It is a description that means almost exactly the opposite of what it seems. “Too big to fail” doesn’t mean a financial institution cannot fail, but that it cannot be allowed to do so. Should that failure occur, it would bring catastrophe to the financial markets and the “real” economy.

As the financial crisis unfolded in 2008, federal regulators judged Bear Stearns, Fannie Mae, Freddie Mac, American International Group and a number of other financial institutions too big to be allowed to collapse, despite the firms’ missteps. The whole notion of rescuing such “TBTF” firms violates the principle of allowing the markets to judge which players sink and swim. But when regulators opted to let Lehman Brothers fail in September 2008, a tsunami roared through the markets and economy.

Now, with the worst of the crisis apparently over, regulators, lawmakers and other experts are debating how to treat such firms in the future. “There are two issues with too-big-to-fail,” says Wharton finance professor Richard C. Marston. “First, there are institutions which are not banks that have proven capable of wrecking our financial system. This includes Bear and AIG. We need a serious regulator to watch over them…. Second, we need to figure out what to do with both the big banks and the big ‘others’ to control future risks.”

The main concern Wharton finance professor Marshall E. Blume has about big institutions “is that if the investors in those institutions perceive the government is going to remove risks [with a bailout], then you’re going to have major distortions of the market.”

While few argue that these firms should be broken up by law, as AT&T was a generation ago, many believe there should be a better mechanism to dismantle failing financial firms in an orderly way, much as the Federal Deposit Insurance Corp. does with ordinary banks. Many say that to impose discipline on these firms, and thus discourage excessive risk taking, such a process should assure that debt holders, not just shareholders, suffer big losses when the government is forced to step in.

According to some experts, the problem of TBTF institutions goes beyond their size, or the number of employees, investors, counterparties and debt holders who get hurt when they fail. Some of these institutions are big enough to warp the way the markets work. In a September 27 op-ed article in The Wall Street Journal, two Federal Reserve officials argue that Fed efforts to stimulate the economy through reduced interest rates were rendered less effective by the big institutions. Normally, a Fed cut in short-term rates ripples through the system, leading to lower rates for businesses and consumers, who are thus encouraged to borrow and spend, lifting the economy. But this time around, the lower interest rates did not work because big institutions and their counterparties were afraid they might not get paid back if they lent to one another.

“The rates that matter most for the economy’s recovery — those paid by businesses and households — rose rather than fell,” write Richard. W. Fisher, president and CEO of the Dallas Fed, and Harvey Rosenblum, the bank’s executive vice president and research director. “Those banks with the greatest toxic asset losses were the quickest to freeze or reduce their lending activity. Their borrowers faced higher interest rates and restricted access to funding when these banks raised their margins to ration the limited loans available or to reflect their own higher cost of funds as markets began to recognize the higher risk that TBTF banks represented.”

Though many small banks remained healthy, they simply did not have the capacity to make up for the big institutions’ reduced lending. Businesses could not turn to an alternate method of borrowing — selling stock and bonds — because in recent years deregulation had allowed the TBTF institutions to become major players in those markets, the Fed officials write.

How to Define ‘Too Big’

At first glance, solving the problem seems straightforward: If an institution is too big to fail, it is too big, and should be broken apart.

But it is not that simple, says Wharton finance professor Richard J. Herring. Forced breakups like AT&T’s in 1982 have been based on antitrust laws designed to preserve competition, and those do not appear applicable in the case of TBTF institutions that do face heavy competition, he says. “You cannot say there have been huge antitrust violations, if any. So it would be a whole new rationale for intervening, and there would be lots of opposition to doing that.”

For one thing, it would be difficult to come up with a widely accepted definition of “too big,” Herring argues. Would it be measured by market capitalization? Assets? The amount of money at risk? The number of employees or lines of business? The amount of leverage? Also, he adds, there is some benefit to having very large financial institutions which can operate internationally, can afford to spend money on innovation and compete with other countries’ big financial companies.

Wharton finance professor Itay Goldstein notes that financial problems are not necessarily caused by firms’ large size; small firms can cause systemic problems if enough of them make the same kinds of bad bets, which can happen when they act with a kind of herd mentality. “In that sense, there may be some advantages to having large financial institutions, because they are less likely to fail in this coordinated [fashion],” Goldstein says. “When you have large institutions, they are more likely to see the big picture, and they are more likely to talk to each other…. And that might be good.”

When a crisis does occur, it may be easier for regulators to gather together executives of a few large firms to find a remedy than it would be to deal with hundreds of small firms, he adds.

“Every country has these” big institutions, argues Wharton finance professor Jeremy J. Siegel. “There are mega-companies in every industry. You go around the world and all the financial institutions are large, and there have always been banks that have been too big to fail…. Just because you have these things that are too big to fail doesn’t mean they are bad institutions.”

While it was trouble with the big firms that threatened the financial system, Blume notes the problem wasn’t caused by their size but by the kinds of things they were doing, like trading poorly understood derivatives through subsidiaries that concealed risks. “So the issue is, if we didn’t have the big banks, would we still have had the problem?” Blume asks. “My view is that we probably would have had these problems…. If you didn’t have one big bank doing it, you would have had small banks doing it.”

The problem, adds Herring, is not so much that the big firms are too big but that they are too complex and too opaque. He notes that the 16 largest financial institutions control 2.5 times as many subsidiaries as the 16 largest non-financial firms. One of the most complex financial firms controls 2,435 subsidiaries, half of them chartered in other countries. Many such subsidiaries are formed to “minimize regulatory burdens” or reduce tax bills, he says. Such complexity makes it difficult for anyone — including regulators and the companies’ own managers and directors — to fully understand all the risks the firms are taking, or how those risks might interact with ones other companies are taking.

The Obama administration has proposed moving to the Federal Reserve much of the financial-institution regulation now divided among various agencies. While many details are to be worked out, the Fed would have new authority to gather information from financial institutions to gauge systemic risks. It also would establish a “council of regulators” to better coordinate efforts of various agencies, and it would establish a system for shutting down failing firms.

A ‘Winding-down’ Plan

Herring suggests that each financial institution also be required to maintain its own “winding-down plan” to explain how risks will be reeled in and the company dismantled, if necessary, in a crisis. Such a plan would provide the firm’s directors and regulators with details on the firm’s operations, such as maps of subsidiaries and procedures for shutting each down, data on insured versus uninsured deposits, and accounts of holdings in derivatives. Preparing such a plan would involve a stress test coordinated with other large firms to assess what would happen in adverse financial conditions. The winding-down plan would even report the number of days it would take to dismantle the entire firm.

Had such a procedure been in place a few years ago, the worst of the current crisis might have been averted, Herring argues. Insight gleaned from the plans’ preparation could have led firms to grow less rapidly, avoid overly complex corporate structures and take less of the kind of risk that threatens the financial system. Regulators might have been alerted to the growing fragility of the system. And, in the worst case, firms could have been wound down in a less disruptive way.

“It’s just asking a series of very simple questions about why are you as big as you are and why are you as complicated as you are,” Herring says. “And if you cannot answer to the satisfaction of your [supervising regulators] and boards [of directors], then maybe [the firm] should be unwound.”

The Dallas Fed’s Fisher and Rosenblum write that: “The problem isn’t just the riskiness of a big bank’s assets, nor even the bank’s size relative to the overall system. It’s important to know whether the bank’s asset holdings are highly correlated with those of other banks. Did they all make the same bad bets at the same time? Did they all bet that real estate prices would rise forever? As we all know, the answer, in this decade, unfortunately, is ‘yes.'”

Among the big issues to be resolved: Which federal agencies should have the oversight task? While the Obama administration would give most of the power to the Federal Reserve, critics argue the Fed did little to avert the recent crisis and may not have the culture and expertise to tackle such matters, even with new powers.

But Marston says the Federal Reserve is the logical choice. “We need the Fed to be given power to regulate [TBTF firms] more tightly if they are deemed ‘too big.’ That means more than just additional capital requirements. I think it means making sure they don’t innovate around any existing restrictions.” It also means such firms should not be permitted to use subsidiaries in a way that suggests they remove risk from the parent firm when they actually do not, he says, adding that the Fed “needs a mechanism to shut down these institutions in the event of failure — the way the FDIC shuts down small banks.”

Many experts believe big firms should be required to maintain larger capital reserves, and that they should not be allowed to bet too much with borrowed money, a key factor in the recent crisis. “Most observers agree on the need to implement and enforce rules that require more capital and less leverage for TBTF financial institutions,” the Fed’s Fisher and Rosenblum write. “Think of it like lower speed limits for the heavy trucks, the ones whose accidents cause the most damage.”

Another problem, according to many experts: The big government bailouts in the recent crisis just encouraged more risk taking. Although shareholders suffered deep losses, many firms’ creditors were rescued by the taxpayer. Debt holders would impose better discipline on firms if they believed that they, too, would be wiped out in a financial crisis, notes Marston. “It’s unacceptable for the Fed to be forced to back up all AIG contracts at face value. We taxpayers will be paying for the AIG mess for years.”

“When the government came in to rescue these large banks, it essentially made whole all of the debt of the banks, including the subordinated debt,” Blume says. “That creates a problem for the future.” In theory, creditors holding subordinated debt are paid only after senior debt holders get all they are owed. That extra risk makes subordinated debt holders more cautious in lending. Rescuing them encourages them to take unwise risks, according to Blume.

One potential remedy, he says, is to require that financial institutions that take on debt incorporate provisions to automatically convert it to shares of stock under certain crisis conditions. Removing debt and expanding equity would bolster balance sheets. And this approach would impose discipline, since stockholders generally suffer bigger losses than debt holders in a bankruptcy or other crisis. Moreover, stock holders have a say in corporate governance; debt holders do not.

“That might solve a lot of the problems with the banking industry,” Blume says.