Titolo

The ''new'' Geithner plan

17 commenti (espandi tutti)

Hypothetically, yes. If I know something is worth x, I am always be willing to sell it at any price larger than x. If I do not know what something is worth, for example I think it may be worth anything between x and y (y>x) than I am always willing to sell it at prices larger than y, but at prices between x and y my behavior is "unpredictable' since I do not know how to evaluate selling versus not selling. What makes things tricky for the observer who does not know what I think is that I may sometimes be willing to sell it and sometimes not.

Notice that the reasoning above follows what we do, and does not assume ambiguity aversion which is what all other models of ambiguity do (including α-MEU). If you were ambiguity averse, than there would be no indecision region since your preference are complete.

Does this help?

It does but it confirms that, by looking at what banks do I cannot tell if the reason is ambiguity or, purely, waiting for Geithner lifesaving jacket under normal risk aversion ... right?

In other words: I know it is worth 30, which is what the market is offering, but given that Geithner will be soon offering 60, I wait. That's exactly what the Japanese banks did ten-fifteen years ago.

I am not sure I understand your point well. Let me rephrase (what I believe is) the question.

There is a lot of people out there willing to buy toxic assets for 30 cents on a dollar. I observe that the owners of the asset (it is actually agents of the owners and this matters a big deal, but let's forget about this) claim that they are really worth 60, so they refuse to sell. Now the question is: How do I tell whether the refusal to sell is due to the fact that the onwers are standard EU maximizers with complete preferences or instead are just plain confused and they have no way to tell what the assets are worth? Is there any observation that would help us in saying it is one or the other? For what I understand from your argument, it looks like that the following experiment might work. Suppose that I ask repeatedly, in a number of independent trials, the bank managers to sell the toxic assets at 30. If they have superior info then they will always say no. If they are confused then sometimes they will say yes. Is that correct?

A more fundamental question is: even if we think that the bank managers are confused and don't know the true value of the assets, shouldn't we push for transparency? Shouldn't it be made clear that the unwillingness to sell at 30 is not due to some superior knowledge but only to inability to find the correct value? The mechanism that I proposed does exactly that.

Let me admit my ignorance upfront and ask: where is the source of imperfect and/or asymmetric information exactly?

Is it a matter of what is inside any given security? This sounds strange to me, as I would guess that those contracts were written in a very detailed way, listing who has to pay whom in each one of many fairly well-specified contingencies (that's what they paid lawyers for!). Anyway, there shouldn't be too much asymmetric information here: why (after a possibly more careful than usual due diligence process) would a potential buyer of those securities know much less than its current holder?

Or (non-exclusive) is it a matter of how future macroeconomic variables will affect the payoffs of those securities? This seems to me the more likely case, but then adverse selection stories would not have much bite. A priori, potential buyers would not have better information than potential sellers. They may have different beliefs (at least if you don't take agree-to-disagree impossibility theorems too seriously) and you can have no-trade results because of ambiguity aversion (as in Luca and Chris's paper) or because of hopes in a "Geithner put" (er no, wait, isn't that a future macroeconomic variable, too?).(*) But the incentive scheme that Sandro proposed would not be of much help in this case, would it?

(*) Edit: another explanation would be that the banks are toast at the expected values of their books, so each banker adopts a strategy of "hold and pray". I don't see how Sandro's plan helps here either.

To evaluate a CDO with copulae you need:

1) the cumulative marginal distribution of default of the members of the pool;

2) the correlation among the members.

In theory you should know 1) but in practice I do not think you really know because they misclassified people on purpose to give them a mortgage....

The correlation of point 2) depends on both the future macrovariables and the classification of the people pooled into CDO.

So I think they really do not know.

 

where is the source of imperfect and/or asymmetric information exactly?

Rody Manuelli and I were just discussing this issue ten minutes ago. It is not an asymmetry of principle, but of fact. Had the Fed/Treasury started to do its job many years ago, or even two years ago, the problem would be much more attenuated as they would have collected all the micro information that come with each MBS and built up a data base and a set of pricing models comparable, hopefully better, to those that the banks have. But they did not.

So, de facto, right know they know the following about the content and potential market value of these securities: shit. Bird shit, not even cow shit. Which is why they are desperately hiring people from the private banking sector (last one: CITI's Chief Economist is now at the Treasury)! In fact, not only they do not have the data sets, but they do not even have the in house expertise: people in the research departments of the various Feds are good at many things, but not at pricing securities! As for the Treasury, I do not know the internal composition, but I doubt they have the human capital. In other words, the info is de facto asymmetric between Treasury and the private sector. Instead it is NOT asymmetric between two private parties that have been in this business for a while. Which is way the stuff in question does not trade, i.e. the reason you put under (*), as we have been arguing for a while.

This observation prompts two considerations:

- Do you want to know how the "capturing" takes place? It takes a place this way, with Fed and Treasury being filled by Wall Street executives in "temporary mission behind enemy's lines".

- If a highly paid private bank executives is willing to go work for the Treasury, how much do you think he is giving up in salary and bonuses? Which implies: EITHER he is, as above, in a secret mission OR the market for these guys is, finally, drying up. Hence, the retention bonuses are even more unjustified as they are being retained only from moving into ... well earned unemployment.

P.S. Sandro's method should help with the asymmetric info described above. It should also help with the "hold and pray" (or "japanese") strategy as one should then announce that the Fed will start to do its supervisory job and sample assets on the book of banks to compare them with the market values of the assets that are now trading. And, if they are on the books for values statistically unjustifiable given observed trading, well ... you know what I mean. But the Fed stopped doing its job, i.e. the one just described, many years ago and has no intention to resume it at all.

P.P.S. Its Chairman, though, has every intention of continuing treating us like idiots. Embarassing!

 

Do you want to know how the "capturing" takes place? It takes a place this way, with Fed and Treasury being filled by Wall Street executives in "temporary mission behind enemy's lines".

If a highly paid private bank executives is willing to go work for the Treasury, how much do you think he is giving up in salary and bonuses? Which implies: EITHER he is, as above, in a secret mission OR the market for these guys is, finally, drying up. Hence, the retention bonuses are even more unjustified as they are being retained only from moving into ... well earned unemployment.

Maybe a few additional words on the once praised US lobbying system could help. US companies have always been keen to hire (or re-hire) people coming from FED/Government/Public Agencies. IMHO what has happened in the last years does not support the view of those who used to approve the apparent transparency of effectivness (for whom ?) of such system...and is not only a matter of regulator capture.

Sandro's method should help with the asymmetric info described above.

Yes, Sandro's mechanism would help against the asymmetry of information between the Treasury and the banks if the former wants to buy assets from the latter. But if the problem is that the banks do not want to sell at the expected values of the assets (which would be more or less revealed by an incentive-compatible mechanism), then there would (and should) be no such sale.

It should also help with the "hold and pray" (or "japanese") strategy as one should then announce that the Fed will start to do its supervisory job and sample assets on the book of banks to compare them with the market values of the assets that are now trading. And, if they are on the books for values statistically unjustifiable given observed trading, well ... you know what I mean.

Ah, that's another story. Yes, the Fed should do that ... and I think I know what you mean. But wouldn't that take a lot of time? Meanwhile, instead of (or in addition to) arranging a toxic misunderstood asset purchase program, wouldn't it be quicker and cheaper to force distressed banks into a Bulow-style bankruptcy procedure? (yes, I remember you scolded me already for bringing up the Bulow proposal, but since Sandro also seems to like it ... I try again!)

to force distressed banks into a Bulow-style bankruptcy procedure? (yes, I remember you scolded me already for bringing up the Bulow proposal, but since Sandro also seems to like it ... I try again!)

I think I should concede on that one. My mistake: by reading the presentation in the press and in the Hall&Woodward's blog I had the, incorrect, impression, that it was another crazy proposal aiming at making miracles (i.e. getting rid of the bad assets by some magic trick). Instead, they have essentially in mind taking over the banks and suggest purely a procedure to keep the bank active and functioning while going through the bankrupcy procedure.

In any case, it seems that unfortunately we will not do that, no? Contracts are sacred, if signed by bank's employees, less sacred if signed by anyone else ...

Scusate l'italiano, mi sembra che Sorana abbia studiato bene il piano "good bank-bad bank". Sto leggendo l'articolo da un paio di giorni e l'unica cosa che credo di avere capito è questa

Much thinking about bank policy takes an old-fashioned point of view by assuming that a bank finances all of its assets through deposits. The good-bank/bad-bank separation has no advantage in that traditional setting. But for a bank that is mostly financed by non-deposit borrowing, moving the non-deposit liabilities to the bad bank has an advantage in dealing with insolvency.

Ma è vero o è wishful thinking? Voglio dire c'è qualche probabilità che il piano così congegnato funzioni veramente?

Temo di non avere affatto `studiato bene il piano "good bank-bad bank"'.

Ho solo letto gli articoli di Hall-Woodward e Bulow-Klemperer e l'idea (carve out a good bank from the existing bank and give the equity of the good bank to the bad bank) mi sembra buona: ottieni una banca "buona" e funzionante in tempi brevi senza preoccuparsi di valutare securities che sono difficili da valutare e senza trasferire ricchezza tra nessuno (poi ci si puo' occupare della banca "cattiva" con calma).

Pero' non sono un esperto di finanza e non so se, per esempio, le banche in crisi hanno uno stato patrimoniale buono abbastanza da tirar fuori una "good bank" (come nell'esempio di Citibank citato negli articoli di cui sopra) o se sia facile determinare quali sono le "good securities" da trasferire alla "good bank".

E' per questo che chiedevo l'opinione degli esperti di nFA.

Mi pare che il piano adesso piaccia sia a Michele che a Sandro, ma purtroppo non a Tim e Barack.

Grazie,

non so se, per esempio, le banche in crisi hanno uno stato patrimoniale buono abbastanza da tirar fuori una "good bank" (come nell'esempio di Citibank citato negli articoli di cui sopra)

era proprio questo che mi lasciava perplesso. Facendo riferimento all'esempio, mi domandavo come facesse la "band bank" a tirare fuori 427 billion dollars di equity. E poi c'è sempre il problema di distinguere gli asset "buoni" da quelli "cattivi".

Facendo riferimento all'esempio, mi domandavo come facesse la "band bank" a tirare fuori 427 billion dollars di equity. E poi c'è sempre il problema di distinguere gli asset "buoni" da quelli "cattivi".

In principio, ma anche in pratica, non è poi così difficile distinguere le cose "buone" dalle "cattive". La cosa difficile è capire quanto valgano le cattive. Il mio mutuo è buono, e si vede facilmente, così come la linea di credito a Google. Questa parte è facile.

Quanto ci sia è irrilevante: quello che c'è c'è, non è che ci si possa far nulla al momento. Se è TUTTO cattivo, allora quella banca è proprio morta totalmente, ma non è così in generale.

Le cose complicate sono altre. Se ho un attimo poi continuo, ora devo scappare. Scusa.

My two cents on the matter. As I explained in a previous post we face two problems: first, we have to determine who is bankrupt and who is not; second, we have to make sure that bankruptcies occur at the lowest possible social cost.

The Bulow-Klemperer plan is, I think, very good for solving the second problem. The social cost of bankruptcy comes from the (temporary) disruption in the credit system. By insulating the ''good bank'' from the rest of the assets, making sure that it remains whole and operating, the BK plan effectively solves this problem. At that point the issue is how much the creditors of the original bank will be compensated, an issue which can and should be solved independently. The existing plans mix up these two aspect, wrongly linking the capacity of the banks to do their business to their capacity to compensate existing creditors. 

The BK plan doesn't solve the problem of determining who is bankrupt and who is not, in fact this is not the goal of the plan. Determining who is insolvent is equivalent to assigning a realistic price to the toxic assets. I guess that one could take two positions about the matter: a) the bankruptcy decision can be postponed; we first separate the good bank from the bad bank [''bad bank'' is a somewhat misleading term, since this would not be a bank; it would be a company owning a good bank and a bunch of toxic assets]. b) we first make an effort to assign realistic prices to the toxic assets. If it turns out that the bank is not bankrupt then no further action is needed. If it is insolvent then we can proceed with the BK plan, carving out the good bank and liquidating the bad bank.

Which approach is best depends, it seems to me, on the value of the good assets in the balance sheet of the banks. If there are enough good assets to create immediately a good bank then probably postponing the bankruptcy decision is a good idea. Otherwise, putting a price on the toxic assets becomes necessary.

The problem with the Bulow-Klemperer suggestion for bankrupcy procedure is that one has to figure out which debt-holders hold claims against the good stuff and which against the bad stuff. Right now they hold claim (in order of seniority) against a mixture of the two. If you do the splitting, what do you put in the balance sheet of the good bank, under liabilities? All the liabilities? Well, no otherwise the good bank is already broke. The most senior ones? That's not what seniority entitles you to, according to most bond covenants and exisiting regulations ... and so on.

In other words, the problem is always the same: who get's screwed? Who picks winners and loosers in this game? Until now, clearly, the government has been deciding who takes home the money and who pays for it, not the market or not even the law. Just the discretion of two different administrations. Not good.

No, I don't think this is a big problem. The liabilities are essentially of two types, deposits and non-deposits. Only deposits go into the good bank and they are therefore safe. But they are safe anyway since (mostly) they are already insured. The good bank also receives the good assets.

Currently, the other liability holders have claims against a mixture of bad assets and good assets but their claims are junior to deposits. Thus, effectively,  they have claims against bad assets and good assets minus deposits, assuming the difference is positive (which is the only case in which the BK operation makes sense). After the BK operation, they have claims against the bad bank, which means a mixture of bad assets and the equity of the good bank, which is owned by the bad bank. Since the value of the good bank is equal to the value of the good assets minus the deposits, they lose nothing. It only becomes clearer that they cannot hope for total reimbursement of their credit if the value of the bad assets is low.

Of course they lose with respect to a situation a la Paulson-Geithner plan, in which the bad assets are overpaid courtesy of taxpayers and the money is used to reimburse creditors. To make the point once again, this is exactly what I like of the BK plan. It separates the issue of how we treat creditors from the issue of keeping the banks running.

...ho fatto un piccolo sforzo e credo di avere trovato l'articolo originale di Bulow e Klemperer. Me lo leggo e spero di capirci qualcosa.