01 Feb 2008, Risk magazine
Criticism of the Basel II capital rules is nothing new, but most of the flak has been directed at technical wrinkles and methodological nuances. The liquidity crisis has changed that. Now, for the first time since it was proposed in 2001, the bedrock of the new Accord is being attacked. Critics charge that Basel II takes aim at the wrong targets, that it is too prescriptive, that it will do nothing to prevent bubbles inflating and little to make the financial system more resilient. Even before the new rules have been fully implemented, officials have already started calling for them to be reviewed.
"There's nothing like a crisis to make people wonder if what they're doing is necessarily the right thing," says Charles Goodhart, director of the regulation and financial stability programme at the London School of Economics (LSE), and a former member of the Bank of England's rate-setting committee.
The criticisms and questions vary in scope. In some quarters, the Basel project is being attacked for encouraging the disintermediation of the banking industry - the dispersion of risk from banks to hedge funds, insurers, pension funds and other investors - which some claim was the main enabling factor in the crisis. "Regulatory capital encouraged banks to shift risk to other parts of the financial system. Bank regulators patted themselves on the back and said "Look - we've made banks safer". But they've only done so by making other parts of the system weaker," argues Avinash Persaud, chairman of London-based liquidity advisory firm Intelligence Capital (see box, page 24).
Regulators see this as grossly unfair. Klaas Knot, director of supervisory policy at De Nederlandsche Bank in Amsterdam and chair of the Basel Committee's risk management and modelling group, says it is right to ask whether lessons can be learned from the crisis, but Basel II should be given a chance to work before being written off. "We should strive to keep improving - that sense is always there at the Basel Committee. But this crisis took place under the old Accord. So let's not rush things and blame the new Accord, which we strongly believe will make banks more resilient to these types of shock in future and will also help strengthen supervision," he says.
For some, this plea comes too late. Andrew Kuritzkes, a managing director with Oliver Wyman in New York, has long had reservations about Basel II. The regulation is too complex and too prescriptive, he says: "The tremendous effort required in Basel compliance led to things like asset/liability risk, liquidity risk and business risk being crowded out. Given a bit more freedom, I'd argue that risk managers would have been more focused on risks outside the Basel II box and would have been better able to anticipate the kind of events that played out from July onwards."
Regulators counter that, had Basel II been in place before the liquidity crisis hit, its worst effects would never have been felt. For example, the new rules levy a capital charge on the liquidity guarantees some banks offered to off-balance-sheet entities. Under the old regime, there was no such charge, so banks deluded themselves that there was no risk associated with the commitment. In the case of Germany's Sachsen LB, that delusion allowed commitments to pile up far in excess of its capacity to meet them.
"Under Basel II, even if the lines are undrawn, you still have to hold capital against them. That makes you much more aware that you have this contingent liability out there and makes it more likely that there would be an early intervention," says De Nederlandsche Bank's Knot.
In addition, the new Accord will ensure more consistent treatment of credit risk between the banking book and the trading book, he says. It will also impose tougher disclosure requirements, which Knot argues would have helped forestall the panic that left conduits and structured investment vehicles unable to roll over their short-term borrowing in July and August: "Under Basel II, investors would have been able to clearly distinguish between conduits with subprime exposure and those without, and the whole process of price discovery would have been completed much more quickly."
Arguments like these have convinced many within the supervisory community that Basel II needs to be embraced more - not less - enthusiastically than before: the Basel Committee, the Institute for International Finance, and even Jean-Claude Trichet, president of the European Central Bank, have all commented on the importance of the new framework. In the US, meanwhile, regulators have finally approved the new rules - a step that was welcomed by the domestic banking industry, says Knot, who insists criticism of the rules is coming principally from a "small band of outsiders".
But support for the new Accord is also tempered with caution. When the German chancellor, Angela Merkel, and the French president, Nicolas Sarkozy, made a joint call for greater financial market transparency in September, they also set in motion a review at the European Commission (EC) level, led by David Wright, the EC's head of financial services policy. That review is expected to focus on the role that rating agencies and hedge funds played in the crisis, but could also result in some tinkering with Basel II, says the head of securitisation research at one large UK bank: "The EC has been told to work out whether there is a need to make changes to Basel. There is a concern that some parts of the framework might be overly reliant on credit ratings."
Earlier this year, the chair of the European parliament's economic affairs committee, Pervenche Beres, called for Basel II to be reviewed, and the EC's Wright reportedly told a meeting of the European Securitisation Forum on November 19 that some of Basel II's annexes might need to be transposed into EU law in order to boost disclosure.
Intelligence Capital's Persaud has heard all of this before: "At the end of every crisis, people always call for more transparency and more risk management. It may be that those things can sometimes help, but they're tangential to the underlying problems. They're cheap to do and they can sometimes make things worse." If the location of risk concentrations is known to market participants, he argues that investors may be tempted to get the jump on each other by anticipating selling pressure - triggering a run on the market as everyone rushes to dump the same securities at the same time.
Rather than calling for more of the same, Persaud says it is time to consider radical alternatives - and the mood at October's G-7 meeting of finance ministers and central bankers in Washington suggests he is not alone. "Everyone was saying the crisis wasn't unforeseen. Everyone had been worrying about a crisis in the credit markets and the impact it could have on housing, economic growth and so on. Yet, despite being unsurprised, the system was unprepared. So, there's currently a much greater willingness and appetite to discuss more radical ideas," says Persaud.
One of those ideas would be to detach capital requirements from risk levels. Because risk is lowest at the peak of a business cycle, regulatory capital under Basel II should also be low, meaning there would be little constraint on a bank's ability to pile on risk - potentially exacerbating the pain of the next downturn, say critics. Conversely, at the market's nadir, risk would appear to be high and capital requirements would also be high - biting into the industry's ability to lend and prolonging the slump. In short, the rules are seen as being pro-cyclical.
"There's a fundamental weakness in the regulatory framework because it puts no constraints on banks' rate of growth when things are going well, then bites deeply later on, which could mean each boom is followed by a credit crunch - when what you really want is to moderate the boom in advance," says the LSE's Goodhart.
Persaud agrees, and says it is possible to address this problem without dismantling the new rules: "The worst thing about Basel II is that it contains no counter-cyclical measures. The best thing is that there's some room for you to do something about it."
The second of the Accord's three pillars is a get-out clause for supervisors, which enables them to adjust the capital requirements that each bank calculates, he says. Regulators could use Pillar II as a way of de-linking bank capital from rising and falling risk, and instead create a new relationship, linking capital to loan growth, for example. The idea is that, irrespective of what risk models are saying, runaway credit supply would be reined in by mushrooming capital requirements, and the loan market would undulate gently between growth and decline rather than seesawing from boom to bust, as the US subprime mortgage market did during the past 18 months. But, despite the current appetite for new ideas, Persaud concedes that banks would not sit by quietly while supervisors moved the goalposts: "It would be politically very difficult. It would take a bold, confident policy-maker to push that through."
Tying capital to loan growth also has its critics outside the banking industry. Oliver Wyman's Kuritzkes sees this as a backward step: "Setting capital in relation to the size of the balance sheet is a very crude mechanism. It's a throwback to what existed in the US before we had the first Basel Accord. I don't think that's an improvement."
He also questions one of the pro-cyclicality argument's basic tenets - that capital levels are at their lowest when the business cycle is peaking. Despite the benign, low-risk period that stretched almost uninterrupted from 2001 through to July last year, banks' capital levels were generally well above the regulatory minimum when the crisis hit, having used the preceding period of calm to retain some of their profits, he says. Instead of looking for ways to decouple risk and capital, supervisors should simply make it clear that low minimum requirements during benign periods should always be generously supplemented, Kuritzkes argues.
De Nederlandsche Bank's Knot contends that Basel II already contains counter-cyclical measures. The main inputs into bank capital calculations are probability of default (PD) and loss given default (LGD), and controversy has swirled about whether those estimates should be a snapshot of the risk level at one point in time or whether they should be influenced by expectations of how a credit would perform when times are hard - the so-called through-the-cycle approach. Knot says there should be no debate: "Basel II specifically says that you should take downturn conditions into account when setting your LGDs. All in all, there are very explicit built-in stabilisers, such as robust data requirements, downturn LGDs, and a requirement to perform stress tests." He insists that Basel II - if implemented properly by national supervisors - would not allow unchecked loan growth during a credit boom.
The devil, of course, is in the detail - and Oliver Wyman's Kuritzkes says there is precious little detail: "There's a lot of easy talk about through-the-cycle PDs, but, to my knowledge, no one has come up with a rigorous way to evaluate these dynamic time effects." There is also a trade-off to be made, he says. One of Basel II's main aims was to curb regulatory capital arbitrage, in which banks took advantage of the old regime's insensitivity by retaining more risk (and its associated higher returns) without having to hold an appropriate level of capital. Tying capital more closely to the underlying risk of an asset was supposed to curb this behaviour - but a through-the-cycle approach would again open the possibility that risk levels at certain points in time could become decoupled from capital requirements.
Despite these criticisms, De Nederlandsche Bank's Knot says he is confident Basel II would have helped dampen the impact of the liquidity crisis had it been in place at the time: "We shouldn't expect that regulation alone can prevent such things from happening, but I think the problems we saw emerging would have been tackled earlier and therefore the effects would have been less severe. I know there will never be a perfect answer to these situations, but I'm confident Basel II is a better answer."
Others are not so convinced. "It's always fair to ask questions and, in this case, I sincerely hope those questions will include the regulatory framework," says the LSE's Goodhart.
DISAPPROVING OF DISINTERMEDIATION
The current credit crisis has come close to making disintermediation a dirty word - short-hand for the process through which exposure to US subprime mortgage loans was originated by banks but ended up distributed around the financial system, spooking investors and causing global panic. "It's bound to cause the originate-to-distribute model to be queried," says Charles Goodhart, director of the regulation and financial stability programme at the London School of Economics (LSE), and a former member of the Bank of England's rate-setting committee.
But this is a tricky issue. Although there is widespread acceptance that banks have embraced the originate-to-distribute model, not everyone thinks that is the same thing as disintermediation - while risk was distributed, much of it remained within the banking industry, they argue. And there is also little agreement on the extent to which the Basel Accord is to blame.
Avinash Persaud, chairman of London-based liquidity advisory firm Intelligence Capital, argues that disintermediation did take place, that the risk ended up with investors who were poorly equipped to hold it, and that these investors all rushed en masse for the exit when prices started falling - causing liquidity to suddenly dry up. He also lays the blame for this at the door of the Basel Committee: "We have a system of bank regulation that is like squeezing a tube of toothpaste with the top still on. If you squeeze in one place, all you do is move the risk to another place." Regulatory capital squeezed banks, he argues, so the risk ended up with other types of credit investors.
But did it? Pierre Cailleteau, chief international economist at Moody's Investors Service in London, points out credit losses have been more concentrated within the banking industry than might have been expected: "So far, we haven't seen a lot of problems with pension funds and insurers. There's no blood in the water. The people who are bleeding are Merrill Lynch, Lehman Brothers and the rest."
Others, however, feel the preponderance of banking industry losses is an illusion. Andrew Gray, a partner at PricewaterhouseCoopers in London, says: "Banks only appear to have suffered a majority of losses because they have tougher disclosure requirements. That's not necessarily the case for hedge funds and other investor groups."
Nonetheless, regulators insist the extent of disintermediation has been exaggerated. Klaas Knot, director of supervisory policy at De Nederlandsche Bank in Amsterdam and chair of the Basel Committee's risk management and modelling group, says "a lot of the disintermediation took place from the perspective of an individual bank, and subsequently the credit risk was traded to a different bank. There was no disintermediation from a systemic perspective - credit risk remained in the banking industry."
This was the fault of the old Accord, he argues, which allocated less capital to credit risk-bearing assets in the trading book than it did to the same risk when it was held in the banking book. Basel II addresses this problem via the incremental default risk charge, which forces banks to hold capital against traded credit risk, says Knot. The precise workings of the charge have been controversial and are one of the remaining areas of the new Accord that has not yet been finalised (see Risk January 2008, pages 96-100). Knot hopes that will happen in the first quarter of 2008, with a view to implementation in 2010. If it had been in place prior to August, banks would have been less keen to sell credit risk-bearing assets and would have held more capital against the assets they bought, he argues.
A separate criticism of disintermediation is that it removes the need for banks to monitor and manage the risk associated with the underlying assets. In the case of the subprime mortgage market, for example, it is alleged that banks stopped caring about the quality of the loans they were making because they knew the risk would be passed on to someone else. As a result, loan terms became too loose.
Intelligence Capital's Persaud suggests this problem could be solved by restricting the proportion of each credit that banks are able to securitise: "Maybe a large chunk of every loan should be kept on the bank's balance sheet. Rather than being able to securitise 100% of every loan, they should only be allowed to securitise 50%. That creates an incentive for the bank to lend more responsibly, to monitor the risk and to manage it. The bank becomes a co-investor."
Whether or not there is any formal attempt to restrict the use of the originate-to-distribute model, the ability to do so currently has been severely curtailed by swirling fears about credit risk, transparency, rating accuracy and valuations, says Moody's Cailleteau: "In the near term, there will be a big premium on commoditisation and standardisation. The ability to securitise will be severely restricted and we'll see some re-intermediation: balance sheets will swell. But I don't think for one second the industry will go back to the old business model."