Streetwise Professor

October 8, 2008

The Humpty Dumpty Option

Filed under: Derivatives,Economics,Politics — The Professor @ 9:45 pm

In an earlier post I mused–in an abbreviated, bloggish sort of way–that reassembling subprime and Alt-A mortgage backed securities by bundling CDO tranches would improve liquidity in this market. My confidence in this intuition was strengthened by reading Gary Gorton’s much more thorough and penetrating analysis of the informational effects of the complexity of CDO structures, and the implications of these effects for the liquidity of the CDO market. (H/T to Marginal Revolution for the Gorton link.)

That has encouraged me to think bigger. What follows is basically a thought experiment, with a critique of the difficulties of making it a reality.

How big? What if . . . ALL assets backed by subprime mortgages were contributed to a single firm? (Still noodling over how synthetic CDOs would fit into this scheme.) The owners of these assets would receive shares in this corporation, which I will refer to as The Firm.

This would offer several advantages.

Most importantly, The Firm would effectively hold basic subprime mortgages. Bundling all the contingent claims on all the subprime mortgages would eliminate all the offsetting embedded options positions, leaving only the underlying mortgages.

Now these would not be easy to value, but they would be a lot easier to value than CDOs and CDOs squared, etc., because these products (as Gorton points out) embed complex option structures. The underlying mortgages have optionality (pre-payment, and especially default). Every slicing and dicing creates additional optionality–CDOs are compound options, CDOs squared are compound options on compound options, and so on. And every additional option increases complexity, destroys information, and makes valuation inherently more difficult.

The reduction in complexity–and the reduction of what Gorton refers to as the “informational losses” that result from the creation of complex chains of contingent claims on the underlying mortgages–would make the shares in The Firm substantially more liquid than the CDOs. Moreover, it is well known that lemons problems are less severe for portfolios than for individual assets that make up those portfolios. As a result, claims on portfolios (e.g., S&P 500 futures) are usually far more liquid than claims on the individual components of the portfolios. Thus, shares of The Firm would be a lot–a lot–more liquid than the CDOs contributed to it, and even more liquid than the underlying mortgages.

If, as is almost certainly the case, illiquidity depresses prices, the shares of The Firm will have a greater market value than the current market values of the CDOs, etc., that are contributed to the firm. This increase in market value would reduce the insolvency issues and questions currently dogging banks and other intermediaries.

This effect is essentially what the Paulson plan is trying to achieve through the creation of a market for the individual CDOs. Given the heterogeneity of the individual CDOs, however, in my view it will be much harder to create a liquid market in these individual components than it will be to create a liquid market for less complex bundles of them. (A sort of alchemy is at work here–putting together multiple complex things in the right way results creates a less complex thing. This is the KISS–keep it simple, stupid–approach to financial engineering.)

Financial institutions that contribute assets to The Firm will replace these assets on their balance sheet with shares in The Firm. This has two important effects. First, since due to their greater liquidity the shares are likely to have a higher market value than the “fair value” of the assets the banks give up, this innovation will increase the capitalization of the banks. Moreover, it will homogenize the balance sheets of financial institutions because the shares are identical and fungible. This will make it easier for banks to evaluate one another. This, in turn, may help rejuvenate the interbank lending and repo markets, ,which are malfunctioning primarily because banks do not know with any precision who is solvent and who isn’t due to the opacity of balance sheets that contain subprime exposures. (The shares can also potentially serve as repo collateral in both private transactions and transactions with central banks.)

Another potential benefit is that The Firm would internalize the benefits of restructuring the underlying mortgages and optimizing foreclosures. This could redound to the benefit of both (a) the shareholders (because the values of the CDOs are reduced because securitization means that those best able to negotiate restructuring of the underlying mortgage debts do not reap the benefits), and (b) the mortgage borrowers who enter into mutually advantageous restructurings with The Firm.

In a nutshell, putting the mortgage Humpty Dumpty back together to create a less complex, more homogeneous financial claim–equity in The Firm–would improve market liquidity and facilitate more efficient restructuring of mortgage debts, thereby easing the solvency and lemons problems that bedevil banks and the money markets.

Sounds wonderful, but the Humpty Dumpty solution as outlined above is akin to Steve Martin’s plan to make a million dollars without paying taxes: First, make a million dollars. I’ve cut straight to the end by assuming the magical creation of The Firm, just as Steve Martin assumed away the hard part of making $1 million without paying taxes.

The problem, of course, is getting there. There are myriad obstacles practical and political.

Most importantly, there is a major scale economy here. The benefits (per dollar of principal or asset value) are increasing in the fraction of subprime mortgage contingent claims that The Firm owns. If it owns all these contingent claims, it will end up with the least complex possible portfolio (simplest, not smallest)–the underlying mortgages, without all of the compound optionality, compound compound optionality, etc., embedded in the CDO structures.

Even though as a result of this scale economy it would be best to put everything in The Firm, it is hard to imagine that this outcome would occur voluntarily, without considerable coercion. Even though this approach makes the size of the pie bigger, and arguably much bigger, it will inevitably have distributive impacts which will induce collective action problems. Moreover, a voluntary approach would likely create adverse selection problems. I’ll consider each point in turn.

Contributors of assets would receive shares in the corporation. Given that the assets contributed are inherently heterogeneous, this raises the issue of how to determine how many shares would be distributed for each asset.

Now, mispricing (which is almost a metaphysical concept in this context) would lead to redistributions of wealth among the contributors of assets. (This is an important difference with the Paulson Plan, in which mispricing leads to redistributions of wealth between current asset owners and taxpayers. This is a potential political selling point for this alternative.) If The Firm gets all assets, this is merely shifting wealth around among asset owners, and from behind the veil of ignorance (which is almost completely opaque here), ex ante any individual asset owner is as likely to incur a distributive windfall as incur a distributive loss.

However, just as everybody thinks–or will assert–that their child is better looking than average, even if some have faces that could stop a clock, asset owners will think–or assert–that their assets are more valuable than average. They will fight tooth and nail to get the highest relative valuation of what they have. This will generate massive transactions costs.

The amount of money that could be redistributed could be immense. Voluntary negotiations among asset owners to form The Firm, and determine the pricing (in shares) of each asset to be contributed, would generate tremendous inefficiencies as the asset owners would negotiate to receive windfalls and impose the distributive losses on others. As Gary Libecap and Steve Wiggins showed in their analysis of unitization of oil fields, even when an agreement leads to substantial increases in aggregate wealth, rent seeking to receive redistributive windfalls can preclude the consummation of the wealth enhancing Coasean bargain. And the magnitude and complexity of unitization negotiations pale in comparison to those involved with the creation of The Firm.

Moreover, if contribution of assets to The Firm is voluntary, adverse selection problems are likely to be acute, resulting in incomplete participation. If The Firm sets a schedule of prices, X shares for AAA (at issuance) CDOs, Y<X shares for AA (at issuance) CDOs, etc., based on a the best available information about the average value of these various securities, asset owners with private information will tend to withhold the assets with higher than average values, and contribute the lemons. Thus, (a) The Firm is unlikely to acquire all of the assets voluntarily, and (b) will be stuck with the lowest quality ones.

There is likely be a critical mass effect here. Due to the scale economy discussed above, the value of the greater liquidity of The Firm’s shares will encourage the owners of assets that are undervalued relative to other assets by the share distribution formula to participate anyways because the alternative is to hold onto the assets or sell them individually in an illiquid market. (I.E., even if I know my asset is receiving a smaller share of The Firm than another asset of the same quality, I may still contribute it in exchange for shares since those shares are worth more than I could get by selling the asset as a standalone.) Since the liquidity benefit should be increasing in the fraction of troubled assets The Firm acquires, once The Firm gets big enough the adverse selection problem should disappear. The challenge, then, is to make The Firm big enough to start this virtuous cycle. This is not a self-starting phenomenon. There is something of a Catch-22; everyone will want to participate if The Firm is big, but how can it get big to start with? Asset owners may re-enact the old Alphonse and Gaston vaudeville routine: After you, Alphonse. No! After you, Gaston. No, I INSIST, Alphonse. . .

It should be noted that the valuation challenges in setting the prices in shares of The Firm for different assets should be less daunting than attempting to determine the prices of these assets in dollars (or equivalently, other goods and services). The values of subprime contingent claims are affected by a set of common factors, most importantly, real estate prices and interest rates. Errors in estimating one of these factors will tend to affect the values of different contingent claims in similar ways. For instance, if in determining values, real estate prices are assume to be lower than they are in actuality, the prices of all mortgage contingent claims will be underestimated. As a result, the errors in estimation of the values of the various contingent claims relative to one another (which is what is relevant in determining the allocation of shares) will be smaller in magnitude to errors in the estimation of the values of these claims to other goods and services.

Of course, due to the optionality of the CDOs, a Z percent error in the real estate prices assumed to determine the number of shares to be given for each asset will not lead to a Z percent error in each of the assets. So there will be errors in determining the relative prices. But due to the fact that all the assets at issue are affected by a set of common factors, it should be possible to set relative prices (i.e., the number of shares to be distributed for each asset) more accurately than it would be to set the prices of each asset in dollars–which is what the Paulson Plan attempts to do.

As a result of the collective action/rent seeking and adverse selection problems, it is highly likely that considerable coercion would be required to create The Firm. That is, the government would almost have to employ the Don Corleone option. (This could involve punitive taxation on the owners that do not contribute their assets, or abrogation of the CDO contracts that are not contributed.) It would truly take all the King’s Horsemen and All the King’s Men, exercising vast powers, to put Humpty Dumpty together again.

Given that the asset owners are numerous and include citizens of myriad nations (not just the US), the political and diplomatic outcry that would result from such measures would be deafening. More importantly, the exercise of such overwhelming coercive power is truly frightening, and would set troubling precedents. As someone who values free contract, and the sanctity of private contracts, and abhors the exercise of government power, I could not in good conscience countenance such a solution. But . . . given that other alternative actions that will be advanced if the Paulson Plan fails will also inevitably involve massive coercion, at the end of the day we may be confronted with choosing between the least coercive of two evils. Or put differently, we may have to determine which alternative offers the greatest value, once the costs of coercion are taken into account. If we come to that sticking point, The Firm, with CEO Corleone, may look good by comparison.

I should also note that one of the potential theoretical benefits of The Firm discussed above may prove problematic. Specifically, although in theory The Firm should be able to negotiate more effectively and efficiently to restructure underwater mortgages (because of its better incentives and information), in practice it would face tremendous political pressures that would impede its ability to achieve this outcome. The owners of The Firm–those contributing assets, or those to whom the asset contributors sell their shares–want to maximize the value of the underlying mortgages. But, the borrowers want to minimize the value of those mortgages, and they will be a powerful political constituency. Indeed, one can imagine that a company that could only be created by a massive exercise of Federal power would be the creature of that power, and would be used as an instrument to redistribute wealth from the owners of mortgages to the mortgage buyers. This would, of course, offset in whole or in part the salutary effects of the creation of The Firm on the balance sheets of financial institutions, and the follow on benefits to the money and credit markets, and the real economy. At the very least, given the trillions of dollars at stake, The Firm would be at the center of the biggest stakes political game in history. Public choice theory tells us that efficiency is not the most likely outcome of such a contest.

So, if we start at the end–the creation of The Firm, the proverbial Bad Bank on massive doses of steroids–this approach of attempting to reverse the effects of years of Frankensteinian financial engineering has some real attractions. It would not be a complete fix–the underlying mortgages are experiencing huge defaults, and as a result the value of The Firm’s shares will still be less than the principal amounts of the underlying mortgages, meaning that banks and others will still have holes in their balance sheets, albeit smaller ones. Even those smaller holes may be big enough to make some institutions insolvent. But we’ll have a better idea of which institutions are insolvent (which will help direct government funds to where they are most needed, and which should help mitigate the uncertainty that is effectively shutting down the corporate paper and interbank lending markets), and we’ll have smaller holes to cover either through recapitalizations or government performance of deposit insurance obligations. This approach avoids many of the problems with the Paulson Plan, which relies on buying the individual assets.

The problem is getting there. It will require coercion–maybe not mobs with pitchforks and torches besieging Dr. Frankenstein The Financial Engineer’s castle, but the government employing its powers of taxation and the ability to abrogate private contracts. Lots of coercion. That is a cost, indeed. But as Ronald Coase never tires of pointing out, in the real world we have to compare real alternatives to real alternatives, rather than to Nirvana. Maybe, as Churchill said of democracy, the creation of The Firm is the worst possible alternative, except for all others that have been considered or tried from time to time.

This concept is still a protean one. I think it has some merits that deserve consideration, but I am sure that there are difficulties that I have not yet discerned. Although it is speculative and theoretical, it is no less speculative and theoretical than the bailout bill just enacted into law. Moreover, I think it has some advantages over that bill. If, as I expect, the bailout plan does not work as hoped (and it is a triumph of hope over experience), the Humpty Dumpty Option deserves a look.

At the very least, I think that this thought exercise helps focus attention on a fact that has not received anything near the attention it deserves: Namely, that any attempt to deal with the subprime morass (or, as my former Naval Academy roommate, now a workout specialist for small banks, puts it, la cosa subprimo) should go to the root of the problem. As Gorton persuasively shows, that root is complexity induced by financial engineering that created dense layers of options on top of options. Reversing that complexity–putting Humpty back together, or, to mix metaphors, putting the toothpaste back in the tube–would go a long way to mitigating the problem. Such a reversal would inevitably involve coercion, so the question will be: is the cost of coercion worth it? As bad as things are, I can’t answer affirmatively today. Tomorrow. . . who knows?

So, for now, consider this a straw man that lays out a particular approach to attempt to strike at the root of the problem. I therefore encourage commentors to (a) point out difficulties, or new complexities that this alternative would create, and (b) propose other ways to attempt to reduce complexity through reverse financial engineering. For if the specifics of what I have outlined here are faulty, I have a strong sense that the fundamental intuition that such reverse engineering is the way out of this mess. If anybody can think of incentive compatible schemes, schemes that are voluntary and will not be scuppered by collective action problems, transactions costs, and adverse selection. . . please share them with the class!

Print Friendly, PDF & Email


  1. Professor,

    This is one of your very best posts—and that’s really saying something!

    You present a compelling case that, if the Panic does not subside soon, something like The Firm will be established using Don Corleone tactics.

    I can’t deny this would pose important legal and constitutional questions but, then again, it’s never been obvious to me what constitutional provisions underlay FDR’s authority to close banks, vitiate the gold clause in privately negotiated bond contracts, and mandate surrender of privately held gold bullion.

    Nevertheless, he was able to order all these things and more. There were some cries of protest at the time but those voices were drowned out by the much larger part of the population who were happy to see the President “acting decisively”. The stock and bond markets rallied strongly in 1933.

    Comment by John McCormack — October 10, 2008 @ 11:46 am

  2. John–Thanks so much for your kind words. I think the idea has a lot of merit. I’ve spent the last couple of days trying to poke holes in it, and find it hard to do from an economic perspective. All of the objections that come to mind are essentially political ones. (Not that those are immaterial, but if the economics are right, it may be worthwhile to pay some political costs.)

    Re your basic point, yes–FDR’s interventions were constitutionally dubious, but were probably essential (though the economics of many of his forays were pretty dubious.) Luigi Zingales made a similar point in an oped (in the India Times, I think. I’ll try to find it and link to it.)

    FDR had the advantage of coming in with a clean slate. Bush is a spent force, and so reviled that it is unlikely that he can do anything. Sadly, I shudder at the prospect of either Obama or McCain grappling with these issues, as their ignorance of economics and finance is frightening.

    Thanks again for the comment, and the compliment.

    The ProfessorComment by The Professor — October 10, 2008 @ 1:15 pm

  3. […] . . I keep coming back to Humpty Dumpty. Go ahead. Call me obsessive. I can handle it. I know it’s radical, but nothing else on offer […]

    Pingback by Streetwise Professor » Died of a Theory — November 12, 2008 @ 7:39 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress