Streetwise Professor

February 7, 2009

The Not So Bad, the Bad, The Really Bad, and the Really REALLY Bad

Filed under: Derivatives,Economics,Politics — The Professor @ 3:50 pm

“Stimulus” package or no, the economy will not resuscitate until the financial markets, and especially the banking system, are stabilized and recapitalized.  Several alternative approaches are on the table.  In this post, I’ll give my appraisal of each.  Sad to say, the leading candidates are the least attractive, in my view, and the most appealing ones have received little traction.  

The four alternatives are (in order from most awful to least bad):

  • Nationalization
  • Government guarantees
  • Bad bank (with a SWP twist–call it “Humpty Dumpty Light”)
  • Bad bank on steroids (“Full Monty Humpty Dumpty”)

Nationalization would involve a government takeover, for some indefinite period, of banks.  Equity holders and bondholders would be zeroed out.  When this option is proposed, it is often said, in a soothing fashion, that the seized institutions will eventually be privatized.

The government (in the US) already has a regime for nationalization: the FDIC can seize banks under certain conditions.  The nationalization proposals being mooted would go beyond this.

Nationalization has one advantage–no bailout of shareholders of the seized institutions (and the bondholders as well).  But it has two major disadvantages, namely: the likely politicization of resource allocation decisions, and soft budget constraints.  Together, this combination is quite dangerous.

Nationalized banks would certainly be under pressure to make lending decisions based on political criteria.  Like FNMA and FHLMC?  Like the CRA?  You’ll love nationalized banks!  Want Barney Frank and Chris Dodd exerted disproportionate influence over how capital is allocated in the US?  Nationalization is for you!

I don’t believe that Fannie and Freddie and the Community Reinvestment Act were the sole, or even the primary causes of the financial crisis.  But they were definitely major contributory factors.  Pressures on financial institutions to achieve political objectives, primarily in housing, contributed to the perverse lending practices that culminated in the subprime meltdown.  

With nationalized banks, these pressures would be felt in all aspects of lending, not just housing.  The banks would become levers that politicians pull to favor influential constituents, and to engage in industrial policy.  

The resulting loans would, almost certainly, be far likelier to default than loans issued based on economic criteria by profit maximizing banks.  That’s where the “soft budget constraint” comes in.  Nationalized banks in financial difficulty would almost certainly receive government financial support, further eroding market discipline and enabling continued extension of credit to bad, but politically favored, risks.

And here’s where soothing promises that nationalized banks will be privatized “as soon as possible” become meaningless.  It may never be possible, as bad loans multiply on nationalized banks’ balance sheets.  Nationalized banks will become the financial equivalent of roach motels–we can get in, but we can’t get out.  

My verdict: nationalizing is Really REALLY Bad.  

As to guarantees: an improvement, but not much of one.  Government guarantees on (some) of the assets on bank balance sheets is equivalent to the Feds writing a massive put.  There are many details to be worked out: What is the maturity of the put?  Who has the right to “put” the assets to the guarantor, and under what conditions?  Is the put on the entire portfolio, or are there multiple puts on multiple portions of each bank’s portfolio?

I’ll put those non-trivial, but secondary, issues aside to focus on other more problematic ones.  Most notably, what is (are) the “strike price” of the put(s)?, and what consideration will the guarantor (the government) receive in return?  

If the guarantee on is something approaching the par (or even book) value of toxic assets on bank balance sheets, the “put” is deep in the money.  That is, it is likely that the market value of the assets is far below the guaranteed level, and therefore it is almost inevitable at some point the bad assets will be “put” to the government, in exchange for the government writing a check for an amount far exceeding the value of the assets.  

If the guarantee is set at a level approximating the current market value of the assets, however imprecisely determined, the option will be at-the-money.  Short options create a short volatility exposure to the option writer, and this exposure is greatest when the option is approximately at the money.  Therefore, making the guarantee approximately at-the-money creates a substantial volatility exposure to taxpayers.  If volatility goes up, the value of the guarantee goes up as well–costing taxpayers money.

If the government receives no consideration for the guarantee, it is a huge subsidy to the financial institution receiving it.  Puts are valuable.  A government guarantee extended gratis to the financial institution is therefore a subsidy to the bank’s shareholders and bondholders.  This value can come from the “moneyness” of the option or the “time value.”  If the guarantee is deep in the money (i.e., the guarantee is at a level far above the market value of the assets), the intrinsic value of the option is large; this intrinsic value will flow to the bank’s current shareholders and creditors.  Even if the strike price of the option the government provides is approximately equal to the market value of the guaranteed assets, the time value is large.  With an at the money option, the guaranteed bank has a “heads I win, tails I don’t lose” position.  If the assets tank further, the government bears the cost as guarantor; if the assets rise in value, the bank can pocket the gain.  This one sided exposure is also valuable, and this value flows to the bank’s current shareholders and lenders.  This value is highly sensitive, moreover, to the volatility of the underlying assets.  

Recognizing this, in recent interventions the government has received consideration in exchange for the guarantee.  For instance, it has received stock.

Receiving the upside on the guaranteed assets is equivalent to buying them outright.  Receiving the upside is equivalent to having a call on these assets.  Writing a put and buying a call is equivalent to a forward contract–a commitment–to buy the assets at the guaranteed value.

Receiving stock gives the government an upside not just on the covered assets, but on the bank’s other assets as well.  But it also gives the government–that is, the taxpayers–an exposure to the downside on the non-guaranteed assets.  

The idea behind the guarantee, presumably, is to shift the uncertainty about the value of toxic assets to the government.  This, in theory, should make it easier for banks to recapitalize; if the government bears the losses on the toxic assets, private investors in the banks need not fear nasty surprises if the value of these assets turns out to be lower than anticipated.  

I don’t like the guarantee approach for a variety of reasons.  

First, there is a high likelihood that it will serve as a stealth bailout of the existing stock and bondholders of the troubled banks.  That is, the value of the put will far exceed the value of the consideration received.  This will represent a transfer from the taxpayers, to the current shareholders and bondholders.  It may facilitate the attraction of new private capital, but there are ways to do this without bailing out the incumbent suppliers of capital.  

The fact that this approach can facilitate a “stealthy” bailout probably explains its political attractiveness. When announcing the guarantee, politicians can say “we didn’t bail out the banks and their shareholders because we didn’t write a check.”  The honest statement would be “we didn’t write a check today, but we promise to write one sometime in the future.”  If the value of the put is greater than the value of the consideration, shareholders and bondholders are happy, because they receive a windfall.  Put differently, the more popular a guarantee is, the more skeptical I become.  It will be most popular when it (a) provides a large stealth subsidy, and (b) allows politicians to escape accountability for the subsidy.   So, if politicians and bankers unite in enthusiasm for guarantees, be afraid.  Be very afraid.  

A second problem with the guarantee is that I am skeptical that it will markedly improve the ability of banks to recapitalize.   Potential investors will reasonably fear that any commitments that the government makes today are not credible.  As guarantor, and as part owner, the government will potentially find it very tempting to expropriate new investors sometime in the future, perhaps by reneging on the guarantee in whole or in part, or imposing additional conditions or restrictions, or by using the threat of such actions to force the banks to engage in politicized (but uneconomic) loans.  Just as a guarantee can be a stealth bailout, it can be a stealth quasi-nationalization.  The government as guarantor and owner will have a major influence on a bank’s operations going forward, and can use that to harm the interests of new investors.  This means that although the guarantee may reduce one kind of risk–the risk of fluctuations in the value of toxic assets–it adds another–political/expropriation risk.  

For instance, let’s say that performing on the guarantee turns out to be very expensive, because the strike price on the assets is far above their eventual value (either because that value is over-estimated today, or because their value falls further.)  The political pressure to scale back on the guarantee, or change its terms, will be immense.  You can just envision the political posturing.   New investors in banks would lose if this were to happen.  Foreseeing this possibility, they may be very reluctant to invest today.

A potential investor in an institution guaranteed and partially owned by the government would not be likely to take the promise that “we’re from the government and we’re here to help you” to the bank, as it were, and invest in it.  

Given all this, in my view the guarantee approach could both subsidize existing shareholders and scare off new shareholders.  As a result, it is highly likely to be just an extremely expensive waystation on the road to nationalization.

So, color that one really bad.

Another proposal on offer is the bad bank.  Strip out the bad assets, and put them in a separate institution.  Again, many devilish details.  

Most bad bank proposals envision that the government would pay for the bad assets.  This raises all of the difficulties associated with determining the price.  Too high prices, and there is a transfer of wealth from taxpayers to existing shareholders and bondholders.  Too low prices, and the reverse occurs, and the already weak capital position of the banks is eroded further.

My twist on this is to adapt something from my Humpty-Dumpty proposals.  Put all the bad assets in a separate corporation, and rather than giving the contributing banks cash, give them stock in the new company.

The valuation issue still exists–since the assets are heterogeneous and hard to value, it will be impossible to make sure that every bank receives stock in the new company that is worth exactly the same amount as the assets it gives it.  Here, however, the valuation errors are shared in the banking system.  Some banks win, some banks lose, but there is no possibility of systematic transfers (one way or another) between taxpayers and the banks’ owners and lenders.  

Moreover, as I argued in my Humpty Dumpty proposal, the equity in the new company would be marketable, and almost certainly more valuable than the sum of the current values of the individual assets.  This is true partially because some of the toxic assets would be various CDOs and other claims with various offsetting optionalities that make them very hard to value.  These offsetting options positions would be eliminated, leaving the simpler, easier to value (not easy–easier) assets underlying the CDOs.  The equity would be more liquid than the underlying assets, and liquidity can enhance value.  Thus, it is highly likely that this mechanism alone will mitigate the capital problems that banks face because the equity that they receive will typically be higher than the value of the assets that they give up.  

Perhaps most importantly, the equity in the bad asset repository could trade in the market and produce a transparent, relatively informative price.  Exchanging a dog’s breakfast of assets that cannot be valued for a homogeneous claim (equity in the bad asset repository) with a transparent price makes it much easier to value the banks, and to determine which are insolvent and which are not.  This will facilitate the resolution of the problems in the banking industry by the FDIC, for instance.  It will also make it easier for potential new investors to appraise their value, and thus they will be more willing to invest, and to pay more, for new equity stakes in the banks.  

So, my verdict is: Bad bank without the issuance of publicly tradeable equity in the firm holding the bad assets–not a good idea because of the daunting valuation issues.  Bad bank with issuance of such equity–a definite improvement over either nationalization or a guarantee.  

But, it is possible to go further, and do the full Humpty Dumpty.  The proposal I’ve just sketched out gets the bad assets off US bank balance sheets, but the bad assets are also held by hedge funds, foreign banks, and others.  If the bad bank option is limited to US banks, the ability to undo complex structures such as CDOs, CDOs squared, etc., will be incomplete.  Since a major benefit of hiving off the bad assets is the ability to compress hard to value derivatives with complex embedded optionality back into less complex, more easily valued underlying securities, this limitation to US banks will also limit the scope of these compression benefits.  

As I argued in the original Humpty Dumpty posts, obtaining the maximum benefit would require putting all of the toxic assets, including all of the outstanding CDOs, etc., in the single bad bank.  This would require participation by hedge funds, foreign banks, and other investors.  

You might think that it would also be legally impossible–how can you make a foreign bank exchange its holdings of CDOs on US subprime mortgages for equity in the bad bank?  I think that there’s a (draconian) legal way around this.  To the extent that the cash flows on the bad assets originate in this country, the US could require that the cash flows from all of the assets underlying the toxic assets that would go into the bad bank be directed to that bank, and that the bad bank would retain all such cash flows, but would also exchange equity to those presenting bona fide claims on those cash flows.  In essence, by holding the cash flows hostage, the government/bad bank would provide a very strong incentive for anyone, hedge fund, foreign bank, whoever, to exchange their claims on these cash flows for equity in the bad bank.  Any valuation errors in determining how many shares would be received for each asset would be shared within the world financial markets, would not lead to a direct transfer between financial institutions and taxpayers (in either direction), and could be structured so as not to favor one set of asset owners over another.    

Draconian, yes.  But nationalizing a set of bad assets (or the cash flows on these assets) is less draconian than nationalizing the banks in their entirety or extending guarantees.  The government would also not obtain an ownership stake in financial institutions, or such a huge financial stake that it could more extensively politicize the financial markets. Nationalization and guarantees also would not enhance price discovery and liquidity–Humpty Dumpty would.  

It’s likely that even if this approach were implemented, some banks would be found to be insolvent.  But that’s a feature, not a bug.  First, by improving the transparency of bank balance sheets through the exchange of untraded and untradeable assets for the tradeable equity in the bad bank, Humpty Dumpty would help identify which banks are insolvent, and by how much.  This will reduce the potential for type I and type II regulatory errors (failing to intervene when the bank is insolvent, intervening when it is not).  Also, if (as is almost certainly the case) the traded equity in the bad bank is worth more than the sum of the value of the assets if they remain on the bank balance sheets, on average bank capitalization will rise.

To sum up a very long post: Some variant on the bad bank is, in my view, the best way to go.  People have soured on the bad bank idea because of the belief that it requires exchanging the bad assets for cash, thereby raising difficult valuation issues that will inevitably transfer wealth between taxpayers and bank stock and bond holders.  This problem can be avoided if the bad assets are paid for with equity in the bad bank.  Valuation problems will occur, but the transfers will be contained within the group of firms contributing the bad assets.  Moreover, exchanging heterogenous, hard to value assets for a homogeneous claim on a super-portfolio of these assets will produce a variety of benefits.   This approach will be more efficacious, the more complete the participation in the program.  Achieving more complete participation will require more draconian, and hence politically difficult, measures.

Nationalization and guarantees present far more acute problems, in my view.  Both will politicize the process of capital allocation.  That will destroy value, and transfer wealth.  Guarantees can serve as subsidies.  Moreover, I doubt that any system of guarantees can be made credible against political reneging or renegotiation in the future.  In anticipation of this, private investors will be reluctant to recapitalize guaranteed banks, perceiving them to be nationalized banks in waiting.

So, create a Big Bad Bank by exchanging bad assets for its stock.  A Big Bad Private Bank.  The bigger, the better.  Clean out the Augean Stables.  Create a market for the unmarketable, and minimize the government’s role in the operation of these financial institutions and the financial markets.  The greater and more enduring the direct government involvement in this process, the worse the outcome, in terms of politicized capital allocation and massive transfers of wealth.   Once all the King’s horses and all the King’s men put Humpty Dumpty back together, they should get the hell out of the way and let him do his thing.  It ain’t perfect, but it’s a damn sight better than the other appalling alternatives on offer.

Print Friendly, PDF & Email


  1. Nice ideas. Is there a chance of (3) or (4) being put forth as a serious alternative by the House and Senate Republicans?
    But how will this help consumer spending return? I feel that the de toxified banks would still be very cautious lending to the public, many of whom are saddled with big debts and poor credit. The US has now morphed into a predominantly service oriented economy. And I dont see a clear path to services demand picking up. Can the SWP spot where the next asset bubble would be? What would be the next job creation engine, apart from the “3M” promised by Obama…….

    Comment by Surya — February 7, 2009 @ 8:21 pm

  2. […] on guarantees extended to still-private banks.  (Is death an option?  Both alternatives are bad, as I’ve written before, and as Willem Buiter concurs.)  A failure to come to a decision as to the fundamental strategic […]

    Pingback by Streetwise Professor » Nero-bama? — February 28, 2009 @ 10:14 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress