Containment, Not Cure
The international banking regulations known as the Basel II Accord have come in for some stick, given the fallout from the banking crisis of 2008. This is, on the face of it, a bit unfair given that Basel II hasn’t yet been fully implemented in most countries and anyway was designed to try to head off some of the problems that have occurred.
Still, most observers reckon that Basel II wouldn’t have prevented the crisis and the tendency of regulators, like generals, to fight the last war means that proposed changes won’t help. Whatever causes the next crisis it won’t be the same as the last one and while regulators are busily building a Maginot Line to stop one kind of problem they’re unlikely to notice that they’re also incentivising banks to invade Belgium, or at least find a way to go around the new regulations. We need a new kind of regulation, one that recognises we can’t stop the disease, but that it can be contained if we act quickly enough.
Predicting the Future from the Past
Basel II aims to monitor risk using models like Value-at-Risk (VaR), which provides a way of estimating the probability and size of a potential loss in a given period. Individual banks are required to report their daily VaR’s to regulators and need to undergo back-testing of their VaR models to ensure that their actual returns are within scope of what they predicted. Failure to get this right leads to requirements to hold additional capital.
Generally banks hate being made to hold additional capital because it reduces the amount they can lend and therefore reduces their profits or, to put it another way, increases their costs. This is also a drag on economic growth as less lending implies less investment. So banks have every reason to try and avoid Basel II capital penalties. And therein lies a problem, because where there’s an incentive there’s an opportunity for a loophole.
The Scale of the Crisis
Leaving loopholes aside for a moment, the scale of the recent financial crisis underlines the problem with using risk models like VaR baselined against historical data. The figures are genuinely extraordinary. In the 58 years prior to the recent financial crisis the S&P500 had only experienced 4%+ gains on 24 days yet in the six months afterwards there were another 12. Prior to the crunch the index had lost over 4% on only 18 occasions but during it there were no less than 15 further drops of 4% or more. Do we really need any more crises to prove that history is a truly rotten way of assessing risk?
As we’ve seen before, disaster myopia means that we continually discount the probability of rare events to zero. Clearly the probability isn’t zero and that means that historical data doesn’t give us a great insight into what will happen tomorrow when it really matters. The fact that Basel II wasn’t in place mandating the use of VaR didn’t make much difference, because most banks were using it anyway, thus helping to co-ordinate their failures. On top of this backtesting of different VaR models shows significant variation over time in their success rates - some models perform better under normal conditions, others under stressed conditions. So all we have do now is detect when normal turns into stress and we're OK.
Of course, if we knew that we wouldn't need the models. D'oh.
Dealing With The Aftermath
The trouble is that these types of crises are always just around the corner and the emphasis in financial regulation in trying to stop them happening is necessary but nowhere near sufficient. As the researchers here argue, having Basel II with VaR forecasting in place, and all of the other paraphernalia it specifies, wouldn’t have prevented the financial crisis and may have exacerbated it. Whenever regulators put rules in place that imply a cost to the regulated then the latter will be incentivised to find a – legal – way of circumventing them.
No, the real challenge for financial regulators is how they deal with the aftermath of a financial crisis. They can’t stop them happening because they’re behaviourally induced and you can’t regulate human nature. However, they can mitigate them once they’ve occurred – with the right sort of regulation.
Leverage and Capital
Although the latest crisis can be traced to issues of excessive leverage, much of it designed to get around regulatory requirements to hold risk capital, and perverse incentives for executives and managers, it’s been the problems that occurred after the first onset of trouble which have created the wider financial crisis. Oddly enough it was the application of regulatory rules that fuelled these after-the-event problems – financial regulation designed for a placid environment helped to fuel the storm when things got rough.
At issue is the need for banks to hold capital in reserve to offset loans. This is perfectly sensible – all banks need to have a buffer of ready cash or equivalents in case of the need to pay it out. The larger the loan book, the larger the cash reserve needed. Unfortunately when banks suffer losses this eats into their capital reserves. At this point they have two options. Firstly they can raise more capital through equity markets. This, unfortunately, is a difficult and expensive option when they’re making losses and can further undermine confidence – the last thing a struggling bank needs.
The second option is that they can reduce their lending base – or sell risky assets, which amounts to the same thing. In the short-term wake of a financial crisis this is usually the option taken, but it has two significant and very nasty knock-on effects.
Exporting Externalities
The first of these is termed the fire-sale externality. In desperation a bank sells its risky assets to raise capital and manages to stabilise its financial position. However, this has a knock-on effect because the effect of the fire-sale is to reduce the value of the assets being sold, which will reduce the value of other banks' assets, placing their capital position in jeopardy. A domino effect can then ensue.
The second knock-on effect is termed the credit crunch externality. If the bank needs to shrink its capital requirements it may do so by reducing lending to parties far removed from the initial problems. So, for example, small business lending may be restricted. Well managed businesses may be forced to forego viable projects or even fail entirely for lack of funding.
In such ways the bank’s problems are exported to innocent third-parties and the costs of this behaviour, although directly traceable to the bank’s failures, are borne by these external entities. Hence, why they’re called “externalities”. For a wider, and thoughtful discussion of these issues, take a look at Rethinking Capital Regulation, by Anil Kashyap, Raghuram Rajan and Jeremy Stein: it's a lucid and wry analysis of the issues before, during and after the crunch of 2008.
Managing The Market Failure
It’s fairly easy to see that from the point of view of the entire market and, indeed, the wider economy the best thing for the bank to do when faced with this kind of crisis is to recapitalise – to go to equity markets and raise fresh capital, diluting existing shareholders. This prevents the crisis from being externalised and ensures that the costs of the problem falls where it should – on the shareholders who’ve failed to keep a handle on the mismanagement of their company by its executives. However, this comes with its own problems, due to reputational risks and the real possibility of a run on a compromised bank.
Unfortunately this type of market failure is one that classical financial regulation quite fails to address – and indeed, it even encourages it, by focussing almost exclusively on ensuring that banks have sufficient capital to remain solvent while ignoring the wider agency costs of how they go about doing this. Alongside this is the problem that any capital regulatory scheme promotes incentives to find ways around it. For example Collateralised Debt Obligations were successful precisely because they enabled banks to avoid onerous capital requirements: the low cost of capital ensured that they were excessively profitable which, in turn, encouraged executives to push this particular boat out until it pretty much sunk the world economic system.
Regulate for Containment
Modifications to Basel II are being made and these are doubtless needed to reflect the learnings of the latest debacle: tools like VaR have their place, as long as we don't end up relying on them unthinkingly like many senior bank executives did prior to the last crisis. However, significantly increasing the capital requirements on banks would have the effect of making the current problems worse, draining much needed liquidity from the global economy.
More is needed and it’s needed in different places. History tells us, beyond all reasonable doubt, that there will be more financial crises afflicting banks in the future. It also tells us that we don’t have a clue how or why these disasters will occur other than it will come through some unexpected combination of financial innovation and human behavioural bias.
But when the next crisis does happen we do know what banks will do. They’ll sell assets and shrink lending and export the crisis to other parties as fast as they can, generally acting with less social responsibility than a dodgy Torquay hotelier. This is where the regulators and legislators need to look: force banks to pay the price of their own ineptitude and make sure they can’t easily externalise their problems. Next time let’s use financial regulation to contain the crisis, not fan the flames: let's aim to fight the next war, not the last.
Related Articles: Quibbles With Quants, It's Not Different This Time, The Business of Capital is Bust
The international banking regulations known as the Basel II Accord have come in for some stick, given the fallout from the banking crisis of 2008. This is, on the face of it, a bit unfair given that Basel II hasn’t yet been fully implemented in most countries and anyway was designed to try to head off some of the problems that have occurred.
Still, most observers reckon that Basel II wouldn’t have prevented the crisis and the tendency of regulators, like generals, to fight the last war means that proposed changes won’t help. Whatever causes the next crisis it won’t be the same as the last one and while regulators are busily building a Maginot Line to stop one kind of problem they’re unlikely to notice that they’re also incentivising banks to invade Belgium, or at least find a way to go around the new regulations. We need a new kind of regulation, one that recognises we can’t stop the disease, but that it can be contained if we act quickly enough.
Predicting the Future from the Past
Basel II aims to monitor risk using models like Value-at-Risk (VaR), which provides a way of estimating the probability and size of a potential loss in a given period. Individual banks are required to report their daily VaR’s to regulators and need to undergo back-testing of their VaR models to ensure that their actual returns are within scope of what they predicted. Failure to get this right leads to requirements to hold additional capital.
Generally banks hate being made to hold additional capital because it reduces the amount they can lend and therefore reduces their profits or, to put it another way, increases their costs. This is also a drag on economic growth as less lending implies less investment. So banks have every reason to try and avoid Basel II capital penalties. And therein lies a problem, because where there’s an incentive there’s an opportunity for a loophole.
The Scale of the Crisis
Leaving loopholes aside for a moment, the scale of the recent financial crisis underlines the problem with using risk models like VaR baselined against historical data. The figures are genuinely extraordinary. In the 58 years prior to the recent financial crisis the S&P500 had only experienced 4%+ gains on 24 days yet in the six months afterwards there were another 12. Prior to the crunch the index had lost over 4% on only 18 occasions but during it there were no less than 15 further drops of 4% or more. Do we really need any more crises to prove that history is a truly rotten way of assessing risk?
As we’ve seen before, disaster myopia means that we continually discount the probability of rare events to zero. Clearly the probability isn’t zero and that means that historical data doesn’t give us a great insight into what will happen tomorrow when it really matters. The fact that Basel II wasn’t in place mandating the use of VaR didn’t make much difference, because most banks were using it anyway, thus helping to co-ordinate their failures. On top of this backtesting of different VaR models shows significant variation over time in their success rates - some models perform better under normal conditions, others under stressed conditions. So all we have do now is detect when normal turns into stress and we're OK.
Of course, if we knew that we wouldn't need the models. D'oh.
Dealing With The Aftermath
The trouble is that these types of crises are always just around the corner and the emphasis in financial regulation in trying to stop them happening is necessary but nowhere near sufficient. As the researchers here argue, having Basel II with VaR forecasting in place, and all of the other paraphernalia it specifies, wouldn’t have prevented the financial crisis and may have exacerbated it. Whenever regulators put rules in place that imply a cost to the regulated then the latter will be incentivised to find a – legal – way of circumventing them.
No, the real challenge for financial regulators is how they deal with the aftermath of a financial crisis. They can’t stop them happening because they’re behaviourally induced and you can’t regulate human nature. However, they can mitigate them once they’ve occurred – with the right sort of regulation.
Leverage and Capital
Although the latest crisis can be traced to issues of excessive leverage, much of it designed to get around regulatory requirements to hold risk capital, and perverse incentives for executives and managers, it’s been the problems that occurred after the first onset of trouble which have created the wider financial crisis. Oddly enough it was the application of regulatory rules that fuelled these after-the-event problems – financial regulation designed for a placid environment helped to fuel the storm when things got rough.
At issue is the need for banks to hold capital in reserve to offset loans. This is perfectly sensible – all banks need to have a buffer of ready cash or equivalents in case of the need to pay it out. The larger the loan book, the larger the cash reserve needed. Unfortunately when banks suffer losses this eats into their capital reserves. At this point they have two options. Firstly they can raise more capital through equity markets. This, unfortunately, is a difficult and expensive option when they’re making losses and can further undermine confidence – the last thing a struggling bank needs.
The second option is that they can reduce their lending base – or sell risky assets, which amounts to the same thing. In the short-term wake of a financial crisis this is usually the option taken, but it has two significant and very nasty knock-on effects.
Exporting Externalities
The first of these is termed the fire-sale externality. In desperation a bank sells its risky assets to raise capital and manages to stabilise its financial position. However, this has a knock-on effect because the effect of the fire-sale is to reduce the value of the assets being sold, which will reduce the value of other banks' assets, placing their capital position in jeopardy. A domino effect can then ensue.
The second knock-on effect is termed the credit crunch externality. If the bank needs to shrink its capital requirements it may do so by reducing lending to parties far removed from the initial problems. So, for example, small business lending may be restricted. Well managed businesses may be forced to forego viable projects or even fail entirely for lack of funding.
In such ways the bank’s problems are exported to innocent third-parties and the costs of this behaviour, although directly traceable to the bank’s failures, are borne by these external entities. Hence, why they’re called “externalities”. For a wider, and thoughtful discussion of these issues, take a look at Rethinking Capital Regulation, by Anil Kashyap, Raghuram Rajan and Jeremy Stein: it's a lucid and wry analysis of the issues before, during and after the crunch of 2008.
Managing The Market Failure
It’s fairly easy to see that from the point of view of the entire market and, indeed, the wider economy the best thing for the bank to do when faced with this kind of crisis is to recapitalise – to go to equity markets and raise fresh capital, diluting existing shareholders. This prevents the crisis from being externalised and ensures that the costs of the problem falls where it should – on the shareholders who’ve failed to keep a handle on the mismanagement of their company by its executives. However, this comes with its own problems, due to reputational risks and the real possibility of a run on a compromised bank.
Unfortunately this type of market failure is one that classical financial regulation quite fails to address – and indeed, it even encourages it, by focussing almost exclusively on ensuring that banks have sufficient capital to remain solvent while ignoring the wider agency costs of how they go about doing this. Alongside this is the problem that any capital regulatory scheme promotes incentives to find ways around it. For example Collateralised Debt Obligations were successful precisely because they enabled banks to avoid onerous capital requirements: the low cost of capital ensured that they were excessively profitable which, in turn, encouraged executives to push this particular boat out until it pretty much sunk the world economic system.
Regulate for Containment
Modifications to Basel II are being made and these are doubtless needed to reflect the learnings of the latest debacle: tools like VaR have their place, as long as we don't end up relying on them unthinkingly like many senior bank executives did prior to the last crisis. However, significantly increasing the capital requirements on banks would have the effect of making the current problems worse, draining much needed liquidity from the global economy.
More is needed and it’s needed in different places. History tells us, beyond all reasonable doubt, that there will be more financial crises afflicting banks in the future. It also tells us that we don’t have a clue how or why these disasters will occur other than it will come through some unexpected combination of financial innovation and human behavioural bias.
But when the next crisis does happen we do know what banks will do. They’ll sell assets and shrink lending and export the crisis to other parties as fast as they can, generally acting with less social responsibility than a dodgy Torquay hotelier. This is where the regulators and legislators need to look: force banks to pay the price of their own ineptitude and make sure they can’t easily externalise their problems. Next time let’s use financial regulation to contain the crisis, not fan the flames: let's aim to fight the next war, not the last.
Related Articles: Quibbles With Quants, It's Not Different This Time, The Business of Capital is Bust
Bank regulation is always doomed to failure because the cost to society of the ultimate penalty, failure of the banking system, is too high. Governments will always blink first.
ReplyDeleteMore useful is how to manage the conflict between agents and principals. In this case the agents are bank managers and prop traders who have every incentive to use the bank's balance sheet to the maximum for their own personal gain.
The principals are the bank's shareholders who want to maximise their returns, but not at the risk of blowing up the bank.
The problem now is that shareholders have virtually no say in how their capital is used. Institutional shareholders and especially tracker funds are notable by their absence on the boards of these banks.
These representatives of the shareholders should provide the bulk of the non-executive directors of the banks. It is their self-interest that is surely the best defence against the agents using the their capital dangerously.
Interesting post. Like Kayshap, Rajam & Stein I share their pessimism about getting balanced adaptive regulation although making explicit plans for what is effectively Catastrophe Insurance underwritten by Government makes sense. In my view Basel II was far too complex, over ambitious, took too long to implement and provided banks effectively with a running commentary during the consultation process. If you are to contain anti-social action it is hardly effective if you publish/debate in public the proposed defintions and possible consequences years in advance. Moreover the attempts at defining capital weightings for all/most asset classes encouraged gaming on an industrial scale. The authors don't really look into the debt/ equity raising decisions but do point out the so-called agency problems. These are no different for non-banks, over mighty CEOs, supine boards, lazy and indifferent shareholders. However there are some built-in behaviousr that have been around for years and should have been red flags. e.g. When UK building societies demutualised, they began to particpate in totally non-standard risks, the Halifax, later HBOS, led a major aircraft leasing deal-why? Equally smaller ex mutuals strayed into US housing via US RMBS and CDOs, what does a relatively small reagional bank whose business was historically based in Yorkshire know about the US housing market? How can it risk assess its particpation or should we not really state the obvious this was just blind yiled chasing. For years the IBs and major commercial banks have originated and distributed largely to what were unglamorously known to all as " stuffee" banks and investors. In short they lacked origination capacity and would buy up assets for yiled without a clue as to the real risk profile. This behaviour has been around at least 30 years, why did boards and institutional investors, let alone smarter retail ones not question this behaviour? This is simple enough stuff.
ReplyDelete