On the subject of bank capital, I can't stress enough that the proper approach is for government to restrict even temporary, fully-collateralized assistance only to those institutions that are clearly solvent, and to promptly restructure the other institutions. What the global economy needs most is not bank bailouts, but to establish and enforce a legal and regulatory structure that allows the streamlined bankruptcy of insolvent institutions (Title II of Dodd-Frank addresses this with a more comprehensive policy than existed in 2008, but it doesn't read as a "clean" solution in my view - putting too many cooks in the kitchen - particularly the Fed and the Treasury).
Again, again, again, the "failure" of a financial institution only means that the institution fails to pay off its own bondholders. Depositors typically lose nothing. For example, "saving" Bear Stearns meant primarily that Bear Stearns' bondholders would be made whole. Saving Dexia a few weeks ago meant the same thing for Dexia's bondholders. The key is not to prevent "failure," but to prevent disorderly failure and piecemeal liquidation. Washington Mutual was a seamless, and therefore nearly unmemorable "failure." Lehman was disorderly and jarring. The difference was that there was a legal and regulatory structure to quickly cut away stockholder and bondholder liabilities in the Washington Mutual instance (which was handled by the FDIC), while there was no similar way to restructure non-bank financials like Lehman in 2008.
From my perspective, weak regulation of bank leverage, inadequate capital requirements, and the need for prompt, streamlined restructuring for insolvent banks are among the most urgent problems that the global economy faces. Consider this. The Financial Times reported on Friday that in 2008, Dexia lent 1.5 billion euros of its capital to two institutional investors, who used the cash to buy newly issued shares in ... wait for it ... Dexia. Remember that as a bank, Dexia operated at leverage of about 50 times its tangible shareholder equity (see last week's comment ). So Dexia's maneuver made it possible to meet regulatory capital standards and take on a huge amount of additional leverage, without actually raising any bona-fide capital. As FT noted, "The unorthodox funding move, which roused Belgian regulators' concern at the time, amounted to Dexia borrowing money from itself to finance a capital increase. This is illegal in most jurisdictions and is now banned in the European Union, but did not break Belgium's existing laws."
On a similarly outrageous note, Bloomberg reported last week that " Bank of America , hit by a credit downgrade last month, has moved derivatives from its Merrill Lynch unit to a subsidiary flush with insured deposits... The Federal Reserve and the Federal Deposit Insurance Corp. disagree over the transfers, which are being requested by the counterparties. The Fed has signaled that it favors moving the derivatives to give relief to the bank holding company, while the FDIC is objecting. The bank doesn't believe regulatory approval is needed." Well, other than that it goes against Section 23A of the Federal Reserve Act , but then, the Fed can make an exemption whether the FDIC likes it or not . And that's what we've come to - government of the banks, by the banks, and for the banks (because banks are people too) .
The Bloomberg report continued, "B ank of America's holding company -- the parent of both the retail bank and the Merrill Lynch securities unit -- held almost $75 trillion of derivatives at the end of June, according to data compiled by the OCC. About $53 trillion, or 71 percent, were within Bank of America NA [the FDIC insured entity], according to the data, which represent the notional values of the trades. That compares with JPMorgan's deposit-taking entity, JPMorgan Chase Bank NA, which contained 99 percent of the New York-based firm's $79 trillion of notional derivatives."
Note that the figures are in trillions, not billions (U.S. GDP is $15 trillion). That said, the vast majority of the "notional value" of derivatives in the financial system represents multiple fully-hedged links in a long chain between final users who actually take the risk, so Bank of America's true risk is most probably a tiny fraction of that notional amount. Unless those derivatives include unhedged short positions in credit default swaps on Greek debt (which we can't really rule out), it's not clear that the derivatives themselves are underwater. The real problem, in my view, is that the transfer is clearly driven by the intent to get around capital adequacy regulations, and runs precisely opposite to the right way to create a good bank and a bad bank . It saddles the good bank - the taxpayer insured one - with the questionable liabilities, while "giving relief" to the holding company. This is really preposterous.
As a final note, it's worth observing that a number of banks reported positive "earnings surprises" last week. If you look at those results for any of the major banks, it is immediately clear that the bulk of the earnings were of two sources: further reductions in reserves against potential loan losses, and an accounting gain known as a "Credit Valuation Adjustment." Those two items, for example, were responsible for nearly 90% of Citigroup's reported "earnings." The Credit Valuation Adjustment (CVA) works like this: as the bond market has become more concerned about new financial strains, the bonds of U.S. banks have sold off significantly in order to reflect higher default probabilities. Under U.S. accounting rules, bank assets are no longer marked to market, but happily for the banks, the decline in the market value of their bond liabilities means that the banks could technically "buy their bonds back cheaper." So the decline in the bonds, despite being due to an increase in investor concerns about bank default, actually gets reported as an addition to earnings! Surprise, surprise.