The whole credit squeeze triggered off by the sub-prime mortgage defaults, assumed alarming proportions with the attendant flight to liquidity by the financial institutions. The massive leveraging (Morgan Stanley boasted $33 of assets for each dollar of capital) during the bubble years magnified the impact of this crisis. The unwinding of positions, with banks scrambling to shore up their steeply depleted reserves, led to massive sell-offs of securities, thereby driving them to bottom levels and leaving many of the asset backed securities virtually worthless and illiquid, without an active trading market. The deleveraging also left borrowers with a shrinking supply of credit and that too accessible at exorbitant prices. This left many of the biggest financial institutions vulnerable to being wiped out and necessitated the government to step in with bailouts.
The original $700 bn Troubled Assets Relief Program (TARP) had sought to identify and then auction the impaired assets on the banks' books and thereby establish a clearing price for them. It was soon abandoned in favour of direct cash injections, in return for some equity stake, so as to enable the banks to recapitalize and then use it to buy some of these impaired assets at a steep discount. With even this not being enough the Fed and Treasury took the extraordinary step of direct interventions of "quantitative easing" through direct purchases of distresed assets and guarantees for deposits and assets, and has so far committed more than $8 trillions for it.
But these bailouts ran into problems on many fronts, none of which were addressed by these bailouts. For a start, the deleveraging and the resultant plummeting of asset prices, had left many assets worthless, illiquid, and with no buyers. It did nothing to alleviate the masive credit and counter-party risks that had frozen the credit markets. Though it had hoped that the bailouts would recapitalize the banks and thereby encourage them to buy up the distressed assets on the cheap, this hope did not materialize as the extent of crisis was much deeper than anybody had fathomed.
In this context, three broad models for taking out distressed assets and bailing out weakened financial institutions have emerged. Let us examine the details of all three.
1. Citibank and Bank of America (BoA) model
This arrangement seeks to inject capital into the distressed banks, in return for some form of stake and commitments on dividend payout and executive compensation, and also provide a sharing (between the bank and government agency) of guarantees for the non-performing or illiquid assets.
The Citigroup bailout provides $40 billion in fresh capital and capital relief (in return for preference shares), and it ring-fences $306 billion of illiquid assets on Citi’s $2 trillion balance-sheet. The bulk of any losses on these, beyond the first $29 billion, will be borne by government agencies.
In the $118 bn BoA deal, the US Government decided to supply BoA with a fresh $20 billion capital injection and absorb as much as $98.2 billion in losses on toxic assets, including residential and commercial real estate and corporate loans. BoA will be responsible for the first $10 billion in losses on a pool of $118 billion in illiquid assets, while the Treasury Department and the Federal Deposit Insurance Corporation will take on the next $10 billion in losses. The Fed will absorb 90% of any additional losses, with BoA responsible for the rest.
In both cases, the Fed will buy up the distressed assets by creating specifically funded entities, which will in turn buy up the mortgage backed securities. This will effectively turn the Fed into a direct lender to homeowners.
In both cases, the equity infusion will be in return for preference shares and commitment to cut back its dividend payout, accept a loan-modification program and put more stringent restrictions on executive pay. The preference share route, instead of taking ordinary common stock shares, was to minimize risk (preference share holders come before common stock holders) and avoid playing a role in the day-to-day management of the banks so assisted. In contrast, the British bailout plan sought to take direct equity stake and ownership of the banks assisted, often majority control. However, in both cases, the bad assets will continue to remain in the balance sheet of the banks.
The problem with this model is that it does little to ally investor fears about the health of the bank, since the distressed assets, whose value is itself uncertain, continue to remain within te books of the bank. Further, it also does nothing to establish a clearing price for the impaired assets on banks’ books nor facilitate its trading. Finally, such bailouts, as in the case of the Citigroup, covers only a portion of the distressed assets, mortgage backed securities here, but leaves out the other similarly dubious assets like its huge credit-card and overseas-loan portfolios, and the numerous off-balance sheet exposures. Its piecemeal approach also creates dangerous moral hazard concerns.
2. Bad bank model
Ring fence all the troubled and non-performing assets of each bank separately and set up a "bad bank", a separate entity which takes ownership of these assets and then manages them in order to maximise their value. This is being tried out in UBS now.
The generic variant of this is to set up a universal "bad bank", which would 'buy up' the non-performing assets from different banks at "fair value", and thereby hope that
a) the uncertainty surrounding counter-party risk is eliminated and buyers know what they are buying
b) a price discovery mechanism for distressed assets is established
c) provide liquidity to a market for trading these assets
This example was successfully tried out in Sweden in the early nineties, and its efforts to restructure and resell distressed loans has been acclaimed a success. This cleaning up the books of banks is likely to restore some semblance of normalcy and thereby make it easier for these banks to raise capital or attract buyers. It is thought that this will have a huge psychological impact on the markets, in so far as it signals a clean break with the past.
In the US, it is now being proposed to set up a similar bad bank, or "aggregator banks", created and capitalized by the federal government, with the sole purpose of buying up bad assets and warehouse them in one place. The precedent cited is the Resolution Trust Corporation (RTC), set up in the aftermath of the Savings & Loans banks in the eighties.
The problem with this model of valuing and isolating bad assets, is what assets to be separated and how to value them. Assigning a "fair value" is an almost impossible task. Unlike the earlier precedents, this time we are dealing with fiendishly complicated structured mortgage assets with risks dispersed far and wide, even identifying and locating their risks, leave alone valuing them is virtually impossible. Since many of these assets are not traded in exchanges and others have been not traded for months and are illiquid, there is no way of valuing them with any sense of accuracy.
Any attempt to buy them by assigning them some valuation, will most likely end up adversely affecting the interests of tax payers, as the government is more likely to end up over-paying. Further, since it is not proposed to seize the distressed banks before the "bad bank" is set up (unlike in case of RTC), such bailout will reward the current owners and shareholders, by offering them the possibility of future gains in the good bank, besides encouraging moral hazard concerns. Paul Krugman has described this proposal "Hankie Pankie II" and compares it to the once abandoned Super-SIV plan.
3. Temporary Nationalization
To the extent that the purpose of all these bailouts is to get the distressed assets off the balance sheets of banks, and thereby re-establish normalcy in the credit markets, there are very serious problems with all these proposals. Stripped off all jargon, the simple reality is that the assets of many major banks are much less than their liabilities. In the circumstances, the only way for the Government purchases to make the banks solvent again is to pay much more than what private buyers are willing to offer (so as to atleast cover up the differential) and also take huge risks on board with the attendant possibility of even more cost for the tax payers.
Paul Krugman has argued that the least painful way to get this done is for the banks themselves to either write them off to zero or to sell them off at whatever price they get. This will penalize greedy shareholders and owners who are likely to get wiped out or have to exit at huge losses.
He has cited the example of the RTC, wherein the Government seized the defunct banks, cleaned out the shareholders, transferred their bad assets to RTC, paid off enough of the banks’ debts to make them solvent and sold the off to new owners. In Sweden, the Government took over whole banks, and then hived off the bad stuff without any kind of valuation at all, and then left them to sit for a while before selling them off. In Ireland in the nineties, the Government had taken over all the bank liabilities, by guaranteeing their unsecured debts, while leaving the assets on the banks' balance sheets, thereby bailing out the debt/bond holders in a massive way. However, the FDIC guarantees in the initial bailouts, along similar lines, was alleged to have been too favourable to bondholders.
All this brings us to probably the bailout of last resort - nationalization! The benefits are many - minimized moral hazard, cheaper cost of capital for the new banks, transparent, least cost to the tax payer, etc. Many influential voices, including even The Economist, have added their weight to nationalization. Awkward Corner and FT Alphavile points attention to the debate on nationalization with links.
Chris Dillow clarifies on the conventional responses against nationalization here. Of all these, Felix Salmon offers the most convincing case in two brilliant posts, where he gets most of the points right
"Since the government has an interest in protecting its own liabilities, rather than maximizing shareholder value, the chances of crazy gambles will be minimized. In any case, since the government is taking virtually unlimited downside, it should by rights have all the upside as well - i.e., ownership. It's (nationalization) transparent and easy to understand: if a bank is insolvent (and the FDIC is good at making those determinations), then simply nationalize it. That's what the Swedes did, and that's what we should do too."
And more here
"Nationalization is what the Nordics successfully did in the early 1990s; it can work here, too. The only real question is how much to pay: my gut feeling is that shareholders should get nothing, holders of preferred stock should take a serious haircut, and that if there aren't very many of those, then unsecured senior creditors should take a modest haircut too. The point of the exercise is to minimize the degree to which the banks' bondholders are bailed out, while at the same time minimizing the systemic consequences of a massive bond default by the likes of Citigroup and Bank of America.
But that decision is secondary. The big decision is whether or not to nationalize - or, rather, whether to nationalize now, and get it over and done with, or to continue to construct ad hoc responses in a desperate and probably doomed attempt to avoid having to nationalize at all. Surely we can all agree that if it's going to happen, it's better it happens sooner rather than later."
NYT makes an interesting point, that nationalization has to be an either-or solution - you nationalize all the banks or none at all - since given the wobbly state of all major banks, any nationalization of a few, would see the floodgates open for the others as investors and depositers exit from them.
Free Exchange has a nice summary of the two fundamental problems associated with any apprioach that involves asset purchases or ring-fencing off assets, like a "bad bank" - which assets are distressed and how to value them? The closer they get to transparent market pricing, the bigger the capital shortfall in the system will look! The narrower the scope of any bail-out, the smaller the up-front bill, but the greater the risk that more will be needed later—and thus the greater the market uncertainty!
The NYT has this article which sums up the central debate in bailing out distressed financial institutions - how to value bad assets?
Nouriel Roubini too feels that nationalization is the only meaningful way forward to clean up bank balance sheets without compromising the interests of the tax payers. He has calculated that they need $1.4 trillion in new cash to return their capital levels to where they were before the crisis began.
William Buiter, Paul Romer and James Kwak discuss the "good bank" solution - instead of creating a government entity to buy toxic assets from existing banks - or nationalizing existing banks, removing their toxic assets, and then reprivatize them - create brand new, good banks with the same government money, enabling them to lend money unencumbered by previous bad decisions, and then privatize them. David Warsh sums up this debate here.
An analysis of the Swedish model here. The authors argue that the Swedish model followed the principles of transparency of asset losses up-front, and honest communication about the extent of public intervention; politically and financially independent receivership; maintenance of market discipline; and restoration of credit flows.