This post contains the comments I made at the 6th Annual Conference: Emerging from the Financial Crisis, held at the Center on Capitalism and Society at Columbia University, on February 20. It contains 22 points, two for each of the 12 disciples, minus Judas, who is otherwise engaged, buying silver futures.
1. It is necessary, for political economy reasons, to rush new comprehensive regulation of the financial sector. While it would be better, holding constant the likelihood of the measures being adopted and implemented, not to act in haste, there is now a unique window of opportunity - a period of extraordinary politics, in the words of Balcerowicz - to actually get the thorough regulatory reform we need. The reason is that the private financial sector is on its uppers - down and out - and will not be able to put together much of a fight, let alone its usual boom-time massive lobbying effort to veto radical measures. It is better to over-regulate now and subsequently to correct the mistakes than to risk another era of self-regulation and soft-touch under-regulation of financial markets, instruments and institutions.
2. The objective of macro-prudential regulation is systemic financial stability, which has a number of dimensions:
a. Preventing or mitigating asset market and credit booms, bubbles and busts.
b. Preventing or mitigating market illiquidity in systemically important markets.
c. Preventing or mitigating funding illiquidity for systemically important financial institutions.
d. Prreventing or managing insolvencies of systemically important financial institutions.
Other micro-prudential considerations (abuse of monopoly power; consumer protection; micro-manifestations of asymmetric information) should be left to the micro-prudential regulator(s).
3. Regulation will have to be comprehensive across instruments, institutions, markets and countries. Specifically, we must:
a. Regulate all systemically important highly leveraged financial enterprises, whatever they call themselves: commercial bank, investment bank, universal bank, hedge fund, SIV, CDO, private equity fund or bicycle repair shop.
b. Regulate all markets for systemically important financial instruments.
c. Regulate all systemically important financial infrastructure or plumbing: payment, clearing, settlement systems, mechanisms and platforms, and the associated provision of custodial services.
d. Do all of (a), (b) and (c) on a cross-border basis.
4. Self-regulation stands to regulation the same way self-importance stands to importance. The notion that markets, including financial markets could be self-regulating, by properly incentivising CEOs and Boards of Directors, and through market-discipline is prima facie suspect. We decide to regulate markets because of market failure. Then we let the market regulate the market. This is an invisible hand too far. The concept of self-regulation is especially ludicrous for financial markets. Finance is trade in promises expressed in units of abstract purchasing power (money). It scales up and down ferociously quickly. If Airbus or Boeing wishes to double the size of its operations, it takes 4 or 5 years to put in place another set of assembly lines. If a bank wishes to scale its balance sheet and operations ten-fold, all it has to do is to add a zero in the right places. Given enough optimism, trust, confidence and self-confidence, financial activity can, through leverage, be scaled up alarmingly quickly. Once optimism, trust, confidence and self-confidence disappear and are replaced by pessimism, mistrust, lack of confidence and fear/panic, the scaling down of bank activities can occur even faster. Such an industry cannot be left to its own devices.
5. Regulation can only take place on the basis of independently verifiable (public) information. Regulators cannot rely on information that is private to the regulated entity. This means that the capital adequacy of the first pillar of Basel II has to be overhauled radically, as it relies for the risk-weighting of assets in part on internal bank models that are private to the banks.
6. Macro-prudential regulation, supervision and intervention requires the close co-operation of the central bank, the supervisor/regulator and the Treasury (fiscal authority). To fulfil its macro-prudential role, the central bank requires the fiscal back-up of the Treasury (ministry of finance). This is because by accepting private securities as collateral in its open market operations and at its discount window, and through the outright purchases of private securities by the central bank associated with qualitative easing or credit easing, the central bank takes private credit risk on its balance sheet. To do this without becoming and active fiscal actor in its own right, the central bank will require a Treasury indemnity (guarantee) for the full amount of its private sector exposure. The fact that the Fed only has a Treasury indemnity for ten percent of its maximal exposure under the TALF, means that the Fed is now a fiscal policy maker. This compromises its independence.
7. Since central banks, supervision/regulation and taxation are almost everywhere national, there is an inherent tension with global financial markets and border-crossing financial institutions. We can expect to see the following:
a. There will be a repatriation of border-crossing banks and ofis, partly driven by market forces and partly by regulatory and other official interventions. This cross-border deleveraging can already be seen at work, especially where banks get tax payer support (at the national level) and commit themselves in return to lend to domestic households and enterprises.
b. Cross-border branches (supervised and regulated by home country (the country of the parent bank), with lender of last resort (LLR) and market maker of last resort (MMLR) access (if any) only through the parent bank in the home country, and with fiscal back-up (if any) only through the home country) will be a thing of the past.
c. Even as regards subsidiaries, there will only independently capitalised, ring-fenced, autonomous cross-border subsidiaries, with regulation and supervision by host country, with LLR and MMLR support (if any)from the host country central bank and fiscal support (if any) from the host country Treasury.
8. There are special problems for Eurozone:
a. There is one central bank (ECB/Eurosystem) for 16 countries
b. Supervision and regulation remain at the national level
c. There is no fiscal Europe/or fiscal Eurozone: no single supranational entity with tax and borrowing powers to back-up and if necessary recapitalise the ECB/Eurosystem, when it suffers losses as a result of Eurozone-wide monetary policy operations, liquidity operations and credit easing operations.
9. There is a clear need for a fiscal Europe to back up the ECB and to permit it to engage in qualitative easing/credit easing. This could come in one of three forms. In increasing order of likelihood these are:
a. A supranational Eurozone-wide tax and borrowing authority, specifically dedicated to fiscal backing for the ECB/Eurosystem.
b. A Eurozone-wide fund, funded initially by the 16 Eurozone governments (in proportion, say, to their relative shares in the ECB’s capital, that the ECB/Eurosystem could draw on (subject to qualified majority support in the Eurogroup) if it were to suffer losses as a result of Eurozone-wide monetary policy operations, liquidity operations and credit easing operations.
c. An ad-hoc, hastily cobbled together fiscal burden sharing rule for the 16 Eurozone national governments, to restore the capital adequacy of the ECB/Eurosystem.
10. There is a need for Eurozone-wide and preferably for an EU-wide regulator/supervisor for border-crossing financial institutions (mainly bank-like entities and insurance companies). Colleges of national supervisors/regulators just won’t do. Eurozone-wide or, preferably EU-wide regulation with national supervision and enforcement is a poor second-best, because countries have been shown to cheat as regards the interpretation of EU-wide rules and regulations.
11. There is the need and opportunity to close down all tax havens and regulatory havens. Tax havens are defined as countries that have bank secrecy, which includes Switzerland, Austria and Luxembourg as well as the usual micro-state suspects (bank secrecy or bank privacy is the legal principle according to which banks can protect personal information about their customers, even from the tax authorities and police authorities of these customers). The anonymity provided by bank secrecy promotes tax evasion, tax avoidance (or fraud), money laundering and hiding the proceeds of criminal activity. Regulatory havens are nations that offer companies the opportunity to avoid global standards for reporting, governance, auditing, transparency, openness etc. Tax havens and regulatory havens are key elements in the global regulatory and tax arbitrage games that have undermined government revenue bases and weakened global regulatory standards.
The means to put tax havens out of business are simple: forbid banks, other financial institutions and private persons from doing business with and engaging in transactions with banks and other financial institutions located in countries that have bank secrecy. To take care of regulatory havens, don’t recognise and enforce contracts drawn up under their laws and do not recognise court judgements originating from tax havens.
12. Financial innovation in products and institutions is potentially beneficial and potentially harmful. There is a need to regulate financial innovation. I propose the model used in the US by the Food and Drug Administration for pharmaceutical and medical products.
a. First, there is a positive list of financial instruments and institutions. Anything that is not explicitly allowed is forbidden.
b. To get a new instrument or new institution approved, there will have to be testing, scrutiny by regulators, supervisors, academic specialists and other interested parties, and pilot projects. It is possible that, once a new instrument or institution has been approved, it is only available ‘with a prescription’. For instance, only professional counterparties rather than the general public could be permitted.
c. Clearly, this approach to financial innovation would slow down financial innovation. It may even kill off certain innovations that would have been socially useful. So be it. The dangers of unbridled financial innovation are too manifest.
13. Rating agencies
a. Rating agencies should be turned into single-activity or single-product line firms. They should provide just ratings, not any other products or services, including advice. The conflict of interest in combining rating activities with advisory services or the sale of other lucrative services or products to customers looking for the best possible rating is obvious and inescapable. Chinese walls don’t work and are aptly named. The Great Wall of China did not keep the barbarians out or the Han Chinese in.
b. The quasi-regulatory role of the rating agencies in Basel II should be eliminated.
c. The customer wishing to have his company, country or instrument rated does not pay the regulator, ex-ante or ex-post. Instead he pays the regulator, who then allocates/auctions the individual rating activity among the population of competing rating agencies.
d. Rating agencies should be paid in part in the securities they are rating. Such securities should be held to maturity and cannot be hedged by the rating agency.
14. Securitisation of transparent, relatively homogeneous assets or cash flows is a useful invention, but only if the originator of the underlying assets/cash flows is required to retain a material chunk of the first-loss or equity tranche.
15. Counter-cyclical capital and liquidity requirements or leverage ratios should be implemented, based either on the balance sheet characteristics of individual banks (e.g. the proposals by Goodhart and Persaud) or based on the behaviour of economy-wide aggregates (GDP, credit growth, balance sheet growth, asset prices) in the host country. These requirements should apply to all highly leveraged institutions deemed systemically important.
16. Capital and liquidity requirements/ratios should not be floors, but buffers. Requirements should be countercyclical. Violations punished according to a steep ladder of penalties.
17. The distinction between public utility banking/narrow banking vs. investment banking: (the rest) has to be re-introduced. I advocate a form of Glass-Steagall on steroids, with a heavily regulated and closely supervised narrow banking sector, engaged in commercial banking (taking deposits and making loans) and benefiting from LLR and MMLR support. The investment bank sector will also be regulated and supervised, but more lightly, and according to the same principles as other systemically important highly leveraged non-narrow bank institutions.
Universal banking has few if any efficiency advantages and many disadvantages. Economies of scale and scope and banking are soon exhausted. They tend to be fat to fail, have a lack of focus, and suffer from span-of-control negative synergies etc. Universal banks or financial supermarkets use their size to exploit market power and try to shelter their risky, non-narrow banking activities under the LLR and MMLR umbrella of the narrow bank that’s hiding somewhere inside the universal bank.
18. Splitting banks into public utility or narrow banks does not solve the problems of banks (narrow or investment) becoming too big or too interconnected to fail. It is therefore necessary to penalise bank size per se, to stop banks from becoming too large to fail (if they are interconnected but small, they are still not systemically important). I would penalise size through capital requirements that are progressive in size (as well as leverage).
19. The problem of the ‘swarming’ of individually not systemically important institutions that collectively achieve critical mass as regards systemic stability, has been analysed by Tobias Adrian and Markus K. Brunnermeier and in the recent Geneva Report by Markus Brunnermeier, Andrew Crocket, Charles Goodhart, Avinash D. Persaud and Hyun Shin.They propose a measure of an institution’s contribution to systemic risk which they call CoVaR . It is the value at risk (VaR) of a financial institution conditional on other institutions being in distress. The increase of in an institution’s CoVaR relative to its VaR is supposed to measure spillover risk among institutions.
There are clearly deep conceptual problems with this measure. It jumps from correlation to ‘spillover’ - which is effectively causation - without much thought about the difference between the two. And, like VaR, CoVaR is likely to be very different under conditions of extreme crisis than under the conditions prevailing on average during the sample that provided the data used to calculate the CoVar. But the problem of individually systemically insignificant institutions can collectively achieve critical mass is one that has to be addressed by the regulator and supervisor if there are large numbers of small institutions.
20. There is no substitute for fair value accounting, or even mark-to-market accounting. Companies must be require to the strictest version of it for the inflation it requires. Even illiquid market prices are better than managerial discretion. Companies should not be able to move assets between the trading book, the banking book and the available for sale book, unless the institution is in administration or in a special resolution regime. Regulatory forbearance should be used to mitigate the consequences of mark-to-market pricing when asset prices are depressed because of market illiquidity.
21. Remuneration is a matter of corporate governance. The incentive structure created by compensation packages does, however, influence the risk-profile of a bank or other financial institution. It should be taken into account by the regulator/supervisor in determining capital requirements and liquidity requirements, assuming the regulator/supervisor understands the impact of the remuneration structure on bank risk taking.
22. Given the manifest failure of the efficient market hypothesis, it is not at all obvious that systemically important financial institutions should be allowed to be listed companies. Financial institutions’ stock market valuations have been notorious will-o’-the wisps and have, through stock options and other stock-market valuation-related executive remuneration components, contributed to the excessive risk taking during the recent boom. Partnerships, mutual ownership, cooperative ownership and various forms of public and mixed public-private ownership may be more appropriate for financial institutions. Perhaps we should even consider removing limited liability for investment banks!
Willem Buiter
London School of Economics and Political Science, University of Amsterdam, and Centre for Economic Policy Research
Regulate the new financial sector now
Posted by Willem Buiter on 26 February 2009
This post contains the comments I made at the 6th Annual Conference: Emerging from the Financial Crisis, held at the Center on Capitalism and Society at Columbia University, on February 20. It contains 22 points, two for each of the 12 disciples, minus Judas, who is otherwise engaged, buying silver futures.
1. It is necessary, for political economy reasons, to rush new comprehensive regulation of the financial sector. While it would be better, holding constant the likelihood of the measures being adopted and implemented, not to act in haste, there is now a unique window of opportunity - a period of extraordinary politics, in the words of Balcerowicz - to actually get the thorough regulatory reform we need. The reason is that the private financial sector is on its uppers - down and out - and will not be able to put together much of a fight, let alone its usual boom-time massive lobbying effort to veto radical measures. It is better to over-regulate now and subsequently to correct the mistakes than to risk another era of self-regulation and soft-touch under-regulation of financial markets, instruments and institutions.
Other micro-prudential considerations (abuse of monopoly power; consumer protection; micro-manifestations of asymmetric information) should be left to the micro-prudential regulator(s).
c. Regulate all systemically important financial infrastructure or plumbing: payment, clearing, settlement systems, mechanisms and platforms, and the associated provision of custodial services.
d. Do all of (a), (b) and (c) on a cross-border basis.
4. Self-regulation stands to regulation the same way self-importance stands to importance. The notion that markets, including financial markets could be self-regulating, by properly incentivising CEOs and Boards of Directors, and through market-discipline is prima facie suspect. We decide to regulate markets because of market failure. Then we let the market regulate the market. This is an invisible hand too far. The concept of self-regulation is especially ludicrous for financial markets. Finance is trade in promises expressed in units of abstract purchasing power (money). It scales up and down ferociously quickly. If Airbus or Boeing wishes to double the size of its operations, it takes 4 or 5 years to put in place another set of assembly lines. If a bank wishes to scale its balance sheet and operations ten-fold, all it has to do is to add a zero in the right places. Given enough optimism, trust, confidence and self-confidence, financial activity can, through leverage, be scaled up alarmingly quickly. Once optimism, trust, confidence and self-confidence disappear and are replaced by pessimism, mistrust, lack of confidence and fear/panic, the scaling down of bank activities can occur even faster. Such an industry cannot be left to its own devices.
5. Regulation can only take place on the basis of independently verifiable (public) information. Regulators cannot rely on information that is private to the regulated entity. This means that the capital adequacy of the first pillar of Basel II has to be overhauled radically, as it relies for the risk-weighting of assets in part on internal bank models that are private to the banks.
6. Macro-prudential regulation, supervision and intervention requires the close co-operation of the central bank, the supervisor/regulator and the Treasury (fiscal authority). To fulfil its macro-prudential role, the central bank requires the fiscal back-up of the Treasury (ministry of finance). This is because by accepting private securities as collateral in its open market operations and at its discount window, and through the outright purchases of private securities by the central bank associated with qualitative easing or credit easing, the central bank takes private credit risk on its balance sheet. To do this without becoming and active fiscal actor in its own right, the central bank will require a Treasury indemnity (guarantee) for the full amount of its private sector exposure. The fact that the Fed only has a Treasury indemnity for ten percent of its maximal exposure under the TALF, means that the Fed is now a fiscal policy maker. This compromises its independence.
c. Even as regards subsidiaries, there will only independently capitalised, ring-fenced, autonomous cross-border subsidiaries, with regulation and supervision by host country, with LLR and MMLR support (if any)from the host country central bank and fiscal support (if any) from the host country Treasury.
c. There is no fiscal Europe/or fiscal Eurozone: no single supranational entity with tax and borrowing powers to back-up and if necessary recapitalise the ECB/Eurosystem, when it suffers losses as a result of Eurozone-wide monetary policy operations, liquidity operations and credit easing operations.
c. An ad-hoc, hastily cobbled together fiscal burden sharing rule for the 16 Eurozone national governments, to restore the capital adequacy of the ECB/Eurosystem.
10. There is a need for Eurozone-wide and preferably for an EU-wide regulator/supervisor for border-crossing financial institutions (mainly bank-like entities and insurance companies). Colleges of national supervisors/regulators just won’t do. Eurozone-wide or, preferably EU-wide regulation with national supervision and enforcement is a poor second-best, because countries have been shown to cheat as regards the interpretation of EU-wide rules and regulations.
11. There is the need and opportunity to close down all tax havens and regulatory havens. Tax havens are defined as countries that have bank secrecy, which includes Switzerland, Austria and Luxembourg as well as the usual micro-state suspects (bank secrecy or bank privacy is the legal principle according to which banks can protect personal information about their customers, even from the tax authorities and police authorities of these customers). The anonymity provided by bank secrecy promotes tax evasion, tax avoidance (or fraud), money laundering and hiding the proceeds of criminal activity. Regulatory havens are nations that offer companies the opportunity to avoid global standards for reporting, governance, auditing, transparency, openness etc. Tax havens and regulatory havens are key elements in the global regulatory and tax arbitrage games that have undermined government revenue bases and weakened global regulatory standards.
The means to put tax havens out of business are simple: forbid banks, other financial institutions and private persons from doing business with and engaging in transactions with banks and other financial institutions located in countries that have bank secrecy. To take care of regulatory havens, don’t recognise and enforce contracts drawn up under their laws and do not recognise court judgements originating from tax havens.
c. Clearly, this approach to financial innovation would slow down financial innovation. It may even kill off certain innovations that would have been socially useful. So be it. The dangers of unbridled financial innovation are too manifest.
d. Rating agencies should be paid in part in the securities they are rating. Such securities should be held to maturity and cannot be hedged by the rating agency.
15. Counter-cyclical capital and liquidity requirements or leverage ratios should be implemented, based either on the balance sheet characteristics of individual banks (e.g. the proposals by Goodhart and Persaud) or based on the behaviour of economy-wide aggregates (GDP, credit growth, balance sheet growth, asset prices) in the host country. These requirements should apply to all highly leveraged institutions deemed systemically important.
16. Capital and liquidity requirements/ratios should not be floors, but buffers. Requirements should be countercyclical. Violations punished according to a steep ladder of penalties.
Universal banking has few if any efficiency advantages and many disadvantages. Economies of scale and scope and banking are soon exhausted. They tend to be fat to fail, have a lack of focus, and suffer from span-of-control negative synergies etc. Universal banks or financial supermarkets use their size to exploit market power and try to shelter their risky, non-narrow banking activities under the LLR and MMLR umbrella of the narrow bank that’s hiding somewhere inside the universal bank.
18. Splitting banks into public utility or narrow banks does not solve the problems of banks (narrow or investment) becoming too big or too interconnected to fail. It is therefore necessary to penalise bank size per se, to stop banks from becoming too large to fail (if they are interconnected but small, they are still not systemically important). I would penalise size through capital requirements that are progressive in size (as well as leverage).
19. The problem of the ‘swarming’ of individually not systemically important institutions that collectively achieve critical mass as regards systemic stability, has been analysed by Tobias Adrian and Markus K. Brunnermeier and in the recent Geneva Report by Markus Brunnermeier, Andrew Crocket, Charles Goodhart, Avinash D. Persaud and Hyun Shin.They propose a measure of an institution’s contribution to systemic risk which they call CoVaR . It is the value at risk (VaR) of a financial institution conditional on other institutions being in distress. The increase of in an institution’s CoVaR relative to its VaR is supposed to measure spillover risk among institutions.
There are clearly deep conceptual problems with this measure. It jumps from correlation to ‘spillover’ - which is effectively causation - without much thought about the difference between the two. And, like VaR, CoVaR is likely to be very different under conditions of extreme crisis than under the conditions prevailing on average during the sample that provided the data used to calculate the CoVar. But the problem of individually systemically insignificant institutions can collectively achieve critical mass is one that has to be addressed by the regulator and supervisor if there are large numbers of small institutions.
20. There is no substitute for fair value accounting, or even mark-to-market accounting. Companies must be require to the strictest version of it for the inflation it requires. Even illiquid market prices are better than managerial discretion. Companies should not be able to move assets between the trading book, the banking book and the available for sale book, unless the institution is in administration or in a special resolution regime. Regulatory forbearance should be used to mitigate the consequences of mark-to-market pricing when asset prices are depressed because of market illiquidity.
21. Remuneration is a matter of corporate governance. The incentive structure created by compensation packages does, however, influence the risk-profile of a bank or other financial institution. It should be taken into account by the regulator/supervisor in determining capital requirements and liquidity requirements, assuming the regulator/supervisor understands the impact of the remuneration structure on bank risk taking.
22. Given the manifest failure of the efficient market hypothesis, it is not at all obvious that systemically important financial institutions should be allowed to be listed companies. Financial institutions’ stock market valuations have been notorious will-o’-the wisps and have, through stock options and other stock-market valuation-related executive remuneration components, contributed to the excessive risk taking during the recent boom. Partnerships, mutual ownership, cooperative ownership and various forms of public and mixed public-private ownership may be more appropriate for financial institutions. Perhaps we should even consider removing limited liability for investment banks!
Willem Buiter
London School of Economics and Political Science, University of Amsterdam, and Centre for Economic Policy Research
Editors’ Note: This first appeared on Willem Buiter’s blog Maverecon, http://blogs.ft.com/maverecon/2009/02/regulating-the-new-financial-sector/. Reposted with permission.