Why Do Bankrupt Firms Have Such Complex Capital Structures?

Complex capital structures are prevalent in many recent high-profile Chapter 11 bankruptcy cases.  One recent example is Toys ‘R’ Us, whose debt structure was, as characterized by Bloomberg Businessweek, “as complex and precarious as a Jenga tower. [1]” It included dozens of subsidiary entities, with separate debt facilities against entities owning the intellectual property, the real estate, and international operations, among other asset groups.  Why do capital structures become fragmented and complex in this way, and what are the implications for bankruptcy law?

In my working paper, Disagreement and Capital Structure Complexity,  I suggest one reason why a firm’s owners may have the incentive to engineer fragmented capital structures, using the idea that investors may disagree about the values of the various assets that make up the firm.  Fragmenting the capital structure horizontally—that is, pledging different assets and asset groups to different creditor classes—allows the firm to sell asset-based claims that are targeted to the investors who value those assets most highly. This targeting is good for the firm’s owners, because it minimizes the firm’s overall cost of capital.

This complexity can become costly, however, when firms encounter financial distress.  The same disagreement-driven fragmentation that allows the company to borrow more cheaply up front can lead to costly valuation disputes in and around bankruptcy, since creditors place a higher valuation on their own collateral than do the other creditors.  This can lead to valuation disputes that are socially costly in terms of professional fees, delays, and lost opportunities.  An example of this is the Energy Future Holdings case.  Following the company’s 2007 leveraged buyout, its capital structure was divided into two silos, with one silo of entities (called the “E” side) holding regulated power assets, and a separate silo of entities holding the non-regulated power assets (the “T” side), with separate creditor groups on each side.  The initial plan to avoid bankruptcy by converting E- and T-side debt into parent-level equity failed after more than a year of negotiations, as the two sides could not agree on the relative value of the two sides.  The resulting bankruptcy took over four years to reach plan confirmation and generated over $500 million in professional fees, to the detriment of creditor recoveries.

My theory has several implications.  One is that disagreement about valuation can lead to inefficient liquidation of viable firms, as creditors may prefer to walk away with the collateral they value highly, rather than fight for that value in a reorganization where the other creditors (from their perspective) are clinging to inflated valuations of their own collateral.  These kinds of forces may have been at play in the Toys ‘R’ Us case.  Certain secured term  lenders, including the hedge fund Solus Alternative Asset Management, believed they were better off monetizing their intellectual property collateral in a liquidation of Toys ‘R’ Us than backing a deal to keep existing stores open[2].  The recent cancellation of the auction of this collateral suggests that these lenders may have been more optimistic than the marketplace about the value of these assets.

From an academic standpoint, the theory provides a new answer to a long-standing question in the literature: Why do we need a corporate reorganization mechanism in the first place?  Why don’t we just auction all firms for cash?  Traditional answers to this question revolve around the need to solve illiquidity problems.  In the presence of disagreement, I suggest an alternative benefit.  A traditional Chapter 11 reorganization allows parties to walk away with securities backed by the assets they financed before bankruptcy, about which the creditors have higher valuations.  Thus, the creditors can continue “agreeing to disagree” about the values of their respective pieces, thus promoting settlement and avoiding socially costly valuation disputes.  This is not possible when the firm is sold as a going concern for cash, since cash has a commonly known value.

Finally, my model emphasizes that when capital structures are fragmented, bankruptcy costs can be driven by haggling and litigation over the value of the parties’ entitlements, even when the parties agree about what to do with the bankrupt firm.  This suggests that the time may be ripe for rethinking and improving the resolution of valuation disputes in bankruptcy.  In a related paper, published in the University of Pennsylvania Law Review[3], Edward Morrison and I review valuation opinions in bankruptcy cases.  We find large disagreements in asset valuations prepared by expert witnesses and persistent manipulation of the discounted cash flow (“DCF”) method.  Valuation experts often recommend methods that are inconsistent with finance theory and evidence, such as company-specific risk premia, contributing to these wide valuation gaps.  A more standardized and constrained playing field that relies more on objective, market-based evidence, and valuation models that are better-grounded in finance theory and evidence, could help mitigate these disagreement-driven bankruptcy costs.


[1] See, e.g., Susan Berfield, Eliza Ronalds-Hannon, Matthew Townsend, and Lauren Coleman-Lochner, “Tears ‘R’ Us: The World’s Biggest Toy Store Didn’t Have to Die”.  Bloomberg Businessweek, June 6, 2018.

[2] See  Matthew Townsend, Lauren Coleman-Lochner, and Eliza Ronalds-Hannon, “Toys ‘R’ Us Nearly Had a Deal to Save Itself”.  Bloomberg Businessweek, June 5, 2018.

[3] Kenneth Ayotte and Edward R. Morrison, Valuation Disputes in Corporate Bankruptcy 166 U. Pa. L. Rev. 1819 (2018), available at https://www.pennlawreview.com/print/?id=619.

This post comes to us from Professor Kenneth Ayotte at the University of California at Berkeley School of Law. It is based on his recent article, “Disagreement and Capital Structure Complexity,” available here.