Economic Brief

According to estimates done by researchers at the Richmond Fed, the federal financial safety net covered $25 trillion in liabilities, or 58 percent of all financial liabilities, at the end of 2008. Such expansion of the safety net has weakened market discipline and contributed to instability in the financial sector. Instead of attempting to address the "too big to fail" problem by breaking up large firms, for instance, policymakers ought to focus on credibly scaling back the safety net and making its boundaries transparent.

In this year's debate over financial regulatory reform, one objective everybody agrees on is ending "too big to fail." TBTF is a policy stance based on the belief that the failure of certain financial firms, under certain circumstances, would be unacceptably disruptive to financial markets and the broader economy. As a result, market participants have long believed that such firms and their creditors enjoyed implicit guarantees from the government. These beliefs have the effect of subsidizing leverage and risk-taking, distorting incentives in a way that contributed to the financial crisis of the last three years.

In the wake of the crisis, ideas for ending TBTF have ranged from the forcible break-up of large financial firms to the creation of a new government resolution authority charged with liquidating distressed firms in a way that ensures creditors bear losses. But to understand the effects of any such policy change, it is important first to understand the problem we are trying to solve. In this Economic Brief, we discuss the size of the TBTF problem — or, more generally, the problem of implicit government guarantees of financial market obligations. We argue that the financial safety net is a large and growing problem. How and why it has grown are the keys to understanding the nature of the problem.

In an article published in 2002, one of the authors of this Economic Brief (Weinberg) and his colleague John R. Walter attempted a measurement of the financial safety net.¹ Using data from 1999 they sought to count the liabilities of private financial firms that had either explicit or implicit backing from the government.

Their estimates were based on conservative assumptions. They included in the safety net only the liabilities of those firms that enjoyed protection based on legislation, or on official actions and statements. Explicit safety net liabilities consisted of the insured deposits of commercial banks, savings institutions, and credit unions, as well as the private employer pension funds guaranteed by the Pension Benefit Guaranty Corporation. Meanwhile, the vast majority of the implicitly guaranteed liabilities were those of the government-sponsored enterprises — Fannie Mae, Freddie Mac, the Farm Credit System, and the Federal Home Loan Banks. But about a quarter of the implicit guarantees were the uninsured deposits of several large banks; federal officials stated in the 1980s that such firms would be treated as TBTF. Walter and Weinberg estimated that in 1999 the federal safety net totaled $8.4 trillion, or 45 percent of the country's financial liabilities.

How big is the financial safety net now? Walter and co-author Nadezhda Malysheva have updated these estimates to the end of 2008.² Their paper (available on richmondfed.org) continues to use the same conservative criteria that were used for the 1999 estimate — only firms that received protection by word or deed were included. This means, for example, that the 19 institutions subject to "stress tests" — many of whom were omitted from the 1999 estimate — were included.

Walter and Malysheva estimate that as of December 2008 the federal financial safety net protected $25 trillion in liabilities, or 58 percent of all financial liabilities, up considerably from 45 percent nine years earlier. These estimates are conservative, as we have argued, and exclude liabilities — such as money market mutual funds — that one might reasonably presume could receive support again, as they did in 2008.

How did the financial safety net come to cover nearly three-fifths of the financial sector? The arithmetic answer is the expansion in the portion of the financial sector benefiting from implicit guarantees — that is, guarantees that are implied or inferred rather than those that result from formally legislated programs, such as deposit insurance. In 1999, the implicit safety net covered only 18 percent of financial sector liabilities. At the end of 2008, 36 percent of the financial sector was covered by implicit government guarantees. In contrast, the portion of the financial sector covered by explicit guarantees fell from 27 percent in 1999 to 22 percent in 2008.

Why has the implicit safety net grown so large? We would argue that ambiguity about implicit guarantees sets up forces that inexorably expand the safety net over time. Letting a firm viewed as TBTF file for bankruptcy would risk a sudden investor retreat from similar firms, on the belief that the government might not support them either. This would just add to financial market volatility in an already volatile situation, which is why the urgency to protect creditors and counterparties becomes overwhelming during a crisis. Policymakers feel compelled to act in ways they find repugnant, even as they recognize the moral hazard costs.

The resolution of uncertainty about implicit safety net guarantees is thus biased toward intervention, which expands the safety net over time. Popular accounts view rescues as necessitated by the distressed firm's large and complex web of financial relationships, which are said to be too costly to unwind. But these direct domino effects have been largely absent in this recent crisis, as illustrated by the relatively smooth unwinding of Lehman Brothers' positions in bankruptcy. Rather, the volatility following Lehman's failure reflected investors reassessing whether other firms might or might not receive government support. The aspect of interconnectedness that mattered most in this crisis was financial firms' common reliance on an ambiguous and uncertain government safety net policy.

Many of the ideas put forward in recent policy debates would perpetuate that ambiguity. Giving the Federal Deposit Insurance Corporation (FDIC) or another government entity the authority to take a failing institution into receivership, culminating in liquidation, seems a natural extension of the FDIC's treatment of troubled banks. But the FDIC would be allowed to provide funds to the receiver that could be used to settle short-term debts as they come due. Even if shareholders are dutifully "wiped out" and the firm ultimately closed, the protection of short-term creditors weakens the incentives of the most critical liability holders. Despite their best intentions, authorities will inevitably err on the side of rescue, which will further weaken market discipline and lead to an ever-widening sphere of intervention and distorted incentives. A provision that provides for clawbacks of funds advanced that prove to be in excess of what claimants would have received in a bankruptcy can restore some discipline to the process. But there is likely to be enough uncertainty — and contentiousness — about clawbacks that the end result will still be that those claimants benefit at the margin from the use of public funds.

It does not really matter whether this additional support is "pre-funded" through fees on financial institutions or funded after the fact. The discretionary use of funds by the government to shield certain creditors from harm is what distorts incentives and leads to excessive risk-taking.

Real regulatory reform would sharply restrict the power of government entities, including the Fed, to provide funds to failing institutions. After the banking crisis of the 1980s, the FDIC Improvement Act (FDICIA) took a step in this direction by imposing a "least-cost resolution" requirement on the FDIC and restricting the Fed's ability to lend to failing banks. But FDICIA provided a "systemic risk" exemption and preserved the Fed's Section 13(3) emergency lending authority. This meant that the creditors of large institutions still had a reasonable expectation of support in the event of a crisis. Real reform would include extending the FDICIA least-cost resolution constraints to all FDIC-resolved failures and abolishing the Fed's 13(3) powers.

Granted, a resolution procedure that does not allow special treatment for short-term creditors could make such funding more expensive. But surely this crisis demonstrated that short-term credit was too cheap and plentiful, and that the financial system was too fragile as a result. Reducing the use of short- term credit to fund illiquid assets would enhance financial stability.

Our discussion of the TBTF problem has not really been about how big a firm is, or even about whether or not a firm fails. Rather, our focus is on ambiguity surrounding the use of public funds to either prevent a failure or soften the blow in the event of a failure. This ambiguity would remain a problem even if large firms were broken into smaller pieces. Small firms that rely too heavily on short-term financing will still be subject to runs by their creditors. And runs will still be viewed as having the potential to destabilize markets more broadly. So from the point of view of financial stability, the benefits of a policy that directly limits the size of firms are doubtful.

Resolution policy is the heart of regulatory reform. Getting it right is essential to credibly limiting the size of the financial safety net, restoring market discipline, and truly ending "too big to fail." If we don't, financial crises are bound to recur.

Jeffrey M. Lacker and John A. Weinberg are president and senior vice president and director of research, respectively, of the Federal Reserve Bank of Richmond. The views expressed are their own.