money, in the form of tax payments, in order to earn fake money, in the form of transforming accumulated other comprehensive income into accounting income."
You could read this as a story of "equity investors are idiots and need to be fed a stream of steady GAAP earnings even if those numbers are bullshit" and, sure, I mean, that's a pretty popular story about equity investors and I guess not untrue? I'd add another story though, this one mostly about banks. In my earlier attempt to make sense of Whaledemort I suggested that part - just part! - of the reason for the complexity of his hedging program, which after all was not just "buy protection against the stuff we lend to" but was instead "buy some protection and sell other protection" - was the desire to smooth accounting income. The trade wasn't to make loans and buy bonds, and hedge the risks by buying CDS against those credits: it was to make loans and buy bonds, and hedge the risks of a serious credit downturn with some sort of convexity trade. There were lots of reasons for that, from profitability to existential what-is-a-bank reasons, but one annoying and obvious reason is that mark-to-market changes in CDS prices basically flow through net income and EPS, while mark-to-market changes on things like loans or the AFS securities portfolio basically don't.
What that means is that if your loans perform well, which after all is what you want (and should happen, like, more than 50% of the time?), then you have an accounting loss on your hedge which is not offset by an accounting gain on your things-hedged. It turns out that this matters, for banks, not only because stupid equity investors are stupid and stupidly want their stupid numbers to steadily increase each stupid quarter, but also because banks care deeply about meeting their capital requirements, since doing so lets them return money to shareholders and not doing so lets them not return money to shareholders and also get made fun of. Bank capital requirements are not based on some platonic ideal of economic value but rather on particular measures of how much equity banks hold. And one rule of the measurement turns out to be that gains and losses on AFS securities don't influence capital: bank capital includes accumulated other comprehensive income (where AFS gains and losses live), but backs out unrealized securities gains from the amount of capital that you can claim.
Banks generally love this because it reduces volatility of capital, but it may have contributed to Whaledemort's efforts to find a hedging program that (he thought, or hoped, at least initially, caveat caveat caveat) would not show significant mark-to-market swings in the absence of huge defaults: because the things hedged would not show any offsetting accounting gains. Now, though, the same rule might bail JPMorgan out: it was able to bank a lot of unrealized gains in its AFS portfolio, and could thus sell the most-appreciated securities to realize $1bn of gains this quarter and undo some of what the Whale wrought. This, perhaps, destroys some value for shareholders, what with the taxes and so forth, but it's important to recognize that it also creates real value for shareholders, and not just in the form of a warm glow around a stable EPS number. It also gets JPMorgan that much closer to repairing the capital damage done by the Whale and thus restarting its share repurchase program. Which, unlike accounting gains either in or out of net income, actually puts money in shareholders' pockets.
* Here is a thing that the former chief accountant of the SEC said:
"They really made two stupid decisions," said Lynn Turner, a consultant and former chief accountant of the Securities and Exchange Commission. The first was taking risks with derivatives that they did not understand, Turner said.
"The second is selling assets with high income that they can't replace," Turner added. In a low interest-rate environment, the bank will struggle to generate as much income with the cash it received from selling the securities, he said.
So! Correct me if I'm wrong but I'm pretty sure that if the market yield for your assets is 6% and you sell them at 6% you can probably go find similar assets at about a 6% yield because um that's how come the market yield was 6%? The fact that you bought them at a 9% yield three years ago is why you can sell them for a profit now, but they're not yielding 9% now oh never mind. Chief accountant of the SEC.
** Because JPMorgan's effective tax rate is 38%! Ha! Ha!