The ‘Good Bank’ Solution

In its January Global Financial Stability Report Market Update (published January 28, 2009), the IMF has raised their estimate of credit losses from bad assets originated in the US and held by US and European banks to $2.2 trillion (from $1.4 trillion estimated in November).  It also estimates the combined need of US and European banks for new capital at $0.5 trillion. Finally, the Fund recommends that the authorities take the distressed assets from the banks’ balance sheets through ‘bad bank’ arrangements.

Note that we are no longer just talking of ‘toxic’ assets, that is, assets whose value cannot be assessed with any degree of certainty because of their complexity.  We are now talking about bad or impaired assets that include the toxic stuff but also a large chunk of plain vanilla assets (real estate loans, simple mortgage products, consumer loans, corporate debt) that have become impaired because the borrowers/issuers are at risk of going belly-up the old-fashioned way.  With a long and deep recession still ahead of us, the quantity of ‘conventional’ bad assets on the books of the banks will escalate. 

There is a problem with creating a government-owned and government-funded bad bank that acquires the bad assets from the existing banks and manages the portfolio of bad assets, unconstrained by liquidity and short-term profitability considerations, either by selling them if and when the markets for these assets recover or by holding them to maturity.  The problem is that the bad assets have to be valued.  Some bad assets are transparently bad.  If there is a liquid market for these assets there is no problem in valuing them.  If there is no liquid market and can be valued using reverse auctions or model-based techniques.  But much of the toxic stuff is so obscure, so heterogeneous and held by so few parties that they are virtually impossible to value.  An auction would become a bilateral negotiation.

The favoured solution of the banks and other institutions holding these toxic assets is that the state pay them over the odds – preferably face value.  That would be unfair and costly from a budgetary point of view.  It would also represent a massive example of moral hazard by creating incentives for future excessive risk taking in the confident expectation that Mother State will bail you out.

Valuing the most toxic asset is not a problem if the banks that hold the bad assets are all in full public ownership. In that case the price paid for the assets as they are shifted from the books of the state-owned banks to the books of the state-owned bad bank is irrelevant – the payment goes out one pocket of the tax payer and into the other.  This provides another powerful argument for temporary nationalisation of all banks with toxic assets and solvency-impairing amounts of other bad assets on their portfolios.

But there are some countries where outright nationalisation of the key banks would be politically difficult.  Could the US political system work itself into a state of fear and/or indignation sufficient to take into full public ownership the likes of Citigroup, Bank of America, JPMorgan-Chase etc.?  It’s possible, but not yet likely.  Were it to happen it would no doubt be called something different: Temporary Public Stewardship, Time-Limited State Trusteeship, Patriotic Conservatorship or some such thing.

The ‘Good Bank Model’

There is an alternative solution to the problem of valuing the toxic assets.  It would not involve nationalising the existing banks.  Instead the state would create one or more new ’good’ banks – all state-owned and state-funded to begin with.  Effectively, some or all of the existing banks would become bad banks.  The good banks  would acquire the deposits and the good assets of the bad banks or legacy banks.  The good assets are, by definition, easy to value.  The creation of multiple good banks may be desirable to encourage competition.  One could even create a good bank for every existing bank: New Citi, New RBS, New ING etc.

New lending business, indeed all new business activity would be undertaken only by the new good banks.  To address the credit crunch, government guarantees or insurance could be provided for new loans and investments made by the good banks.  No further guarantees should be extended to existing assets, either in the good banks or in the legacy bad banks.  The good banks would receive their capital from the state.  Other funding would be provided by the transfer of the deposits from the bad legacy banks, through loans from the state or through the sale of bonds by the new good banks to the state.  The state could also guarantee new loans to the new good banks from the private sector or bonds issued by the new good banks and purchased by the private sector.

As regards the the legacy bad banks, the easiest and cleanest way to proceed is to stop them from doing any new business on the asset side of their balance sheets: no new lending and no new investment.  They would also not be permitted to take new deposits.  A simple way to ensure this is to take away their banking licenses.  They would exist only to manage and ultimately to run down the portfolio of bad and toxic assets they hold on their balance sheets.  Maturing liabilities could be refinanced if that made more sense than accelerating the sale of the assets.

Taking away the banking licenses of most of the existing large universal banks in Europe and the USA would be appropriate because recent developments have demonstrated that the existing institutions, managements and boards are not fit for purpose.  They have failed as banks, even if they have not (yet) failed in the technical, legal sense of becoming insolvent.  For most of them the past, present and anticipated future financial support of the state is the only thing that stands between them and bankruptcy.

The bad banks (i.e. the existing banks minus their deposits and good assets (and including the compensation for the difference between these two)  could retain their existing ownership structure.  They would not receive any further financial support from the state, whether through capital injections or through guarantees of their assets or liabilities.  They would be left to swim or sink on their own, without any further financial support from the state. If they were to become insolvent, they would go into the normal insolvency procedures for non-bank financial institutions.  The existing unsecured creditors could, in reverse order of seniority, be converted into shareholders of the insolvent company, thus ensuring that the incentives for prudent lending to banks will be stronger in the future than they have obviously been in the past.

The new good banks would initially be wholly state-owned and either state-funded or privately funded but with a sovereign guarantee to begin with.  They should, however, be managed commercially – at least at the micro level, that is, at the level of the selection of individual investments and the making of individual loans.  I think it is unavoidable and even necessary in the current credit crunch, that the new good banks be set aggregate targets for  lending, but only at the most aggregative level, e.g. so much to the non-financial enterprise sector, so much to the household sector and no more than so much to the financial sector.  Falling short of the quota or exceeding it would result in the forfeit of the shortfall/excess.   If the performance criteria for executive remuneration were structured properly, the individual project and loan selection could still be long-term profit maximising (or loss minimizing) but subject to an aggregate lending constraint.

The aggregate lending target would be dropped as soon as credit markets normalised.

To minimise the risk of political micro-management of lending and investment decisions, management should be incentivised to act commercially by having in their remuneration package a large bonus component (!) that depends on long-term profitability.  A fiercely independent and qualified board would have to be appointed, instead of the old duffers and cronies that make up most bank boards today.  The CEO could not be a member of the board, let alone its chair.

In my proposal, the existing banks would become the bad banks and retain their existing ownership structure.  No government resources would be wasted propping up the valuations of existing assets.  All government financial support would go to the new state-owned good bank.  Even there, government guarantees would only be provided for new bank borrowing (if this is from the private sector) and for new bank lending and investment.  There is no point here either in propping up the valuation of existing assets.

It is possible that the legacy bad banks would initially have balance sheets that are much larger than the balance sheets of the new good banks, which would contain only the good assets bought from the legacy bad banks and any new assets built up since the new good banks were created.  The combined balance sheet of the legacy bad banks in the USA could easily run into multiple trillions of US dollars (probably somewhere between $3 trillion and $5 trillion, depending on how strictly you apply the criterion that the good banks only purchase easy-to-value assets).  But with all new lending and investment performed by the new good banks and with the balance sheets of the legacy bad banks shrinking by construction, the new good banks would be growing relative to the legacy bad banks.  With a bit of luck, they could grow into bad banks again!

As credit markets normalise and the economy recovers, the aggregate lending targets and the government guarantees for new lending and for borrowing from the private sector would be eliminated.  In due course, but probably not before the third year of their existence, the privatisation of the new good banks could be contemplated.


The ‘good bank’ model with its state-owned and funded good bank(s) and its legacy bad banks that retain their current ownership structure is superior to the ‘bad bank’ model in which the state owns the new bad bank(s) and the legacy banks (now good banks since the removal from their balance sheets of the toxic assets) retain their current ownership structure.  First, the good bank model only requires the good assets to be valued, which is a lot easier than the bad bank model.  Second, the good bank model concentrates government financial support on new lending and investment flows and on new bank borrowing, and does so through an institution owned by the tax payers and where the tax payers have a chance of some upside when the banks are privatised in due course.

Finally, which politician would not prefer to be associated with the creation of a good bank rather than with the creation of a bad bank?

Maverecon: Willem Buiter

Willem Buiter's blog ran until December 2009. This blog is no longer active but it remains open as an archive.

Professor of European Political Economy, London School of Economics and Political Science; former chief economist of the EBRD, former external member of the MPC; adviser to international organisations, governments, central banks and private financial institutions.

Willem Buiter's website

Maverecon: a guide

Comment: To comment, please register with, which you can do for free here. Please also read our comments policy here.
Contact: You can write to Willem by using the email addresses shown on his website.
Time: UK time is shown on posts.
Follow: Links to the blog's Twitter and RSS feeds are at the top of the page. You can also read Maverecon on your mobile device, by going to