The Treasury Department’s sweeping report on regulatory reform includes the recommendation that the U.S. reduce its reliance on the advanced approach to large-bank capital requirements, favoring instead the standardized approach. Because this recommendation is included with a raft of other, often highly controversial proposals, many have overlooked its appeal.
Backing away from the advanced approach for risk-weighting assets would mean moving on from a method that has had limited value; eliminating it has no downside risks. In fact, in our recent paper measuring the cost-benefit trade-off of the advanced approach, we found that demanding advanced models for risk — which is already well captured by the dozens of other capital rules — not only adds burden but also poses a lot of unintended risk. (The paper was funded by the Regional Bank Coalition, but reflects the views only of Federal Financial Analytics.)
Our analysis started by counting all of the capital standards governing large, noncomplex U.S. regional banks and bank holding companies. Many of the capital requirements, including the advanced approach, originally stem from international Basel Committee accords. Suffice it to say that each bank must comply with over two dozen capital standards. One is tempted to say that the problem here isn’t so much that regulators don’t trust big banks, but rather that they don’t trust themselves or their fellow regulators — and have piled on the capital rules just in case someone misses something somewhere.
Of course, so extreme an abundance of caution might seem warranted given the grave cost of the financial crisis. But the onslaught of cascading rules — imposing heavy burdens on banks to comply with each one — has its own risks. Indeed, when regulators try to assess the costs versus benefits of a particular rule, they do not sufficiently consider the impact of other rules.
The way to assess the tradeoff of the advanced approach is to deploy a methodology we call marginal cost-benefit analysis (M-CBA). This methodology looks not just at the advanced approach on its own, but also in its context taking into consideration all of the other capital rules and all of the other rules — e.g., new liquidity ratios — that directly intersect with the capital rules.
Regulators are more and more looking at the costs and benefits of each rule, but none yet does so from a marginal-CBA perspective — that is, to see whether a new rule has sufficient benefit to outweigh its costs given all the other rules.
The advanced approach relies heavily on banks’ internal models to determine risk weights for various assets. But the first thing that becomes immediately apparent about the advanced approach from an M-CBA perspective is that the models-based standards are rarely, if ever, the binding capital standard for large, noncomplex regional banks — let alone for any other big bank subjected to them. The reasons for this are simple and sharply differentiate the U.S. from the global framework.
First, and very importantly, the U.S. requires its banks to hold the higher of the standardized or advanced approach. Basel is wracked by seemingly unresolvable controversy over the relationship between these two risk-based metrics, but the U.S. has simply solved this by requiring banks to meet the tougher threshold every time. As a result, banks are prevented from gaming the system by meeting the lower bar.
Second, the U.S. stress tests are far tougher than those mandated in most other nations. As a result, stress tests for complex risks such as derivatives capture complexities that the standardized approach on its own could miss.
And, even if neither the standardized approach nor the stress test, run through the Federal Reserve’s Comprehensive Capital Analysis and Review, is right for an individual bank’s risk profile, each regulator has more than enough legal and regulatory authority to stipulate more capital. Internal risk management is also charged with ensuring ample capital regardless of what the rules say, with regulators more than able to hold the bank accountable for any lapses a supervisor might miss.
The most important cost of changing a capital rule is the risk posed to a bank’s solvency and by inference to the rest of the financial system. But, as our paper makes clear, the advanced approach has very little bearing on risk because, as described above, it’s almost always not a bank’s binding constraint. On top of the stress tests, prompt corrective action and early remediation fully equip regulators to intervene long before capital ratios falter. That regulators all too often fail to use the tools they have is no justification for simply piling on more costly rules.
If those other tools fail, does the advanced approach make large regional banks and BHCs easier to unwind, without causing further havoc to the financial system? The answer is no. Taking away the advanced approach would not take away any of the separate resolution-planning requirements or all the other capital, liquidity, governance and remaining sanctions applicable to these companies.
In fact, retaining the advanced approach undermines all these other rules by diverting hundreds of hours of staff time and millions of dollars away from effective risk management on a forward-looking basis. Board and senior management resources are also diverted from managing risk appetite to scrutiny of the 200 million or so models that the lower-level staff must construct to make the advanced approach tick.
Perhaps your only reservation about agreeing with the Treasury recommendation is a concern that reducing reliance on the advanced approach would further distance the U.S. from the Basel rules and thus fragment global standards. But, let me remind you: The U.S. rules do not, as Basel requires, apply the advanced approach to “internationally active” banks. Instead, U.S. regulators picked a proxy asset size — $250 billion — and slapped the advanced approach even on large, noncomplex banks and BHCs with no material activities outside our shores. As a result, U.S. regulators are free to realign the standardized and advanced approaches.
Treasury has rightly started the debate about the marginal value added of the advanced approach. When regulators finalized it in 2007, they recognized that this complex capital standard posed an array of potential pitfalls. They thus promised to carefully re-evaluate the rule as it was implemented. A decade later, no such review has been announced, let alone conducted. It is past time to do so.