The great financial crisis was caused in large part by complexity: complex products, complex institutions, and complex counterparty networks. Yet, post-crisis regulatory reforms designed to prevent future financial crises, principally Basel III, are equally complex. Sophisticated actors easily exploit complex systems, and large banks certainly are sophisticated – if nothing else.
Several empirical studies have shown that large banks holding similar asset portfolios often come up with widely different capital charges, thereby leading to excessive variability in risk-based capital ratios. Such discrepancies deprive the ratios of meaning, making it difficult for investors and the public to assess the overall health of any given financial institution. The Basel Committee has grown wise, and at the beginning of this year passed a work program designed to reduce the variability in risk-weighted assets (RWAs) with the goal of finalizing all changes by year-end. But that goal is in serious jeopardy, as several member jurisdictions have begun to distance themselves from the Committee’s work. Now with the rise of populist political parties around the globe, it’s not just the Committee’s work program that is in trouble. Rather, the entire Basel edifice is at risk of crumbling down, taking with it three decades of international cooperation on banking regulation.
Basel’s History
The failures of Herstatt Bank in Germany and Franklin National in the U.S. in 1974, awakened banking regulators to the contagion risks posed by internationally active banks operating in an increasingly interconnected world economy. The regulatory cooperation that ensued gave birth to Basel I in 1988, resulting for the first time in internationally agreed upon minimum capital standards. Basel I only accounted for credit risk, and did so in a rudimentary fashion, applying risk-weights to assets based upon four broad categories. This method of calculating RWAs later became known as the standardized approach.
Basel I quickly outlived its usefulness, as banks became more complex and developed sophisticated risk modeling techniques with the help of computers. Basel I was updated in 1996 with the adoption of the market risk amendment, which required banks to hold capital for potential losses in their trading portfolios. The market risk amendment ushered in a new era in capital regulation by allowing banks to use their own value-at-risk (VaR) models as an input when calculating the market risk capital charge.
After passage of the market risk amendment, banks continued to press regulators to allow for greater use of their internal models when determining how much capital to hold for other types of risk. Regulators obliged with passage of Basel II in 2004. Basel II created the internal ratings based (IRB) approach which allowed the largest banks to use risk parameters determined by their internal systems as inputs into a formula developed by supervisors for calculating minimum regulatory capital for credit risk.[1] Basel II also required banks to hold capital for operational risk events, and created the advanced measurement approach (AMA) to do so. The AMA allowed banks to use their own models, subject to regulatory approval, to calculate how much capital they needed to set aside for operational risk. Basel II also maintained, and refined, the standardized approach introduced by Basel I.
U.S. banks were in the process of implementing Basel II when the financial crisis came. The crisis exposed Basel II’s weaknesses and limitations, the principal one being that it didn’t require banks to hold enough capital. And although many critics of Basel II decried its reliance on banks’ internal models, Basel III doubled down on the practice, as we shall see.
Basel III
Basel III was approved in 2010 and most significantly:
- Established more stringent measures of capital and risk-weighted assets;
- Increased the required level of capital;
- Introduced liquidity requirements;
- Created a minimum international leverage ratio of tier 1 capital to total on-balance sheet assets and off-balance sheet exposures; and
- Included risk-based capital buffers and surcharges for global systemically important banks (G-SIBs.)
However, Basel III maintained the use of internal models-based approaches to calculating RWAs, and large banks typically choose this method over the standardized approaches. Banks may claim they prefer to use their own models because it provides for a more accurate picture of their risks, but the reality is that they get away with holding less capital by using internal models-based approaches.[2] And when each bank is using their own internal models to determine how much capital they need to hold, it’s not surprising when they come up with different results for similar portfolios. The Basel Committee itself has supplied evidence supporting this finding. In 2013, the Committee’s analysis found “significant variation in the outputs of market risk internal models used to calculate regulatory capital.” A similar study of RWAs for credit risk also identified “considerable variation” across banks, and that this variation cannot be explained simply by differences in the risk composition of assets. Significant variation has also been found in operational risk RWAs.
Reform Efforts
Allow me to quickly recap where we’re at. To the surprise of no one – except the Basel Committee perhaps – providing large banks the opportunity to use their own models when determining how much capital they must hold has led banks to: (1) use their own models, (2) reduce the amount of capital they would otherwise hold, and (3) come up with widely different RWA calculations for similar types of portfolios. This outcome has eroded the credibility of regulatory capital ratios as it makes assessing the health of any one bank vis-à-vis its peers extremely difficult.
The Basel Committee is attempting to solve the problem of excessive variability in RWAs by the end of the year. On December 1st the Committee met in Santiago, Chile where supposedly “very good progress was made towards finalising the Basel III post-crisis reforms.” Over the past year, the Committee has published four consultative documents that would revise the current standards for: (i) the standardized approach for credit risk; (ii) operational risk; (iii) the internal ratings-based (IRB) approaches for credit risk, including a potential “output floor”; and (iv) the leverage ratio. Let us quickly look at each of these proposals.
Credit Risk
Rather than completely do away with banks’ internal models, the Committee is proposing a constrained internal ratings-based (Constrained IRB) approach for the calculation of credit risk-weighted assets.[3] This would require that the standardized approach be used for certain credit exposures, such as financial institutions, while establishing input floors when models-based approaches are used. These input floors would apply to banks’ estimates of probability of default, loss given default, and credit conversion factors for portfolios that remain eligible for the IRB approach. Floors would also be established for determining counterparty credit risk RWAs.
The Committee has also proposed strengthening the standardized approach for credit risk.[4] The new approach would reduce the reliance on external rating agencies[5] while enhancing the risk sensitivity and calibration of specific risk exposures, such as real estate.
Operational Risk
The Committee has proposed eliminating the advanced measurement approach (AMA) which allowed banks to rely entirely on their own models when determining how much capital to hold for operational risk events.[6] Instead, banks can use the standardized measurement approach (SMA) which relies on a combination of financial statement information and banks’ internal loss experience.
Output Floors
Because the Committee’s proposed reforms don’t completely eliminate the use of banks’ internal models, they are considering replacing the existing capital floors which are based on the Basel 1 framework. So for example, a bank using the IRB approach to calculate credit risk RWAs cannot come up with a value that is below 80% of what the credit risk RWAs would be under Basel I. The Committee is considering a more risk sensitive output floor.
The Leverage Ratio
The leverage ratio is designed to serve as a compliment to risk-based capital ratios. It is calculated by dividing Tier 1 capital by on-balance sheet assets and off-balance sheet exposure. The Committee’s proposal to reform the leverage ratio revises the method for calculating certain derivatives exposures and introduces a leverage ratio surcharge for G-SIBs. The proposal stops short of eliminating central banks placements from the leverage ratio denominator, which the Bank of England did in August.[7]
Pushback
Over the summer, the Committee attempted to assess the impact their proposals would have on bank capital levels by conducting a quantitative impact study (QIS). The results haven’t been publicly released yet, but judging by their reaction, it’s clear the Europeans did not like what they saw. In September, European Commission Vice President Vladis Dombrovskis announced that the EU will not implement any of the Committee’s final rules. Dombrovskis believes “the proposals Basel has issued for consultation would imply significant capital requirement increases in all areas,” and that “at a time when we are focused on supporting investment, we want to avoid changes which would lead to a significant increase in the overall capital requirements shouldered by Europe’s banking sector.”
If the goal is to reduce RWA variability without increasing aggregate capital levels, then some jurisdictions will end up holding less capital while others will have to hold more. Clearly the QIS results indicated that the Europeans would have to hold more, which was expected. Compared to the U.S., the Europeans have adopted less stringent capital requirements post-crisis. This partially explains why the market has been so concerned about European bank capital levels and why the shares of European banks have been performing so poorly. But Europe has also been struggling with anemic economic growth, and Euro area politicians are weary of imposing any additional requirements on their banks that may hinder the flow of credit to the real economy.
Trumped
The EU’s decision to ignore the Basel Committee’s final proposals places the future of the Committee on shaky ground. But it may be the Trump administration that finally topples the tower of Basel. President-elect Trump’s economic advisor David Malpass was recently quoted as saying: “As regulations are given from Basel into the U.S. financial regulators, we now have kind of an agreement…where the U.S. is absorbing regulatory guidance, regulatory suggestions from the international community.” According to Malpass, this system does not work “for the average person.” If the U.S. takes its ball and goes home, the Basel Committee will exist in name only.
What started off as an innocuous attempt by the Basel Committee to simplify post-crisis capital regulations has led to the future of the Committee being cast into doubt. This is partly due to political currents beyond the Committee’s control, but there is no doubt the committee bears some responsibility. Faced with a once-in-a-lifetime financial crisis, the Committee missed a golden opportunity to simplify financial regulation while strengthening the health of internationally active banks. Upon realizing their error, the Committee rolled out a series of reforms that can best be described as adjustments around the edges. Eight years after the crisis, even these adjustments were too much for some members to bear. The bonds of cooperation that have tied financial regulators together since 1988 have been broken, and the future is uncertain.
[1] Examples of some of the inputs banks can provide are: probability of default, loss given default, and exposure at default.
[2] In the U.S., the Dodd-Frank Act contained the Collins amendment, which requires U.S. banks to use the more conservative of the standardized approach or the internal models-based approach. Thus, U.S. banks are unlikely to be significantly impacted by the reforms under consideration by the Committee.
[3] Reducing variation in credit risk-weighted assets – constraints on the use of internal model approaches, March 2016.
[4] Revisions to the Standardized Approach for credit risk, December 2015.
[5] The Dodd-Frank Act eliminated the use of external credit ratings in determining regulatory capital
[6] Standardized Measurement Approach for operational risk, March 2016.
[7] http://www.bankofengland.co.uk/publications/Pages/news/2016/062.aspx
