How the G-20 Should End Too-Big-to-Fail

Call in the reinforcements.

Photographer: Hannelore Foerster/Getty Images

In the darkest days of the 2008 financial crisis, the leaders of 20 developed and developing nations gathered in Washington and set out an ambitious plan for reform. Never again, they agreed, should the world's biggest banks be able to present such a threat to the global economy that governments would be forced to rescue them.

Six years later, as the Group of 20 prepares for this week's summit in Brisbane, the goal of ending "too big to fail" remains frustratingly elusive.

Ahead of the meeting, regulators have been working to improve an important part of the financial architecture: resolution regimes, which are supposed to provide an orderly way of managing failures of large, systemically important banks without resorting to taxpayer-backed bailouts. This is a tall order, because the biggest bank holding companies contain thousands of subsidiaries moving money and risk across borders with such speed and complexity that their own executives often can't keep track. Any sign of distress could quickly turn into a fiasco as markets guess at the scope of the impact and national authorities rush to seize assets to cover local creditors' claims.

Regulators think they have hit upon a solution. Instead of trying to deal with all the pieces of a distressed bank separately, they will simply swoop in and recapitalize the holding company, allowing all the subsidiaries to keep running as if nothing was wrong. To that end, they are drawing up plans to require the holding companies to issue a special kind of long-term debt, which the regulators can convert into equity in an emergency. They have also nudged banks to insert a stay in derivatives contracts, a feature that will temporarily hold creditors at bay while regulators perform their salvage operations.

Although both moves are necessary, neither is sufficient to ensure that regulators can handle even a single large failure, let alone a crisis like what the world endured in 2008.

Bloomberg News has reported that the special "bail-in-able" debt requirement might increase the largest banks' total loss-absorbing capacity to 25 percent of risk-weighted assets. This achievement would be smaller than it seems, because banks have become adept at minimizing the denominator in that equation. As a share of total assets, with no risk-weighting applied, it could be as little as 6 percent. In a market meltdown, where falling prices make it difficult to assess the extent of losses, it's hard to imagine how such a thin buffer would give authorities and creditors the comfort they need to refrain from grabbing assets and cutting off funding. The derivatives stay would only slightly alter the nature and timing of the disaster. The only fix would probably be a blanket guarantee from the home country's government -- that is, an old-fashioned, taxpayer-backed bailout.

Given the inevitable difficulties of handling big bank failures, a better approach would be to make the banks less likely to experience distress in the first place. If, for example, banks financed themselves with much more equity than the currently proposed global minimum of 3 percent of assets (5 percent in the U.S.), shareholders would in most cases absorb losses with no need for regulators to get involved. True, in good times, higher equity levels would reduce profitability measures such as return on equity, which might not be good for executives' bonuses. That said, it might also prompt shareholders to consider simplifying the banks' structures -- a move that could both boost their value and make them easier to dismantle in a crisis.

When the G-20 leaders meet this weekend, they should consider whether their approach to ending too-big-to-fail is more complicated than it needs to be.

To contact the senior editor responsible for Bloomberg View’s editorials: David Shipley at