Counterfactual: the tough-love bank bailout of 2008-2009

by Andrew Sprung

When I read assertions that the financial crisis has been "wasted," that is, not used to create a less risky, more accountable, more sustainable banking system, I usually think, "let's see how financial reform legislation turns out before we judge."

When I read that the bailouts of 2008 and 2009 created massive moral hazard and thus made the financial system riskier than ever, I wonder, "should the government have liquidated megabanks that appear not have been insolvent? Would the systemic risk have been worth it?"

John Gapper, writing in the FT, has an answer.  The bailouts may have been necessary. But they should have been much tougher on the banks.

Tarp may have achieved its financial aims but, in terms of systemic risk, it has failed. By propping up banks indiscriminately, on soft terms, Tarp not only outraged voters but also magnified moral hazard in the financial system and made effective reform harder.
"Given the severity of the crisis, the government had no choice but to intervene, but there were not enough strings attached and the moral hazard problem has worsened," says Matthew Richardson, a professor at New York University.

There was not much detail in the rest of the piece as to what kinds of strings should have been attached. I asked Gapper by email how the bailouts could have been handled differently, establishing along the way that a) he was not suggesting that any given bailed out bank should necessarily have been liquidated, and b) it would not do in the U.S. to break up bailed out banks by fiat, without new regulations mandating separation of banking functions (which he recommended in a prior piece). From Gapper's response:

...bondholders did not suffer at all. Could have been a haircut - In particular could have bailed out AIG CDS at less than par. Could have stopped dividend payments on equity. Could have insisted on shareholders co-investing as price for stopping institutions going bust. I don't think these would have solved moral hazard problem but the problem of giving everyone a 100 per cent bailout might have been eased a bit.

Gapper further referred me to the NYU White Papers Project. There,Viral V. Acharya and Rangarajan Sundaram elaborate on a criticism that Gapper mentioned in brief - that "Mr. Paulson lumped Goldman Sachs and Citigroup together."  Addressing whether the bailout terms were fair to the taxpayer, they make these points:

  1. By adopting a one-size-fits-all pricing scheme that is set at too low a level relative to the market, the US loan-guarantee scheme represents a transfer of between $13 billion and $70 billion of taxpayer wealth to the banks. In contrast, the UK scheme, which uses a market-based fee structure, appears to price the guarantee fairly.
  2. By offering very little in terms of optionality in participation, the US loan guarantee scheme is effectively forced on all banks, giving rise to a pooling outcome. The UK scheme, in comparison, provides considerable optionality in participation, which, combined with its pricing structure, has induced a separating equilibrium where healthy banks have not availed of government guarantees but weaker banks have. Implicitly, the US scheme encourages a system where banks are likely to remain (and to want to remain) on government guarantees until the crisis abates, whereas the UK scheme has paved the way for a smooth transition to market-based outcomes.
  3. The US recapitalization scheme has also provided little in terms of participation optionality for the large banks, but it too is otherwise generous to the banks in that it imposes little direct discipline in the form of replacement of top management or curbs on executive pay, and secures no voting rights for the government.

Simon Johnson, in  written testimony to the Congressional Oversight Panel on TARP last month, puts additional meat on the bones as to how to do a tougher bailout:

The money used to recapitalize (buy shares in) banks was provided on terms that were grossly favorable to the banks. For example, Warren Buffett put new capital into Goldman Sachs just weeks before the Treasury Department invested in nine major banks. Buffett got a higher interest rate on his investment and a much better deal on his options to buy Goldman shares in the future.
As the crisis deepened and financial institutions needed more assistance, the government got more and more creative in figuring out ways to provide subsidies that were too complex for the general public to understand. The first AIG bailout, which was on relatively good terms for the taxpayer, was renegotiated to make it even more friendly to AIG. The second Citigroup and Bank of America bailouts included complex asset guarantees that essentially provided nontransparent insurance to those banks at well below-market rates. The third Citigroup bailout, in late February 2009, converted preferred stock to common stock at a conversion price that was significantly higher than the market price – a subsidy that probably even most Wall Street Journal readers would miss on first reading. And the convertible preferred shares that will be provided under the new Financial Stability Plan give the conversion option to the bank in question, not the government – basically giving the bank a valuable option for free.
Note that this strategy is not internally illogical: if you believe that asset prices will recover by themselves (or by providing sufficient liquidity), then it makes sense to continue propping up weak banks with injections of capital. However, our main concern is that it underestimates the magnitude of the problem and could lead to years of partial measures, none of which creates a healthy banking system.

Obama may have been right that the Treasury's first principle was rightly "do no harm," and that we are better off for not having precipitously nationalized banks that were not clearly insolvent. But "do no harm" does not mean "inflict a minimum of pain."