by Dr Xinyu Hou, Research Associate, Cambridge Centre for Finance and Cambridge Endowment for Research in Finance
In the early days of Federal Express, the company once was down to $5,000 in its checking account, not able to cover the $24,000 jet fuel bill due the following Monday. With the firm hanging on the edge, the founder, Fred Smith, flew to Las Vegas and gambled the $5,000 to $32,000 – saving the business at a critical time. When asked by another partner of the firm how he could do that, Fred shrugged, “What difference did it make? Without the funds for the fuel company, we couldn’t have flown anyway.”
Gambling, or in another word risk-taking, certainly made a difference – it saved the owners and avoided bankruptcy costs with some probability. It may even benefit other claimants, including the fuel company, who were unlikely to get much in bankruptcy. However, gambling oftentimes has a bad reputation as in asset substitution, which says that the shareholders would be willing to take on a project with higher volatility and even negative net present value, so that they can benefit from the upside of the volatility but substitute out the downside to the bondholders (thanks to limited liability).
So, when is gambling good and when is it bad? Phil Dybvig and I recently have a theory paper, “Gambling for Redemption or Ripoff, and the Impact of Superpriority“, that studies pure gambling by the firm using derivatives. Gambling using derivatives is a sharper tool, allowing the firm owners to have more control over the payoff distribution and without sacrificing project efficiency as in a typical asset substitution (for example, you can write a digital option which pays £100 20% of the time and 0 80% of the time). We have a simple single-period model which shows that the impact of gambling can be understood through two polar cases: gambling for redemption, which means gambling just enough to stay in business, is good for the owners, the creditors, and for overall efficiency, similar to the FedEx gambling. Redemption is optimal for the owners when the owners’ net gain from bankruptcy is negative (in the simplest case, when firm’s continuation value in excess of cash, ie, the total non-cash asset value which reflects firm’s future investment decisions if the firm continues, is greater than the face value of the debt). Gambling for ripoff, which means taking a very big risk, benefiting the owners at the expense of the creditors and overall efficiency – like an “asset substitution on steroids.” Ripoff is optimal when the owners’ net gain from bankruptcy is positive (again in the simplest case, when their continuation value in excess of cash is smaller than the face value of the debt). This is because to obtain the continuation value, the owners need to repay the debt – a negative net gain from continuation. Therefore, the owners choose gambling that fails as often as possible, even if the short-term assets such as cash could cover the debt.
Gambling for ripoff is of special current interest because of controversial legislation in the US before the financial crisis that exempts repos and other derivatives from important provisions of bankruptcy, including the automatic stay and clawbacks, causing some people to call them superpriority claims. In the United States, it has traditionally been difficult to redeploy assets for large gambles. While the Common Law allows for asset seizure in satisfaction of debts, the bankruptcy automatic stay provision “stays” the assets in the firm’s estate, stopping the creditors from taking further actions against the firm. Even if the assets were seized before bankruptcy, seizure or sales can be clawed back by the court. One important purpose of these laws is to prevent “asset rat race” in which claimants rush to take a piece from the firm, similar to a bank run. This gives a breathing space to the firm when the firm is in financial destress and ensures orderly resolution. Consequently, any promise by the firm to transfer assets to pay off on a failed gamble would not be credible unless the gambling counterparties are sure that the firm will not be pushed into bankruptcy, and gambling is only operating at a limited scale. However, the superpriority treatment for derivatives sidesteps the laws. The owners of a shacky firm can now pledge their assets as their gambling bets because the gambling counterparties know that the assets would not be stayed in bankruptcy. With more funds to gamble, the paper shows that gambling for ripoff becomes more appealing to the owners.
This is consistent with the claim of Mark Roe, a Harvard law professor, who suggests that the superpriority laws have made the firms more fragile and accelerated the 2008 financial crisis. Roe argues that superpriority provides a cheaper way of financing, facilitating more liquidity that otherwise would not occur. This shifts the firms away from using traditional financing and lower the incentives of derivatives counterparties to monitor the firm. Because of the heavily used superpriority claims, the “too big to fail” problem was worsened. Our paper provides another angle to look at the impact of law. In our single-period model, superpriority laws seem good for firm owners because they enable large-scale gambling, but perhaps only because the amount of debt and the continuation value are both exogenous. We also provide a multiperiod model that endogenizes these variables and some others. The results show that superpriority typically reduces firm value because bond investors realise that superpriority increases the likelihood of gambling for ripoff, and this is reflected in bond pricing.
If the firm owners are potentially worse off because of the laws, as claimed in the multi-period model, the owners would have incentives to use more defensive measures (operating leverage, secured debt, short-term debt, and even repos) to protect against the laws. Also, the bondholders cannot rely on protections in bankruptcy through negative pledge covenants which preclude asset sales but may rely more on perfected security interests (collateral) which are still honoured under UCC Article 9. This is supported by some empirical evidence that shows an overall increase of these “more defensive assets” in recent years, but more rigorous empirical work needs to be done to test the theory.
References
Dybvig, P.H. and Hou, X. (2022) “Gambling for redemption or ripoff, and the impact of superpriority.”
Frock, R. (2006) Changing how the world does business: Fedex’s incredible journey to success – the inside story. Berrett-Koehler Publishers.
Myers, S.C. (1977) “Determinants of corporate borrowing.” Journal of Financial Economics, 5(2): 147-175
Roe, M.J. (2011) “The derivatives market’s payment priorities as financial crisis accelerator.” Stanford Law Review, 63(3): 539-590
Schwarcz, S.L. and Sharon, O. (2014) “The bankruptcy-law safe harbor for derivatives: a path-dependence analysis.” Washington and Lee Law Review, 71(3): 1715-1755