Insight

All Thing Considered: Examining the Too Big To Fail Subsidy

This week the debate around the “Too Big to Fail” problem (TBTF) heated up again with the release of a new report by the Government Accountability Office (GAO). The GAO report, at the request of Senator  Sherrod Brown and Senator David Vitter, delved into the question of whether and to what degree the largest banks in the U.S. receive an implied subsidy vis-à-vis other banks.[1]

Those listening to the debate should recognize that:

  • Government rescues over the past century have come to firms both large and small
  • More recently, extraordinary assistance came to nonfinancial firms (such as automakers) using much the same justification.
  • A variety of economic factors determine the large-small bank funding differential, a differential that exists between large-small firms across industries.
  • Any calculated differential is necessarily a snapshot, and overlooks the changing nature of this difference, which at times may even be negative. As the GAO report notes, “such advantages may have declined or reversed.”
  • TBTF, such as it is, is a problem of policymaker discretion manifest in bank creditor decisions. Banks themselves may benefit, but only passively.

These insights are built on a recent paper, regarding the question of TBTF. The belief that some firms should not benefit from an artificial advantage conferred by policy is unanimously held. But determining whether that advantage exists (and to what degree) is extremely difficult, as acknowledged in the GAO report.

The theory holds that in the case of a potential firm failure, policymakers will step in to provide assistance precluding failure, therefore making the firm’s investors or creditors better off than they otherwise would be. As a result, such firms will be able to borrow and raise funds more cheaply than others, distorting the otherwise efficient allocation of capital.

As Douglas Elliot points out, analytically the “gross subsidy for each class of deposit or security would be based on the multiplication of three factors:”

P(firm’s failure) x P(government rescue) x E(security’s relevant loss portion

conditional on rescue)

 

One would then subtract the gross subsidy (if any) of smaller banks to determine the net subsidy.

To see why such a calculation is so difficult, if not impossible, consider the history. The probability of a government rescue is determined mostly by policymaker discretion, and only in a limited indirect way by economic factors. For instance, the second half of the 20th Century may be littered with examples of government rescues, but they are nonetheless few and far between, not to mention highly erratic. Moreover, extraordinary assistance has come to firms both large and small, and to financial firms as well as to nonfinancial firms (e.g., automakers). Isolating those firms that might someday be “eligible” for a rescue from those that are not is beside the point, as each rescue must be determined and executed according to policymakers’ preferences and ad hoc authority at each time. That is, since there is no statute or law governing the use of bailouts, policymakers are only bound by their perceived authority at the time,[2] the political constraints which may or may not support rescue, and their risk-aversion to broader economic losses.

On the third of Elliot’s three factors, what seems straightforward is again complicated by the discretionary (some might say arbitrary) nature of firm rescues. The concept of a “bailout” is not monolithic in type, but may take many forms. Indeed, perhaps the most well known bailout in history – the Troubled Asset Relief Program (TARP) – started as one type of rescue and ended as another. Asset purchases, perceived by government officials as either unworkable or insufficient, were replaced with capital infusions by way of stock purchases and warrants. Loans, guarantees, direct capital infusions, deposit insurance – all forms of government assistance which may engender moral hazard problems and cause economic distortions, but with varying effects on different parties.

TARP itself was further modified on the fly to fund the rescues of GM and Chrysler. One upshot of this latter change was a reordering of the longstanding creditor priority from what would’ve been expected under typical bankruptcy procedures. So, to the extreme uncertainty surrounding the determination of the likely beneficiary, add the possibility that some creditors may actually be made worse off under a bailout scenario.

Of course a high degree of uncertainty does not translate to zero benefit. After all, financial markets price all kinds of highly variable, unpredictable, and even low-occurrence events. And as long as counterparties and investors perceive some probability of future rescue, interest rates for bonds and deposits will reflect that. But a differential in funding costs between large and small banks may not be (entirely) evidence of TBTF. Large and small firms exhibit this differential across industries (with banking somewhat in the middle of the pack). There are many economic reasons why a large firm can borrow at a lower cost than a smaller counterpart, including liquidity, diversification, lower information asymmetries, and so on. Describing that difference as evidence of TBTF means ignoring the “size-related factors independent of perceptions of government support.”[3]

For its part, the GAO notes its estimate of the TBTF subsidy is highly dependent on variable choice, and changes over time, even showing higher funding costs for the larger banks in 2013. They echo Elliot’s model by saying “changes over time in our estimates of the relationship between bond funding costs and size may reflect changes in investors’ beliefs about the likelihood that a bank holding company will fail, the likelihood that it will be rescued by the government if it fails, and the size of the losses that the government may impose on investors.” They go on to say they “cannot precisely identify the influence of each of these factors.”

Eliminating (or at least reducing) use of bailouts and their attendant economic distortions in the future is a function of decreasing discretion: the more policymakers can tie their hands in a time of crisis, adhere to rules-based policy, and provide a credible procedure for the resolution of firms, the more appropriately markets will price funds to reflect economic risk.



[1] This report is actually the second part of a larger project: the first piece was released last November and sought to quantify the value of the (explicit) support extended by the federal government during the financial crisis.

[2] Of the Dodd-Frank Act, many (including its namesake authors) argue that it’s most important aim is in fact to prohibit use of bailouts in future crises. Whether it accomplishes this aim or not, extraordinary assistance in the form of bailouts (contra forms of deposit insurance) has arguably never been explicitly authorized by law, but rather read into non-bailout authorities in existing law (e.g., the Federal Savings & Loan Insurance Corporation rescue under the “essentiality doctrine.” Thus one must consider the likelihood of extraordinary assistance as determined by factors other than existing law.    

[3] Randall Kroszner, “A Review of Bank Funding Cost Differentials,” Presented at “Too Big to Fail” and Its Implications on Bank Funding Costs, NYU School of Business, October 8, 2013, http://www.stern.nyu.edu/cons/groups/content/documents/webasset/con_044532.pdf

Disclaimer