Basel II: The View from the US
The Basel II implementation process in the US is not going very well. To begin with, the US will be implementing Basel II at least two years behind the rest of the world. Also, capital floors will apply for three years instead of two, and capital may not fall more than an aggregate of 10%. Recent hearings in both the House of Representatives and the Senate reveal that lawmakers believe capital should not fall in the US, but they are also concerned that, should capital fall elsewhere, US banks could be at a competitive disadvantage. Initially, US regulators proposed to implement only the most advanced version of Basel II, but it now appears that other versions may be allowable. Different opinions have clouded the issue and the fact that there are at least five regulatory agencies involved is not helpful, but this is not the source of the problem.
To answer these two questions properly, we need to look at the history of Basel II’s development. The initial implementation of the Basel Accord in 1988 was not without controversy. Indeed, it did not become fully effective until 1992 and at that time required the largest bank holding companies in the US to raise some US$50bn in equity for additional capital. This was certainly not an insignificant sum, and it depressed industry earnings greatly. And while the Basel Accord was called a risk-based capital standard, it was that in name only since it required an 8% capital-to-assets ratio (4% for mortgages).
But the Basel Accord did help stabilize the banking system and it raised capital substantially. It was also highly effective as an international capital standard so long as institutions retained their assets on their balance sheets. The advent of economic capital also helped the industry better understand how much capital it should hold relative to the risk it took. It made little sense to retain risks on the balance sheet that required less than 8% capital, and these were securitized. Clearly, a capital standard that encouraged banks to hold the riskiest exposures on their balance sheets was not effective.
Meanwhile, as the industry began to develop more complex products and enhanced internal risk management systems, it became clear that the Basel one-size-fits-all capital standard was becoming increasingly ineffective. In 1996, the Basel Committee implemented the market risk amendment, requiring institutions to hold capital for the risk imbedded within trading book activities.
The market risk amendment represented an important step toward Basel II. It allowed institutions to use their own internal capital estimation and allocations processes (their models) to help establish the regulatory capital requirement. Of course, the regulators required a multiplier add-on to the internal capital estimate.
Many within the industry (RMA included) argued that the Basel Committee should further amend the Accord to allow banks to use their own internal models to establish capital for credit risk. The Basel Committee rejected this idea and instead developed Basel II.
The most advanced version of Basel II allows banks to use their own internal estimation of probability of default, but that is about it. The parameter estimation process for loss given default (LGD) and exposure at default (EAD) has become so prescriptive that neither reflects a bank’s internal estimate. This is particularly the case in the US with the latest version of the notice of public rulemaking (NPR) and is one reason why some have suggested that the standardized option be open to banks here. The AIRB has moved so far away from internal practice that it can no longer be used for internal risk management purposes; for institutions with highly advanced risk management systems, it would be little more than a very expensive compliance exercise. Naturally, if you have to develop a capital system that cannot be used for internal purposes, the least expensive version is attractive.
Some have suggested that Basel II allows the industry to establish its own level of capital. This is not the case. The industry plugs its own internal parameter estimates into a regulatory model, and it is the regulatory model that determines the level of capital that an institution must hold.
When the Basel II framework was released in June 2004, the Basel Committee specifically stated that it was designed to evolve over time as risk management practices within the industry evolved. Without this implicit promise, that Basel II would evolve, it is unlikely that the US banking industry would have supported moving forward with the new capital framework. The 2004 framework held out the hope that, over time, the industry would be able to use its own internal risk measurement and modeling processes to determine regulatory capital. However, it is becoming clear that the prescriptiveness embedded within the NPR could stifle further evolution of the framework and break this all-important promise.
It is important to remember that the safety and soundness of the banking system is dependent not only on capital but also on risk reduction. Indeed, the CAMELS rating system in the US stands for:
US institutions have steadily improved their risk management practices over the past decade. Many learned the hard lessons of the last significant downturn during the late 1980s and early 1990s. Those that did not are no longer around. The earnings capacity of the industry remains strong. Many believe that the greater diversification and emergence of truly national franchises have strengthened the industry.
Much, but certainly not all, that Basel II seeks to accomplish in terms of improving risk management has already been achieved in the US. Since 1999, large complex banking organizations have undergone a regulatory examination to determine the adequacy of their capital estimation and allocation practices. These examinations, known as 99-18 exams, have become increasingly more thorough over the last few years. And it is worth pointing out that with or without Basel II implementation, US regulators have the authority to require US institutions to improve their capital estimation and allocation practices. It should also be pointed out that the US has required institutions here to hold capital for residual risk from securitization activities since 2001.
Some in the US have lately argued that, since risk management practices have been significantly improved (a primary objective of the Basel II process), it is not necessary to bother with tying the regulatory capital standard to Basel II. Moreover, since capital cannot fall 10% without Basel II forced into recalibration, some argue ‘what’s the point’? Also, it is highly unlikely that rating agencies here would allow a major institution to lower its capital significantly without also suffering a downgrade – a risk no institutions would likely take. Therefore, it is highly improbable that capital would fall in the US, with or without the tougher floors imposed by the US regulatory agencies in the fall of 2005.
In the US, the regulatory agencies and the rating agencies will not allow capital to drop, at least not at this juncture. And on top of that, the leverage ratio (5%) will remain in place. So, the US industry is examining the NPR very carefully and will adopt parts of the proposal that truly advance risk management practices and reject those that do not. Hopefully, US banks will not be forced to develop costly regulatory capital-allocation processes that cannot be used for internal purposes.
Otherwise, additional Congressional hearings will surely follow and the Basel II implementation process will be delayed further.