Forget Bad Banks, Why Not New Banks?

I’ve written here before about my puzzlement over the widespread international enthusiasm for “bad bank” proposals and I haven’t changed my mind since.

A more attractive proposal, which is getting less attention, is the idea of establishing new banks.  It could be argued that this directly addresses the problems being created by the weak capitalization of the international banking system, without the extreme moral hazard problem associated with TARP-style over-paying for bad assets.  The financial intermediation function performed by banks is crucial to the efficient functioning of the economy and, for a number of reasons, undercapitalized banks do not perform this function well. Understanding this, the approach of govenments everywhere has been to use taxpayers money to prop up undercapitalized banks. But while banks play a crucial role that doesn’t mean that we necessarily need the current set of banks to perform this role.

Kevin already posted a link to Willem Buiter proposing something like this but the idea is now being given wider prominence. Here for instance is a nice clearly-written piece from today’s Wall Street Journal by Stanford’s Paul Romer.  An important benefit of this type of plan, as Romer notes, is that it seems more likely to attract additional private sector equity capital relative to the various plans to attract new investors for existing banks with failed management and murky balance sheets. Indeed, I first read about the idea of new banks in this 2009 predictions piece from celebrity uber-bear bank analyst Meredith Whitney and she was focusing purely on the private sector opportunities. She said:  “I think you’ll see more new banks created. We’ve already seen more applications. And it’s a great idea: You start with a clean balance sheet and make loans today with today’s information. Plus, right now you’ve got a yield curve that’s good for lending.”

Skeptics could point out that these new banks will lack the branch network or knowledge capital of existing banks.  However, branch networks could be purchased pretty cheaply these days and the newly-minted banks could be attractive places for those bankers with good track records to work. Perhaps the best argument against this idea is the time lags involved in getting new banks set up. But the problems in the international banking system seem likely to be with us for some time so useful long-term solutions may be called for.

No doubt I’m being wide-eyed and innocent here.  Perhaps our trusty band of loyal commenters can give me a word to the wise.

37 replies on “Forget Bad Banks, Why Not New Banks?”

The term “bad bank” has been accumulating so much baggage recently that it is no wonder it may seem like a “bad idea”. In particular, it has come to seem like an easy option — just drop your bad assets into the bad bank and proceed as if nothing had happened. This seems to have been Paulson’s version.

Traditionally, though, the term “bad bank” is simply used to define an entity created exactly to achieve the goals proposed by Buiter and Romer — a clean break with the past for the banks that are going to be allowed to go forward. A bad bank is specifically a receptacle for housing and managing toxic assets.

So, if there is an identifiable large block of problematic and hard to value assets, you will likely need to have a bad bank structure to hold these assets separately from the rest of the failing bank’s business. That allows the authorities to sell off, or restructure, the viable assets and the rest of the bank’s business.

Isolating the problematic and hard-to-value assets is useful, whether or not the going concern banks that remain in the system are “de novo” legal entities or a restructuring of existing legal entities. And it has no bearing on whether or not the shareholders and management of the old bank stay in control.

What, fundamentally, is wrong with letting the banks go bankrupt? Wiping out the current shareholders and management is probably a good thing. As far as I can see, the only problem, and a big problem at that, is that bondholders will end up owning the bank (or its various pieces) by having shares in a fairly worthless bank. (That’s my limited understanding, not being an expert on anything, especially not bankruptcy). And it wouldn’t be a good idea for all the mortgage debt in Ireland to be owed to assorted foreigners investors. Therefore, the state must bail it out, giving billions of free euros to bondholders who, in the normal run of events would be left with scraps of shares.

The ideal solution would be to allow the government to totally wipe out the shareholders and management of any bank that has taken a cent of taxpayer money, taking the bank into public ownership and bypassing the usual queue of creditors. The difficulty is in how to explain this “theft” to the bondholders and other creditors, who also deserve to be wiped out (with Anglo, they got a high return for the risk, and hence must face the reality).

This is where we need an international agreement to change bank bankruptcy procedures. This should allow states to bankrupt a bank and nationalize it in one step, leaving the creditors all over the world with the derisory shareholdings they deserve, but keeping the control and majority shareholding with the host state until they deem fit to privatize. Only individual depositors should be entitled to benefit from any bailout.

In future, all countries should limit foreign ownership of banks (or anything bank-like that needs to be bailed out like a bank). This’d make it much easier for a state to take what steps it needs. Rescuing your banking system should be essentially a ‘private’ matter and shouldn’t require giving free billions to assorted foreigners!

Maybe we should just ban banks and let building societies expand again. Funding loans from local depositors seems much more sensible.

And finally, somebody somewhere has to owe China 2 trillion dollars? We can’t just play pass the parcel between bad banks, somebody has to owe it to them!

The problem with letting banks go bankrupt is that the government guarantee would be activated immediately, and the government would have to pay out compensation. The compensation would count as current expenditure. And any more current expenditure is just not an option. That’s my view on what the issue is anyway.

Rescuing a bank has nothing to do with foreigners that I can see. If you let the bank go down and you’ve given a guarantee, you’re going to have to pay out to all the depositors including J Foreigner. You shouldn’t have to pay out to shareholders, no matter where they are from.

If you ban banks from expanding, they will be even more dependent on funding loans from external sources. A country like Ireland needs capital a lot more than a country like Germany. It makes sense to balance the two, somehow.

It is quite true that creating ‘bad banks’ is just pass-the-parcel, and hoping something good has happened before the music finally stops.

My opinion is that we should focus on creating not just good banks, but great banks, and that banks should have sufficient scale to operate with relatively low margins (i.e., offering cheaper finance on a rational, managed basis). It is worth looking at the example of LTCB/Shinsei in Japan, a bad bank made good.

I get the impression that Anglo, like so many banks, got money not from deposits but from borrowing from institutions. Therefore, the debts that have dragged Anglo down are not deposits but these bonds issued by Anglo.

Presumably, these bonds would have been bought by institutions inside and outside Ireland. The way that Anglo was nationalized represents a corrupt, unnecessary, free gift to gamblers. Idiots that lend money, via bonds, to mismanaged banks and businesses are supposed to lose money when that business goes bankrupt.

As far as I know, when an ordinary business goes bankrupt the business is taken from the current shareholders and given to the creditors. The creditors (e.g. banks that lent the business money) are no longer owed money by the bankrupt business, but are instead given shares in the new business.

So, if Anglo was allowed to go bust in the normal way, these bondholders would be left with shares worth a small amount of money. They would no longer be owed 22 billion euros. My argument is that the government should be allowed to bail out a bank while also giving the bondholders what they deserve, i.e. a few measly shares.

And the deposits don’t need to be paid out immediately. That’s why I propose that the bank be effectively bankrupted and nationalised overnight – the depositors would still be owed money by their bank, it would just be a nationalized bank instead.

The only change I am proposing from the status quo is that the bondholders, who deliberately gambled 22 billion on a mismanaged bank, should see their bonds cut back drastically. Maybe the nationalized Anglo should owe them only 22 million in bonds. This is entirely fair because they lent to a dodgy bank and such businesses go bankrupt every day, leaving the bondholders with little or nothing.

My understanding is the following. In general terms, a bondholder is not a shareholder. A bond is essentially a loan (a promise to pay back money at some future date, a bit like a bank deposit). The government guaranteed all bank debt for two years, so that banks could borrow more money day-to-day and keep trading.

If you actually allow the bank to go bankrupt, nothing much will happen that didn’t happen in the case of Anglo Irish. The only difference is that if the bank actually shuts the doors of the bank and refuses to pay depositors on demand, all pandemonium will break loose. You would quite likely end up with something looking very much like a riot on Dame St as retail depositors rushed to the Central Bank to collect on their deposit guarantees, which they would be entitled to do immediately, and the Central Bank would have a tough time actually verifying or processing these claims. So what you are describing is essentially what the bank eventually did with Anglo, although they wisely avoided the part where the doors actually closed. Now you could argue that they should have done that back in October, rather than putting in the guarantee. And with the bnefit of hindsight, it would be very difficult to definitively prove you wrong.

One very practical problem with defaulting on bonds is that when it happens, it is likely to cause problems somewhere else in the world, screwing with someone else’s economy. That has to be taken into consideration. The people who ultimately are owed money on those bonds are most likely ordinary people dependent on them for their pensions or businesses that happened to have cash reserves but which now face tough times. We’re all in this thing together.

So the whole thing is like a very shaky, very unsafe condemned terraced building that contains some very valuable furniture and features. Ultimately, you have to have your detailed plan to knock the thing down or buttress it up, but you have to go about things in a pretty delicate way, or else the whole street will come crashing down around you.

I agree with what you’re saying about a bankruptcy. What I’m saying is that, in the process of nationalization, the government should default on bonds while keeping the bank open as usual.

You say that bondholders are likely to be ordinary people, but the same goes of shareholders (and mortgage holders). I don’t care if they are ordinary people or not, banks bonds and bank shares are something that gamblers enter into of their own accord. You surely wouldn’t recommend that the stay buy all shares and houses at the price they reached at the top of the boom? It’s inconsistent to bail out the bondholders while leaving the shareholders and mortgage holders fend for themselves.

We can’t bail out everybody, and the bondholders deserve to lose out just as much as the shareholders.

yield curve does favour lending and the idea of a bank without the problems inherent in existing banks sounds attractive. As always though there are several issues

1. where does the deposit side of the balance sheet get funding? is anybody going to deposit hard to come by cash in a ‘we’ll pay interest in a few months once we get lending going’ bank?
2. The likelihood is that they would not have a large capital base and in a dead securitization market they would stop lending fairly rapidly as they could not surpass the new level capital requirements (which are part of what got us here to begin with)
3. The lending margins would have to be huge in order to balance the deposit side (ie: offering attractive deposits to get funding) but the only people who would sign up for this would be desperate borrowers so in effect you might be creating a ‘new’ bad bank. Financially strong clients are already getting good deals from the present banks.

there is a moral sincerity in a ‘fresh start’ but you have to fix the sytem to begin with before you start creating new banks within the existing broken one.

I think we should not underestimate the informational capital existing banks have. Notably, in his study of the Great Depression, Bernanke stressed that bank failures exacerbated the informational asymmetries that plague credit markets.
Moreover, I’m wary of suggestions that government recapitalisation will induce flows of private capital into banks. To me, this makes little, if any, sense. If investors desire a diversified portfolio, then government investment would simply crowd out private investment. After all, since the public have a claim on the returns, government investment is de facto *their* investment too.

Letting banks, any bank, go bankrupt abruptly is not a good idea. A nanosecond later (or on market open) the other banks are sold to zero, the bond and institutional lenders demand repayment, the system siezes up and then the state is holding not a contingent liability but an actual liability of 400b.
Apart from that, its a good idea.

Brian, Antoin,
I’m starting to see your points. Looking at Anglo’s balance sheet (in the 2008 Preliminary results below), the deposits alone (Customer Accounts 51 billion) are more than the value of the loan book (72 billion or, in my opinion, 72 Million!). Hence, even if the bondholders were wiped out, the bank would still be a bankrupt wreck.

The ‘Debt Securities in Issue’ (bonds?) are only 17b, less than the 22b reported by the Irish Times; I suspect the Irish Times article I linked to above was using out of date data from 2007

Here’s the 2008 figures.

Once the State gave the guarantee, we lost the option of a “good bank” i.e. letting the bad banks sink, at least until October 2010. My guess is that the guarantee will be extended before the end of this year because the banks won’t be able to borrow if there is less than one year left in the guarantee.

As events unfold, it becomes clear that the guarantee was a stroke of genius by the banks. The Government thought they were throwing the banks a lifeline, in fact the guarantee has lashed the country’s fortunes to the future of the banks.

Brian Cowen told Marian Finucane this morning that he sat in silence while the bankers made their case on that (in)famous night of 30 September. He said the bankers then left the room and the Government made the decision independently. The course of our economic history was changed on that night and we may never know the full story.

indeed . as I put it they had to throw something on the landmines of the banks toxic assets and they threw the economy.
I have to say that I doubt very much that the government made the decsiion independently. They had to have advice and they made it on that advice. I assume that is what they mean – I cant see the two Brians and Mary deciding alone. But, maybe….

Absent the Government guarantee, in the event Anglo was let go broke the stockholders are wiped out and the creditors — bondholders and depositors — are left owning the bank. Assume the deposit guarantee system takes care of the depositors, then the bondholders must settle for, say, 60 cents on the dollar with the receiver. This bankruptcy would only create problems for other banks if they are seen to have same solvency issues. If Anglo is seen to be an “outliner” in terms of asset quality, then they may be OK. One option would be to give guarantee to systemically important banks — AIB and BOI — and let the others go broke and skin their bondholders.

There is a lot of careless talk about letting bondholders go bust, when the conversation turns to toxic banks. People might be surprised by how many of these bonds they actually own themselves…….in their pension funds!

Furthermore, who owns most of the shares in Irish banks. Again, I would place a sizable bet that our pension funds are stuffed full of those too!

The point is that there is no panacea that’ll keep everyone happy. Maybe our pensions will have to take a hit to save the broader economy and our government from defaulting. I’m probably oversimplifying; that isn’t necessarily the choice we are facing.

But something somewhere has to give sooner or later. Perhaps we just have to hang tight until the EU decide to devalue the euro. “We’ll vote Yes to Lisbon if you devalue the euro”

Devaluation is arguably just another means of taking money off people with savings. There is no easy solution.

there is anotherset of issues that Romer and Buiter and Karl I think forget. a_ Lets say we start the Bank of Mani Puliti with 8b in government loans. Now, we need to get say 100b in deposits and loans from the markets. Despite what Karl thinks there is no guarantee that in that instance the now not to be capitalised banks will realise that they are zombies and fold their tent. At least it will take time for the system to realise that they are dead. People do not move their money rationally and so small deposits will take time to move. There will be no firesale of branches so it will have to be an online bank (and we already have Rabo, backed by the Dutch, so wheres the competitive advantage…). So the logistisc of attracting deposits shouldnt be underestimated
b) The interbank markets will not pull their money from the existing instantly and if they did we would have a disorderly collapse, which we are all agreed would be poor for the country. So, BMP will have to raise money on the interbank markets at , I suspect, a rate higher than the effective present rate of interest that AIBOI have to pay. After all, they are new, and are run by the Irish state, which has not got a great track record. Similarily, the deposit rate to entice people to bank with us as opposed to the Dutch would have to be higher. The overall effect is that the cost of funds from BMP will likely be higher than that from AIBOI in the short to medium term. this raises the cost of corporate funds, which depresses economic activity. This is not like the USA where the sovereign equity holder is less risky than the esisting equity holder. Au cointreau….
c) Private equity coming in to the irish banks would exacerbate this. PE would want, and deserve, 20% plus for the risk they are taking on.

Bottom line – setting up a replacement banking system is logistically a nightmare and there are significant second and third order effects especially in a small economy that need to be carefully thought out. We do not live in a model – we live in a society and a networked economy, and we need in our policy presctiptions and musings to remember this. Else there is a serious danger that the best will become the enemy of the good in relation to the rapidly accellerating mess we are in. Its best, imho, to fix the exisiting banks (and yes, have a toxic bank) than to experiment with ideal world models. We are in a storm and redesigning the ship on the hoof is probably not a good idea – best to patch the holes and sail for calmer water

“sume the deposit guarantee system takes care of the depositors”
The problem with this is that the value of the deposits of AIB/BOI/Anglo is what, 300b? So the guarantee if it were called in would be worthless. Pretending otherwise is interesting but not realistic. And if one were to be called in then all would be. Once we gave the guarantee to them all they all became as strong only as the weakest link in the chain – which may not be Anglo. Is INBS stronger or weaker than Anglo?

Just to clarify for Brian, my comments related to the general international discussion about how to deal with banking system problems (note the word “international” and the links to proposals relating to UK and US banking systems.) Without doubt, the government’s bank guarantee greatly complicates the current Irish situation.

That said, I still think these proposals provide some useful food for thought in the Irish case. They help in clarifying that the issues of dealing with failed banks on the one hand, and getting new lending going, on the other hand, can (at least in theory) be separated. For sure, with deposit (and other) guarantees in place, it is in all governments’ best interests to deal with insolvent banks in the most orderly way possible and to get as much as possible in return for their assets. But using taxpayers money to deliberately overpay for bad assets or fully recapitalize existing banks are not necessarily the only options.

“Letting banks, any bank, go bankrupt abruptly is not a good idea.”

This being the case, what is the case for having a private banking system? I mean, I thought the market is supposed to be better than the state at providing goods/services because the inefficient go out of business. If we can never let banks go bust, then what?

I think the issue is “abruptly” : orderly winddown is one thing, collapse another. I think we saw that with Lehmans.
Yes, the international element is key, your right.

Say we nationalise the banks today. And then sell off the good bits to foreign or domestic investors. Is this not essentially equivalent to the good bank proposal, and why would it not work?

The nationalization route is being actively discussed elsewhere. It would “work”. Whether doing so now would be the most expedient for the taxpayer and for the economy at large at this moment is less clear. (Of course there are also other interests in play).

And there are other, more subtle, financial engineering solutions available (some of them discussed here already) which might give better protection against moral hazard and againt risk of short-termist political influence over lending decisions. Some of these are also being discussed elsewhere.

But Ireland is a bit exposed now to be innovating here. That’s why I have been suggesting a wait-and-see attitude, in anticipation of announcements from US, Germany and UK in the coming weeks. We could more safely follow an established exemplar.

Any intervention now does limit our options. Saturday morning’s newspaper articles suggest that what is currently on the table is generically the same as UK-type intervention of last October. (But we assume this is a selective leak). As such, it may be less limiting than some other options.

But still…a little patience seems best, even if the banks are frustrated at their lack of financial independence, and even if allowing undercapitalized banks to operate entails some risks.

Willem Buiter has written again on the subject. His piece includes this:

“The US, the UK and several other continental European countries are at risk of emulating Ireland, where the government first guaranteed all the liabilities of the banks (other than equity) and only after that began to nationalise the banks. This leaves the Irish government today in the not too enviable position of having to choose between sovereign default and bleeding the tax payer and the beneficiaries of normal public spending to make whole all the creditors of the banks.”

I think the Buiter article linked to by, ahem, Brian J. Goggin …. has the answer to Kevin’s question of “Say we nationalise the banks today. And then sell off the good bits to foreign or domestic investors. Is this not essentially equivalent to the good bank proposal, and why would it not work?” If “the good bits means all the deposits and other guaranteed liabilities but only the good assets, then these banks would be undercapitalized.”

Buiter writes about his “new good banks”: “Under my new good banks proposal, the new good banks would take on the (guaranteed or insured) deposits of the legacy bad banks (which would lose their banking licenses) and would buy the good assets of the legacy banks. Should deposits exceed good assets, the state would have to make up the difference initially with government debt on the balance sheet of the new good banks. Should deposits be less than good assets, the new good banks would be able to borrow from the sovereign to finance the acquisition of the good assets from the legacy bad banks. This would cleanse bank balance sheets and transform them into good banks but leave them undercapitalized. Soros suggests that $1 trillion of the estimated $1.5 trillion required to recapitalise the existing banking system should be directed to the cleansed banks. Soros believes or hopes that some of the money required to capitalise the new, cleansed banks could come from the private sector. Under my proposal, and that of Stiglitz, the state would initially capitalise the new banks on its own. “

That makes sense, and from my layman’s perspective it seems preferable to use government money to set up banks that will work, than to use government money to prop up banks that are not working.

@ Karl:
I understand that there may be some person, with a name similar to my own, employed in some branch of trade or commerce. I am happy to state that that person is not related to me in any way: nobody in my family has ever had any talent for making money.

On the other hand, we’ve never had the opportunity to lose quite such large amounts of it either.


I have been following your prosposal and ensuing discussion with interest. I was initially sceptical, but am starting to see the case for what you suggest — even for Ireland.

Can you let me know if I’ve got the key elements straight?

Good banks–
Assets: “Good” assets from the legacy banks, government bonds, and new loans over time
Liabilities: All guaranteed liabilities from the legacy banks
Ownership: Initially government, but private equity raised over time, with eventual full privatisation
Branch network: Network of the legacy banks
Not quite sure what happens to previous owners of legacy banks. Some options: (i) wiped out completely; (ii) compensated based on remaining value of legacy bank; or (iii) given an ownership stake in the new bank

Bad banks–
Assets: Remaining “toxic” assets from the legacy banks
Liabilities: None
Ownership: Government
Challenge: Managing the asset portfolio to recoup value for the state. (Niggling questions: Who would do the managment? How much forbearance is optimal? How much forbearance is likely given inevitable politicisation?)

A great advantage is that it provides more clarity on the commitment of the state, which I’m assuming would be very large. Of course, there would still be uncertainty, including the ultimate value of the toxic portfolio. A disadvantage is that this clarity is likely to come with some sticker shock.

One final worry: Will there be many willing creditworthy borrowers for the good banks to lend to in the immediate future? Maybe we just need to be patient.

Here’s another idea.

The government gets involved in the property. It buys property that is likely to be of value to it for around half (or a third or whatever) the peak price. That will effectively allow cash to be returned to the banks, although not too much. It will also draw out any other land purchasers. Worst case scenario, the property would still not be worth a lot, but the government would actually own something and would avoid getting too deeply involved with the management of the banks.

The next big issue is going to be commercial debt as the recession takes hold. It’s a lot harder to quantify the value of commercial debt than it is property debt.

Let’s be clear. There is absolutely nothing heterodox or controversial about the “new bank” proposal.

It is in fact equivalent to the nationalization+bad bank proposal.

Thus, obviously the new bank would acquire all or most of the branches and take on all or most of the staff (but not top management). It benefits from an injection of government capital. And it is to be sold back into the private sector as quickly as possible.

So it is essentially identical to the solution adopted in WaMU and Bradford and Bingley as evoked in my paper for the crisis conference last month.

(One small but significant difference — subordinated bondholders in those banks took a bath, whereas some of them have a government guarantee for the next 20 months).

Agreed Patrick. It’s not a particularly controversial approach. That said, for all its merits, it does not seem to be the route the Obama administration (or the Irish government) is going to go down. At least not yet.

One of the problems with discussions on this issue is the confusion created by multiple possible meanings of “bad bank”. The good meaning (Patrick’s meaning) relates to the best way to minimise taxpayer losses after nationalisation: Putting together a team of workout experts to get the best prices over time for the bad assets. This is what the Resolution Trust Corporation — apparently many people’s model for the “bad bank” — did with failed S&Ls.

The bad “bad bank” proposal involves taking over one insolvent bank (for instance) and using it to go out and pay over the odds for other banks’ toxic assets. This is my interpretation of most of the casual references to bad banks lately and this just seems like a bad idea. Insuring the losses on bad assets is just as bad — the only difference being that governments will be paying out in a drip-drip fashion instead of straight off.

Beyond all of that, given the multi-faceted problems related to dealing with failing privately-owned banks, there still seems to me to be some merit in using at least some of the large amounts of money being used internationally for re-capitalizations ($350 billion is a lot of money) to set up some completely new banks (not just the “new WAMU” or “new B&B” banks associated with FDIC-style purchase and assumption procedures.) But perhaps it’s just not practical.

According to today’s Eurointelligence, for what it’s worth:

European Commission sets out guidelines for bad banks

FT Deutschland has obtained a paper prepared by the Commission for today’s Economic and Finance Committee meeting in Brussels, which includes a detailed set of procedures European governments should be following in setting up bad banks. The Commission wants that toxic assets are written off at market value, and that any write-offs will first have to come out of shareholders’ capital. Once the capital falls below the Basel ceilings, it is up to the government to provide new capital, or to force the bank into bankruptcy procedures. All fine and good, but how about the small problem that there is no market for most of these assets, and thus no market price? The article says the Commission wants to leave that little detail to the bank supervisory authorities.


you will be glad to know that I have already spotted the above and I am currently applying for a banking license. I think depositors will be tempted through sheer annoyance with the current set up. watch this space

Comments are closed.