WASHINGTON — Weeks after JPMorgan Chase disclosed its massive loss from its trading activities, most of the policy discussions have focused on whether to toughen the Volcker Rule that bans proprietary trading.

But another — and possibly more significant — debate is taking place behind the scenes: what does the loss say about the state of big banks' capital requirements?

The $2 billion-plus loss has bolstered those who argue that regulators are over-relying on data-driven exercises like stress tests, which failed to see JPM's London hedges as a threat, when they should be even more focused on higher capital requirements across the board.

"This only highlights the need for capital. If you are going to control, the main thing you want is that they would have a lot of loss absorbency. So whatever stupid things they do, it's not going to be somebody else's problems," said Anat Admati, George G. C. Parker Professor of Finance and Economics at the Graduate Business School at Stanford University. "Solvency problems are at the heart of any banking problem and they just have to be reduced and that's what capital is about fundamentally, reducing the solvency concerns."

Yet Admati and other capital proponents see flaws in even the revised set of standards proposed by the Basel Committee on Banking Supervision, which the Federal Reserve Board is set to formally embrace at a meeting Thursday.

"The problem that we have — and the recent JPMorgan episode illustrates that — is what happens when banks start taking on a lot of risk in parts of the bank that are not historically where the risk has been taken," said Robin Lumsdaine, a former Fed official and now a professor at American University. "The capital regulations are designed to focus on the riskiest areas of the institution. If risk is being taken in other parts of the firm, then the capital regulations are understandably going to be somewhat inadequate."

In the case of JPMorgan, for example, capital regulations wouldn't have touched the firm's office of the chief investment office, which is largely considered a risk management function.

"You assume that the CIO office is the one that's looking across the firm, doing that detailed granular evaluation, identifying risks, and managing them properly," said Lumsdaine. "That means the only positions they have, the only risk exposures they should be taking on, are hedging positions. They're not supposed to be the part of the firm that's speculating."

To be sure, in the case of JPMorgan, the firm was easily able to absorb the losses, which were small relative to the size of the firm. That is in no small part due to its high capital levels.

U.S. firms have built up their capital since the Federal Reserve began stress testing institutions in 2009. The 19 bank holding companies, including JPMorgan, that participated in those and subsequent tests have increased their Tier 1 common capital levels to $759 billion by the fourth quarter of 2011 from $420 billion in the first quarter

of 2009. Tier 1 common ratio for these firms has also increased to 10.4% from 5.4%.

Under Basel III, banks will be required to hold 7% capital by 2019, while the largest firms will have to hold an extra cushion of between 1% to 2.5%.

Still, many experts wonder whether that's enough. On its own, the JPM loss wasn't a systemic threat. But what happens if other banks engage in similarly disastrous trades — or losses emerge from activities that were previously considered low-risk (as mortgages were prior to the financial crises)?

"The key is what's the risk?" said Admati. "Is this the worst case? Is this what we're preparing for? The answer is, No. This is just a suggestion of the risk that are in and around their balance sheet. It's not a fortress. The risks — you don't see them there on these accounting numbers of this balance sheet."

Observers are particularly concerned about the stress tests, which regulators increasingly are relying on yet failed to foresee the trading loss at JPMorgan. It also didn't account for the reputational hit that the firm took when it finally disclosed the losses after Jamie Dimon, the firm's chairman and chief executive officer, previously dismissed them as a "tempest in a teapot."

In its stress test results released by the Fed, the New York-based institution held a Tier 1 capital ratio of 9.9% under normal expected circumstances and 5.9% capital ratio under extremely adverse conditions such as a 13% rate of unemployment and a sharp decline of gross domestic product to minus 8%. The results did show a $27.7 billion trading and counterparty losses, which included mark-to-market losses, changes in credit valuation adjustments and incremental default losses, as well as loan losses worth of $55.8 billion from first lien mortgages, credit cards and other loans.

But it would not have included, or even detected, individual hedges that would have pointed to a potential risk that led to JPMorgan's $2 billion trading loss.

"How might the stress tests reveal anything about what ultimately happened in this case?" said Richard Spillenkothen, the former head of supervision at the Fed. "It's just a question of how the stress tests were carried out, what the focus was. How broadly was it defined to deal primarily with credit losses across the portfolios?"

Lumsdaine said it proves the stress tests should be even broader.

"It really highlights how important it is for stress tests to be comprehensive and enterprise-wide," she said. "At some level, each stress test is about a firm, so you could argue it is enterprise-wide. But the stress tests are designed to assess the effects on a bank's capital — as a result they are focused on the parts of the bank where historically the risk has been taken."

William Isaac, the former chairman of the Federal Deposit Insurance Corp., argues that regulators are too focused on stress tests — and even capital levels. For Isaac, regulators aren't capable of right-sizing capital requirements, an effort they have tried to successfully undertake for the last two decades, and exercises like stress testing don't make up for good governance.

"You cannot regulate this banking industry by remote control," said Isaac. "We're trying to take the human factor out of this and judgment out of supervision of banks. You can't, otherwise you get the results that we had in 2008 and 2009. Regulators should be inspecting the institution and making sure they know what's going on and that the bank has proper control systems in place and governance systems in place."

But others said that's an impossible task, and said the focus on capital remains the simplest, even if imperfect, solution.

"We should not try to micro-manage banks, we should insist that they always have adequate capital cushions," said Mark Flannery, a finance professor at the University of Florida. "We shouldn't ever let ourselves get in a position that a bank is almost out of capital. We ought to start a lot sooner at a lot higher capital level and make up losses as they occur. That's almost the opposite of micro-managing that focuses on the ultimate protection against misbehavior or bad luck. You are bound to lose if you decide you are going to micromanage."

Spillenkothen agreed, saying it's unrealistic to expect regulators to get it right 100% of the time.

"Even if you successfully prevent three out of four, or five out of six, or nine out of 10 bad things from happening, one thing out of 10 that can happen could have serious consequences, so you need to have really strong capital," he said.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.