How to reduce risk on Wall Street? Make the banks pay.

By Matthew Richardson and Nouriel Roubini
Sunday, April 11, 2010; B01

Between the fall of 2008 and the winter of 2009, the world's economy and financial markets fell off a cliff. Stock markets in the United States, Asia, Europe and Latin America lost between a third and half of their value; international trade declined by a whopping 12 percent; and the size of the global economy contracted for the first time in decades.

When economists and Wall Street types toss around the term "systemic risk," that's pretty much what they're talking about. The particular risks that led to the crisis -- i.e., big institutions with too much leverage, too little capital and too many implicit and explicit government guarantees -- were not impossible to anticipate. (In fact, some of us warned about the financial pandemic that was to come.) Now, the question is: How do we keep this all from happening again?

To create a truly safe financial system, we have to focus on two goals. First, we have to drive a stake through the heart of the "too big to fail" mantra that only fattens our financial beasts. Second, we should stop focusing on the problems of individual banks and look at the broader risk that the largest and most complex financial institutions pose.

We can accomplish both goals by charging such institutions an annual fee, or tax, or surcharge, or levy, or whatever the politicians need to call it. The current reform proposals in Congress call for something like this, but they don't go nearly far enough. The amount of the fee would vary according to each bank or financial firm and would include two key elements: an insurance premium based on whichever of the institution's debts carry a real or implied government guarantee (akin to the FDIC system already in place), and a fee that reflects the institution's contribution to a potential large-scale, systemic crisis.

All large and interconnected firms will want to minimize the fee they must pay, so each will have an incentive to choose less risky activities and to take on less debt, leading to a safer and sounder financial system. And unlike the bills in Congress propose, the money collected should not go into a resolution fund to help wind down failed institutions; instead, it should compensate those who suffer the collateral damage from systemic financial crises -- the solvent financial institutions and businesses in the real economy that suffer when credit markets panic.

To understand why this initiative is needed, we need to remember what really led to the crisis. Systemic risk emerges when financial institutions don't have enough capital to cover their debts and their bets. As a result, when those bets go sour, the institutions fail or the credit markets freeze -- and without credit, commerce plummets and economies fall into recession. That is precisely what happened with some of our largest institutions: Fannie Mae and Freddie Mac, Lehman Brothers, AIG, Merrill Lynch, Washington Mutual, Wachovia, and Citigroup, among others.

At the heart of the problem were the incentives such institutions had to take on massive risks. These were not dimwitted or reckless bankers. They simply had access to cheap financing from capital markets because of implicit government backing (under the "too big to fail" mentality) or explicit support, as in the case of depository institutions or outfits such as Fannie and Freddie. So they exploited regulatory loopholes to take on trillions of dollars in one-way bets on financial instruments involving residential and commercial real estate and other consumer credit. These were largely safe investments, except if an severe economic downturn materialized. The bankers knew that all the benefits of these activities would accrue to their shareholders, while all the costs -- in case everything came crashing down -- would be borne by society.

Consider Federal Reserve Chairman Ben Bernanke's oft-cited analogy for why bailouts, however distasteful, are sometimes necessary. Bernanke has described a hypothetical neighbor who smokes in bed and, through his carelessness, starts a fire and begins to burn down his house. You could teach him a lesson, Bernanke says, by refusing to call the fire department and letting the house burn to the ground. However, you would risk the fire spreading to other homes. So first you have to put out the fire. Only later should you deal with reform and retribution.

But let's change the story slightly. If the neighbor's house is burning strongly, putting the fire out might risk the lives of the firefighters. You can still call the fire department, but instead of saving the neighbor's house, the firefighters stand in protection of your house and those of your other neighbors. If the fire spreads, they are ready to put it out. This approach could save lives, and it has the added benefit of chastening your guilty neighbor into refraining from smoking in bed, or perhaps installing new fire alarms.

This is the purpose of a systemic-risk fee on big financial institutions.

How would it work in practice? For the first part of the fee, the government would need to decide which bank liabilities -- whether foreign deposits, interbank loans, private debt, long-term debt, etc. -- have implicit or explicit guarantees. Then, as with FDIC insurance for domestic deposits, the government should charge the banks a premium for these guarantees. And when a financial institution goes bankrupt, it should go through a credible and pre-established resolution process -- a sort of "living will" arrangement -- in which the firm is not bailed out, but its liabilities that are not guaranteed by the government are converted into equity shares. In that way, creditors, not taxpayers, bear the costs of failure.

The second part of the fee is more critical. Banks and their lobbyists will say it is impossible to measure a particular firm's share of the systemic costs of a financial crisis. Don't believe them. It's not rocket science.

There is plenty of evidence and research on the magnitude of these costs covering many crises over a number of years. These costs are huge, because they include not just the price tag of government bailouts but the overall losses the economy suffers because of a banking crisis. Based on historical evidence of crises worldwide, it is reasonable to estimate a financial crisis coming once every 50 years and costing about 5 percent in forgone gross domestic product. Today, this would imply annual levy of $14 billion on the U.S. financial sector -- hardly chump change.

To determine how an individual firm contributes to systemic financial risk, we must be able to estimate the size of its liabilities, its leverage, how its losses mirror those of the overall financial sector in a crisis and how interconnected the institution is with the rest of the financial system. Except for this last factor, all can be calculated from publicly available data. Indeed, with our colleagues at the Stern School of Business, we did just that in a recent paper titled "Measuring Systemic Risk," which analyzed the risk for major financial firms in June 2007, before the crisis erupted. The six firms with the greatest systemic risk, estimated per dollar of each firm's capital, should sound familiar: Bear Stearns, Freddie Mac, Fannie Mae, Lehman Brothers, Merrill Lynch and Countrywide Financial.

Reducing or at least controlling systemic risk should be the top priority for regulators and lawmakers. Among the thousands of banks in the United States, almost all the assets are held by the 50 largest bank holding companies, which together account for approximately $14 trillion. Remarkably, two-thirds of that is controlled by just six financial institutions: in order, Bank of America, J.P. Morgan Chase, Citigroup, Wells Fargo, Goldman Sachs and Morgan Stanley. Add to this the government-sponsored entities such as Fannie and Freddie, plus the large insurance companies, and you pretty much have the entire U.S. financial system.

Ironically, this crisis has only worsened, rather than reduced, systemic risk by leading to the creation of more institutions large enough to post a threat to the overall system: Bank of America merging with Countrywide and Merrill Lynch; J.P. Morgan with Bear Stearns and Washington Mutual; and Wells Fargo with Wachovia. And MetLife, the largest U.S. life insurer, has entered into an agreement to buy AIG's international life insurer, increasing MetLife's assets by almost 15 percent.

It will not be so easy to put this genie back in the bottle, but we believe that our proposed fee would propel firms to become smaller, alter their risk-taking behavior and reduce their leverage. There is some hope. Several governments are already moving in this direction -- as the United Kingdom prepares for next month's elections, both major parties agree on the need for a bank tax, while Germany and France have signed on as well.

Unfortunately, these proposals, along with those in Congress, still focus on the firms themselves, rather than the costs their risk-taking behavior imposes on the rest of society. Unless the banks face the prospect of paying up front for the damage they could inflict, they will continue to pollute the financial system with unnecessary risk -- and increase the chances of another crisis.

From the archives: For recent Outlook coverage of Wall Street and the financial crisis, see "Teddy showed the banks who's boss. Will Obama?" by Simon Johnson and James Kwak and "Who's crazy now, Wall Street?" by Steven Pearlstein.

Matthew Richardson and Nouriel Roubini are professors at New York University's Stern School of Business and contributors to the forthcoming book "Regulating Wall Street: The New Architecture of Global Finance."

View all comments that have been posted about this article.

© 2010 The Washington Post Company