Streetwise Professor

NB: I have slightly augmented this post, with some additional amplification of how the “plan” will provide the least support to the most troubled banks, and generate the least price discovery for the most troubled assets.

The Treasury has released, at long last, additional details of its plan to address toxic assets.   (I refuse to use Treasury’s Orwellian Newspeak “Legacy Assets.”   That sounds like a funeral home advertising slogan.)   I say “additional details” rather than “final plan” quite deliberately.   As I will note below, there are still many parts to-be-colored-in-later.

My overall assessment is very different from the stock market’s initial appraisal.   Mine is negative, the market’s positive.   We’ll see.   I only remark that the market gave its hearty approval to TARP I too, and we know how that worked out.

I’ll divide my analysis into two parts: an evaluation of the big picture, and some comments on the details.

With respect to the big picture, in my view the Treasury approach does not respond to the major problem facing the banks.   It nibbles at the edges, and throws a lot of money at the problem in a very indiscriminate way.   This raises the serious risk that (a) the plan will fail, and (b) when it does, the resources to address the real problem will not be forthcoming for economic and political reasons.

The main problem with the banking system is that many banks, including some of the biggest, are quite likely insolvent.   Maybe, on a consolidated basis, the entire banking system is insolvent.   Addressing this problem requires two basic steps: (1) identify the insolvent institutions, (2) restructure the insolvent institutions, by stripping out bad assets and recapitalizing, both with public funds to cover any deposit insurance obligations, and with private funds.  

The Treasury plan will not do either of these things directly, and are unlikely to accomplish them even in a roundabout, indirect way.   The likely outcome is that some assets will change hands, but that the worst ones, and the ones held by the worst banks won’t.   As a result, considerable uncertainty about what banks are solvent and which are not will remain. Moreover, insolvent and near-insolvent banks will continue to operate, leading to serious incentive problems that can exacerbate the problems down the road.  

In a nutshell, the Treasury plan offers equity and debt financing to private investors who will purchase assets from financial institutions.   An important factor to note is that the current owners of these assets will have full discretion over what assets they sell, and the prices they accept.  

Geithner justifies his plan by saying:  

The Public-Private Investment Program will purchase real-estate related loans from banks and securities from the broader markets. Banks will have the ability to sell pools of loans to dedicated funds, and investors will compete to have the ability to participate in those funds and take advantage of the financing provided by the government.

But banks already have the ability to sell, and private investors already have the ability to buy.   They haven’t done so.   What Treasury adds to the mix is generous funding via equity investments, and more importantly, non-recourse loans.   Its hope is that this funding will lead to more voluntary transactions for toxic assets.

I seriously doubt this will occur to the extent necessary because the Treasury plan seems to overlook entirely the major reason WHY banks with bad assets on their balance sheets have been very reluctant to sell.   Namely, that many of these banks are insolvent, or borderline solvent.   What’s more, they know they are insolvent, or teetering on the edge.   They further know that the absence of active markets for these securities and loans gives them substantial discretion regarding how to value these assets for accounting purposes, and by assigning optimistic valuations they can maintain the illusion of solvency and continue to operate under current management.  

Selling assets at market prices that reflect private investors’ realistic appraisals of value would result in serious writedowns to bank assets, reducing their capital, and raising the likelihood that their insolvency would become readily apparent, leading to government seizure/receivership, and the consequent wiping out of shareholders, some debtholders, and management.  

Bank managers, equity holders, and even some creditors of insolvent institutions realize that there is considerable option value to remaining in operation while insolvent.   Things could turn around, and the bank could get lucky and escape insolvency.   Moreover, its management has an incentive to take additional risk in order to gamble its way out of insolvency.   Management, the shareholders, and even some debtholders realize that the downside of this strategy is small, because the government and some debtholders will take the hit if the gambles turn out bad.   (The government will pay off on deposit insurance, and may be forced to bail out too-big-to-fail firms.)   On the upside, managers have the ability to keep their jobs longer, and if their gambles succeed, may keep them indefinitely.   Similarly, shareholders of insolvent institutions, protected by limited liability, are in a heads I win, tails I don’t lose any more situation.

So, what is likely to happen is that the really bad assets will stay on bank balance sheets. Some of the better assets will be sold.   Zombie banks will continue to exist.   Uncertainty and asymmetric information about the solvency of individual banks will persist, as the banks will still be stuck with large amounts of bad assets.   This will continue to impede interbank lending and the recapitalization of banks.   Moreover, the bad incentive effects of allowing insolvent or borderline solvent banks will continue, leading to substantial dissipation of value.

In other words, this plan is only a slight improvement on the status quo.   It does not strike at the heart of the problem: the facts that (a) there are many banks, including some very large ones, that are insolvent, and (b) that nobody knows exactly which ones are insolvent or by how much.

Additional material begins:

Since the worst-off banks are least likely to participate in the program, they will receive the least amount of cash from the sales of bad assets.  Moreover, since the assets that are currently most misvalued on bank books (which are likely the worst assets) are least likely to be sold under the program, it will contribute least to price discovery for this class of assets.  

Both of these outcomes strike me as perverse.  The least amount of cash flows to the banks that need it most, and the assets that are most imprecisely valued will be the least likely to benefit from improved liquidity and price discovery.  We would like the Treasury (i.e., taxpayer) dollars to go to the banks that need it most, not least.  Especially, as if these banks continue to operate in their zombie state they are likely to wind up on FDIC’s or Treasury’s doorstep somewhere down the line.  Where will the money needed to satisfy government obligations be then?  Well, it might have already flowed to other institutions whose need for it was not as great.  We would like to get better pricing information about the worst of the assets.  We won’t.  This will make it all the more difficult to evaluate the solvency of the banks holding these assets.

In other words, it seems that this plan fights the wrong battle in the wrong war at the wrong time.  Other than that, it’s great.

End additional material.

What is needed here is a kind of Judgment Day for banks.   This is most effectively accomplished through a compulsory stripping of bad assets from banks, either through some variant of the good bank/bad bank approach, or my Humpty Dumpty option.     Once the bad assets are out, it will be easier to identify which banks should live, and which should die.   I favor Humpty Dumpty because the equity claim on the entire pool of bad assets that this approach creates is likely to be more easily valued than the individual bad assets.   It also shares the valuation risks more efficiently among the banking system, and limits taxpayer exposure to valuation errors.  

The Treasury plan defers Judgment Day.    It gives us Groundhog Day instead.  We will continue to repeat, day after day, the experience of living with a banking system of doubtful solvency.

The problem is that every day we wake up to Sonny and Cher, the existing problems of weakened interbank markets and difficulties of banks in attracting capital will fester, and indeed, due to the perverse incentives that zombie bank managements and shareholders face, new problems and risks will appear.  

We’ve seen the effects of delaying Judgment Day before.   The government refused to grasp the nettle when many S&Ls began to run into trouble in the mid-1980s.   It devised many ingenious ways (e.g., regulatory capital, accounting fixes) that allowed insolvent institutions to survive to gamble another day.   (Proposals to permit banks to rely on historical cost accounting today would be similarly stupid.)   Then, due to Garn-St. Germain, Congress gave the S&Ls the ability to take even greater risks, which they did with gusto to gamble their way out of trouble—only to dig their holes deeper, and force the taxpayers to fill them in.   In a few short years a serious but tractable problem metastasized into a hugely expensive one.   (Fortunately, we haven’t seen a repeat of the Garn-St. Germain fiasco this time around, but banks have enough freedom to take on additional risks without any new, “liberating”, “deregulatory” legislation.)

It should also be noted that the Treasury plan in essence gives the worst banks the worst incentives.  The banks with the worst assets, and the banks that are in the worst financial position, have the weakest incentive to take part in the program. Anything that leads to more accurate valuations of their balance sheets increases the odds they have of going from Zombie to Dead.   Managers of zombies figure that that state is preferable to death, and what’s more, that while there is no coming back from death, one can get lucky and escape zombiehood.   Indeed, one can take risks on the government’s dime, which if they pay off, will permit the return to a normal, non-zombie life.    And if they don’t–well, that’s somebody else’s (the taxpayers’) problem.

This scathing assessment obviously raises the question: Why would Treasury do such a thing?   I have no definitive answer to that question, but I would note that banks have no doubt fought Judgment Day tooth and nail.   That is, a political economy perspective suggests that Treasury may well have been captured by the banks. Treasury may also have considerable room to doubt whether Congress would support Judgment Day either, because of a lethal combination of aforementioned bank opposition and likely populist outrage.

So, to sum up the SWP take on the big picture: high likelihood of disaster.   By deferring Judgment Day, Treasury is at best perpetuating, and arguably exacerbating the most important problem confronting the banking system.  

Now a few comments on the details, such as they are.  

In brief, this plan, despite months of preparation, is still not ready for (sub)primetime.  

It is not, in other words, a completed plan.   Many holes still exist.   The document released by the Treasury states that the “Legacy [read Toxic] Loans Program will be subject to notice and comment rulemaking.”   In other words, this is not ready to go.   There will be a time consuming, and perhaps contentious, period of public comment.   With respect to the Legacy [Toxic] Securities Program, the Treasury says: “Haricuts will be determined at a later date [!] . . . .Lending rates, minimum loan sizes, and loan durations have not yet been determined. [!!!!]   These and other terms of the program will be informed by [!] discussions with market participants. [What have they been doing for the past weeks/months?]”  

If haircuts (i.e., the amount of leverage), loan size, rates, and maturities have not been determined, pray tell what has?  

At best, the plan warrants a grade of “I” for incomplete.   As a result, it will take months to get off the ground.

Moreover, the program does not touch one of the most important categories of toxic assets: CDOs, CDO squared, etc.   The Legacy [Toxic] Securities Program is limited to residential mortgage backed securities, commercial mortgage backed, and asset backed securities (ABS).   CDOs, etc., are collateralized by ABS, so they would not be eligible.   Thus, a good slug of the bad assets in the system will inevitably stay on bank balance sheets.  

In sum, in my view, based on both the big picture and the details, this plan is a failure on its own very modest terms: “to restart markets for troubled assets, begin the process of repairing balance sheets, and eventually lead to increased lending.”   The worst banks and the worst troubled assets will avoid the program like the plague.  

Sad to say, addressing the fundamental problem of the banking system—insolvency and near insolvency of major banks—will require compulsion of some sort.   This compulsion could come through existing legal processes, notably bankruptcy.   It could come through new forms of government compulsion, such as Humpty Dumpty or the compulsory formation of bad banks.

I am not, as you would know if you read this blog regularly, in favor of government compulsion.   But, in many respects the problems we face are the direct consequence of government policies, notably deposit insurance and the implicit guarantee to large financial institutions through the too-big-to-fail doctrine.   Given these conditions, purely voluntary approaches create serious incentive problems that only exacerbate the perverse effects of these policies.     Given the guarantees (explicit and implicit) baked into the cake, market participants (namely, bank managers, shareholders, and debtholders) will not hold Judgment Day on their own.   It must be forced on them; Groundhog Day works just fine for them, as there is at least a chance that some day they will wake up renewed.   The Treasury plan fails to attack the solvency question, and hence will only defer the inevitable result.   In the interim, unfortunately, things are unlikely to get much better, and may get considerably worse.