Ruminations on banking

Original Link: http://blogs.ft.com/maverecon/2009/04/ruminations-on-banking/

By Willem Buiter

(1) The autodafé of the unsecured creditors is coming to a US bank near you

A binding budget constraint sure concentrates the mind, even for the US Treasury. There is just one way to make the US government’s policy towards the banks work. That is for the Congress to vote another $1.5 trillion worth of additional TARP money for the banks - $1 trillion to buy the remaining toxic assets off their balance sheets, and $0.5 trillion worth of additional capital. The likelihood of the US Congress voting even a nickel in additional financial support for the banks is zero.

There is no real money left in the original $700 bn TARP facility - somewhere between $ 100 bn and 150 bn - to do more than stabilise a couple of pawn shops. The Treasury has been playing for time by raiding the resources of the FDIC (which, apart from the meagre insurance premiums it collects, has no resources other than what the Treasury grants it) and of the Fed. The Fed has taken an open position in private credit risk to the tune of many hundreds of billions of dollars. Before this crisis is over, its exposure to private sector default risk could be counted in trillions of dollars.

In addition to looking for money in off-budget and off-balance sheet places (and out of sight of Congress), the US Treasury has also tried to hide true extent of the problems of the US banks. I addition to supporting the FASB’s recent proposals for increasing managerial discretion as to the way illiquid assets are accounted for (that is, condoning the issuance of another license to lie), the Treasury appears to be using the ‘Stress Tests’ announced as part of the Financial Stability Plan, as a mechanism to play for time and gamble for resurrection.[1] I base this on what I have been picking up about the reality of these Stress Tests.

1. The actual decline of the real economy thus far is already steeper and deeper than assumed in the Stress Tests.

2. The Stress Tests focus on the 40% of the banks’ balance sheets consisting of securities, rather than on the 60% consisting of conventional loans. The securities (including the toxic waste) is where most of the old problems of the banking sector are concentrated, that is the problems incurred as a result of the pre-August 2007 speculative frenzy. The loan book contains the stuff that will go bad as a result of the steep and deep contraction in real economic activity the US has been in since Q4, but that will not show up in the banks’ reports until this summer at the earliest.

3. The bulk of the information provided to the authorities by the banks is private information to the banks that is virtually impossible to verify independently. Too many banks have lied about their exposure too many times for me to feel confident about the quality of the information the banks have been providing as part of these Stress Tests.

As a result I now expect a clean bill of health for the banks from the Stress Tests. For most banks this will turn out to be incorrect before the end of the year. At that point, the de facto insolvency of much of the US border-crossing banking system will become so self-evident, that even the joint and several obfuscation of banks and Treasury will be unable to deny the obvious. There still will be no fiscal resources available to sanitise the banks’ balance sheets by purchasing or guaranteeing the old toxic assets and new bad assets.

At that point, only the ‘good bank solution’, which requires either a serious hair cut for unsecured creditors or a mandatory conversion of debt into equity will be viable, simply because the bad bank solution requires additional public money which isn’t there. (You creating a new good bank out of the assets of the old bank and the insured deposits and counterparty claims on the old bank, leaving the unsecured creditors of the old bank with a claim on the equity in the new good bank; a bad bank requires funds to buy the toxic and bad assets from the old bank and addition resources to capitalise the bad bank).

We will have wasted a lot of time - the good bank solution and the slaughter of the unsecured creditors should have been pursued actively as soon as it became clear that most of the US border-crossing banking system was insolvent, but for past, present and anticipated future tax payer support. If the Treasury can be pushed into a pro-active policy by declaring, just before the beginning of the weekend, that most of the banks undergoing the Stress Test have failed them and moving these wonky institutions straight into the FDIC’s special resolution regime where they can be restructured according to the good bank model, we could have well-capitalised banks capable of new lending and borrowing by the beginning of next week.

The same policy should be pursued wherever banks have failed: it never makes sense to put the interests of the unsecured creditors before those of the tax payers. It is bad economics in the short run and in the long run. And it is political poison. I fear, however, that only in those countries where there is no fiscal spare capacity (as in the US, for political reasons or in Iceland, for economic reasons), the right solution to bank restructuring will be adopted. Elsewhere the unsecured creditors will continue to feed off the carcases of the tax payers and the beneficiaries of public spending programs that will have to be sacrificed to foot the bill.

(2) Too big to fail means too big

No country should ever find itself in position of Iceland, with systemically important banks or other financial institutions that are too large to save. The fiscal spare capacity of the government (its ability to raise future primary (non-interest) surpluses has to be sufficient to take care of any possible solvency gap in the systemically important part of their financial institutions.

If the too big to save problem can be resolved rather easily, the too big to fail issue has been with us for so long that one assumes there must be powerful forces sustaining large and complex financial institutions. The assumption is correct, but these forces are political, not economic or efficiency-related. Economies of scale for banks are exhausted well before a balance sheet size of $100 bn is achieved. Synergies between commercial banking and investment banking activities are essentially non-existent. The multiplication of products and services offered and of roles played by a financial institution can be privately profitable, because it allows the exploitation of the gains from conflicts of interest. Chinese walls, the industry’s answer to potential conflicts of interests, are aptly named. The Great Wall of China never kept the barbarians out or the Chinese in. Chinese Walls in banks, auditing companies, rating agencies and other financial enterprises don’t stop any information that is commercially profitable from getting across the boundaries. Financial supermarkets lose focus and ultimately become Citigroup - a conglomeration of worst-practice from across the financial spectrum.

There is no ‘too interconnected-to-fail’ problem separate from the ‘too-big-to-fail’ problem. I can be infinitely interconnected. If I operate on a small scale, I and my interconnections are immaterial from a systemic stability perspective.

Banks and other financial supermarkets want to be large for two reasons. The first is monopoly power. This causes banks to want to be large in any specific activity. It’s common to every industry and to all human activity. That’s what we ought to have anti-trust or pro-competition policies for. The second reason is that financial supermarkets can shelter non-systemically important profitable operations under the heavily subsidised public umbrella provided by the state to a few systemically important operations (deposit taking, payment, clearing and settlement systems, counterparty and custodial services) though lender-of-last-resort support (provided by the central bank) and through recapitaliser-of-last-resort support (provided by the Treasury).

There is no economic reason for large banks. Therefore banks should be kept small. An obvious mechanism (apart from aggressive anti-trust policy) is to tax bank size. One way to do this is through making regulatory capital requirements increasing in the size of the bank’s activities. For instance, tier one capital as a share of (unweighted) assets could be made an increasing function of the value of the assets. Gary Becker has made a similar proposal.

Governments everywhere should be focusing on breaking up banks and keeping them small. If some banking activity (or indeed any other economic activity) is deemed to have a minimum optimum scale that is makes it too large to fail, it should be publicly owned. Small is beautiful for banks.

(3) The repatriation of cross-border banking

Even if we agree that a particular bank or other financial institution is too large to fail, as soon as there are multiple fiscal authorities and border-crossing banks, there remains the unanswered question as to who should bail it out. The host country fiscal authority? The home country fiscal authority? Both together through some ex-ante or ex-post negotiated sharing rule? A dedicated supranational fiscal authority?

Recent events have made it clear that, as my colleague Charles Goodhart puts it, international financial institutions are international in life, but national in death. Any systemically important financial institution has to be backed by a central bank (for short-term liquidity support) and by a Treasury or ministry of finance (for recapitalisation and other long-term financial support when insolvency is the issue). If both your liquidity and your solvency are publicly insured or guaranteed (at highly subsidised rates), you need to be regulated and supervised, lest you be tempted to take insane risks.

This means that conventional border-crossing banks and other systemically important financial institutions will become a thing of the past. We will not see the kind of cross-border branch banks that we have seen during the past decades for very much longer. These foreign branches are not independently capitalised, have no independent, ring-fenced sources of liquidity, are often effectively managed from the home country (the country of the parent bank), are regulated and supervised by the home country regulator and supervisor, with lender-of-last-resort support (if any) from the home country central bank and fiscal support (if any) from the home country Treasury.

In the European Union, the relevant sections of the financial services action plan are now dead and need to be replaced unless we (a) get a single European supervisor and regulator for border-crossing banks and other systemically important financial institutions and (5) create a supranational fiscal Europe capable of dealing with insolvency threats to border-crossing systemically important financial institutions. Without a single regulator-supervisor and a sufficiently developed ‘fiscal Europe’, the principles of mutual recognition and the “single passport”, a system which allows financial services operators legally established in one Member State to establish/provide their services in the other Member States without further authorisation requirements, will not be permitted to operate in the field of banking and other systemically important border-crossing financial institutions and products. Need I say Icesave? (for a further discussion of the issues involved see the paper by Charles Goodhart and Dirk Schoenmaker (2009), “Fiscal Burden Sharing in Cross-Border Banking Crises” International Journal of Central Banking, Vol. 5, No. 1, 141-165,

There will no doubt continue to be cross-border banking subsidiaries. These will, however, be independently capitalised. Their liquidity will not be pooled between parent and subsidiaries, but will have to be ring-fenced for each subsidiary. They will be regulated and supervised by the host country (as they often are today). Lender-of-last-resort support, if any, will be provided by the host country central bank (as it often is today) and fiscal support when insolvency threatens will be provided by the host country fiscal authority.

The reason for host country supervision and regulation is simple: the pain is local, so the control will have to be local. The reason for host-country bail-outs is even simpler: tax payers are national. They will not accept the use of their taxes in bail-outs of foreign shareholders, unsecured creditors and counterparties. The US authorities have been able to channel somewhere between $40 bn and $50 bn of US tax payers’ money to foreign counterparties of AIG, but that is unlikely to be the new status quo. They got away with it, thus far, because the foreign bail-outs by the US tax payer was hidden in obscure, indeed barely comprehensible financial transactions.

But we should not expect the US taxpayer to stand behind foreign subsidiaries of US banks and other systemically important US financial institutions, unless a convincing case can be made that there is a material direct US financial exposure. If all the parent has at stake in its foreign subsidiary is its equity in the subsidiary, the US tax payer will not bail out the foreign subsidiary. If the parent has further exposure, through parent-to-daughter loans, through the parent having invested in securities issued by the subsidiary or as a counterparty, there may come a point at which the financial exposure of the parent becomes sufficiently large to induce a response by the US tax payer. But given the current mood of the tax payer, I don’t expect to see rescues of foreign subsidiaries anytime soon.

The same applies to the UK and to other European countries with banks that have large foreign subsidiaries. I don’t expect the British government to bail out the foreign subsidiaries of RBS, Lloyd’s Group, Barclays or HSBC. I would expect the British government to look seriously at bailing out UK subsidiaries of foreign banks, should these be threatened with insolvency, especially if most of the substantive economic activity of these banks is in the UK. For instance (what follows is a pure hypothetical - I know of nothing that would lead me to question the financial soundness of Abbey), I would not expect the Spanish government ever to bail out Abbey (owned by the Spanish bank Santander), but I would expect the UK Treasury to show an active interest.

Individual parent banks may decide to try to rescue their foreign subsidiaries, even when the subsidiaries in question appear to have gone belly-up by normal commercial standards. An extreme example of this is HSBC - reported to have incurred losses estimated to be somewhere between $30 bn and $62 bn (accounting does not appear to be an exact science) on its US credit card issuer and subprime lender, Household Finance Corporation (HFC), now renamed HSBC Finance, which it acquired at the end of 2002. Even though HSBC announced, in March 2009, that it would shut down the branch network of its HSBC Finance subsidiary in the U.S, it is continuing to operate the HSBC Finance credit card arm.

HSBC could, in the opinion of many observers, have saved itself a bundle by cutting its losses and walking away from HFC when the scale of the subprime debacle became apparent late in 2007. There is a clear clash here between reputation and goodwill considerations on the one hand and throwing good money after bad on the other. This, however, is a matter for the shareholders and other stakeholders of HSBC and its management. It is a corporate governance problem. It is not a matter for the UK tax payers, unless HSBC’s HFC-related losses come to haunt it in the future and drive the parent into the arms of the UK tax payer after all. Should that happen, I hope the British authorities will take to heart all three of these ruminations.