Up in Smoke: Bankruptcy by Contract In the Legal Cannabis Industry
by
William A. Organek*
Access to the federal Bankruptcy Code cannot be modified by contract. Its procedures are meant to preserve the value of a bankrupt’s limited assets for all claimants, not just those who negotiate for protection. While critics of this inflexibility argue that many debtors and creditors would benefit from contractual modifications to the Code’s one-size-fits-all approach, such agreements are rare because courts universally reject them. However, the state-legal cannabis industry is different: since cannabis is federally illegal, the protections of the Code, as well as its prohibitions on bankruptcy contracting, are entirely inapplicable to industry participants. This Article exploits this anomaly and leverages a novel, hand-collected data set—consisting of almost 75,000 pages from 1,167 publicly-filed documents disclosed by thirty-four publicly-listed cannabis companies—to discover whether, and how, companies use this unique chance to contract around insolvency. It concludes that parties largely ignore this opportunity, but that some classes of participants make advantageous bankruptcy contracts.
This Article offers five key findings. First, industry participants are aware, and largely publicly disclose, that they cannot access the Code. Second, despite this awareness, the data suggests cannabis companies are mostly unwilling or unable to include contract terms that might replace or improve upon correlated provisions in the Code. Only around 6.6% of cannabis industry contracts contain such provisions. Third, despite this overall rarity, bankruptcy contracting is present in around 29% of secured lending contracts for cannabis companies, meaning it is more than four times more likely in these contracts than overall. Fourth, companies more frequently rely on structural mechanisms to allow them (or their counterparties) to take advantage of the absence of bankruptcy protection for the industry. Around 44.8% of documents examined include provisions to tailor capital structures or business operations in this manner, though only around 6.5% employed marijuana-specific strategies. Finally, despite predictions or suggestions by academics who support bankruptcy contracting, no company issues any form of exotic securities that would contractually automate parts of the insolvency process.
Bankruptcy contracting could be more common in secured lending agreements because of lenders’ heightened negotiating leverage and sophistication, greater financial stakes, and longer-term relationship with the borrower. However, for most parties, it may be that informational and transactional costs make bankruptcy contracting inefficient. Forms of bankruptcy structuring might serve as a cheaper alternative. By showing how legal cannabis companies use, or fail to use, their freedom of contract to tailor results in insolvency, this Article provides evidence of the limits and potential inequities of bankruptcy contracting.
Introduction
Frank Arenas, a disabled pensioner, started a small agricultural and retail business with his wife Sarah.[1] They used one unit of a two-unit commercial property they owned in Denver to grow crops and leased the other to a company, DPG, that would sell their crops.[2] The Arenases obtained all required governmental licenses and operated their business legally and successfully. But after some time, disputes with DPG caused Frank and Sarah to face large claims for damages and attorneys’ fees.[3] With insufficient assets to pay these bills, they did what many similarly “honest but unfortunate”[4] companies and individuals do under such circumstances: they filed for bankruptcy. Perhaps the Arenases believed relief would be forthcoming because insolvent debtors are almost never denied access to bankruptcy courts.[5] Yet this was not the case because the crop the couple grew and sold was state-legal and state-licensed marijuana.[6] Although licensed marijuana businesses may be legal on a state level, they are illegal on a federal level.[7] Unfortunately for Frank and Sarah Arenas, the Bankruptcy Code is federal law.[8] The court in the Arenas’ case therefore reasoned that “while the debtors have not engaged in intrinsically evil conduct, the debtors cannot obtain bankruptcy relief because their marijuana business activities are federal crimes.”[9] Thus, a trustee could not be appointed to sell their illegal assets or distribute the proceeds of illegal crimes; nor could the Arenases be permitted to manage their company and continue to violate federal law with the imprimatur of a federal court.[10] The Arenases were “unfortunately caught between pursuing a business that the people of Colorado have declared to be legal and beneficial, but which the laws of the United States—laws that every United States Judge swears to uphold—proscribe and subject to criminal sanction.”[11] As a result, their bankruptcy petition was dismissed.[12] They—along with every state-legal cannabis bankrupt before and since[13]—were ineligible for federal bankruptcy protection.[14]
This prohibition is uniquely anomalous in American bankruptcy law. The goal of this Article is to examine the legal cannabis industry’s contractual response to the prohibition. No other industry is comparably foreclosed from bankruptcy protection—far from it.[15] Instead, use of the Bankruptcy Code, without alteration, is forced upon individuals and companies, who are not permitted to voluntarily waive their rights to file for bankruptcy or alter even part of its mandatory rules.[16] Yet the legal marijuana industry is effectively forced to waive bankruptcy protection. Scholars have written for decades about the potential benefits and drawbacks of amending, modifying, or waiving some or all aspects of bankruptcy’s otherwise comprehensive and mandatory scheme—so-called bankruptcy contracting.[17] But, because access to bankruptcy cannot be waived, there have until now been few direct opportunities to test their propositions. The blanket prohibition on bankruptcy protection for the legal marijuana industry inadvertently grants companies freedom to create contractual alternatives to bankruptcy.
Using a novel, hand-collected data set of material contracts, and public filings made by large industry participants, this Article investigates whether, and how, companies modify their contracts in an industry where the provisions of the Bankruptcy Code are almost entirely inapplicable. This research bridges the gap between theoretical arguments and real-world practice. By hand-coding analyzing almost 75,000 pages of 1,167 publicly-disclosed documents by thirty-four publicly-listed cannabis companies, this Article adds rare empirical data to the typically speculative debate on the merits of contractual alternatives to bankruptcy.
The Article presents five key findings from the data. First, this data set indicates that marijuana companies are typically aware of their inability to access bankruptcy protection. Second, and counterintuitively, despite this awareness companies rarely alter their contracts to provide alternative mechanisms for resolving financial distress. Only about 6.6% of contracts contain tailored provisions to address insolvency in the absence of the Bankruptcy Code, making bankruptcy contracting unusual even in the one industry where it could be commonplace. Third, the bankruptcy contracting that does exist is largely concentrated among one type of creditor: secured lenders. Here, bankruptcy contracting still remains rare, with around 29% of secured lending contracts containing insolvency-specific remedies that would alter, amend, or even incorporate remedies that would be available under the Bankruptcy Code. Notably, this 29% prevalence in secured lending contracts is more than four times greater than the 6.6% prevalence in all contracts. Fourth, companies more frequently tailor their business operations or capital structures in response to the inability to file for bankruptcy rather than using specific contractual language to address their lack of bankruptcy protection. Overall, this alternative strategy is used in around 44.8% of contracts. Yet 85.5% of these contracts (i.e., 447 of 523 documents) structure around bankruptcy by trying, with varying degrees of success, to “opt out” of any potential violation of US cannabis law. Only around 6.5% of documents use marijuana-specific structuring strategies. Finally, despite predictions or suggestions by bankruptcy scholars, no company examined issues any form of exotic security meant to contractually automate portions of the bankruptcy process. The greater (though still low overall) prevalence of bankruptcy contracting among secured lenders, and the overall dearth of bankruptcy contracting generally, adds useful data to debates about the role and limits of bankruptcy contracting.
The inability of parties to waive bankruptcy’s mandatory rules is understood as essential to its operation. Without a collective, compulsory system, creditors would race each other to the courthouse to collect on the remaining assets of a dying company. This process would hasten the company’s demise, waste value, harm some creditors, and increase borrowing costs. But critics question the necessity of a compulsory system, noting that freedom of contract and avoidance of mandatory systems stands behind almost all other aspects of corporate law. These contractualists argue at least some debtors and creditors might benefit from a customizable set of rules through lower credit costs and more efficient transactions. But despite these arguments and growing support for contractual solutions to a host of corporate law issues,[18] contracting out of bankruptcy has been consistently rejected by courts. Consequently, one frustrated proponent of bankruptcy contracting, who envisioned that parties would write bankruptcy contracts if able, complained more than twenty years ago that “[t]his prediction is not testable today because bankruptcy contracts are unenforceable.”[19] The situation is little changed since.
The state-legal marijuana industry thus provides a singular—perhaps unique—opportunity to test the predictions of supporters and opponents of bankruptcy contracting. Despite an enduring, and well-known in the industry, prohibition on bankruptcy filings for cannabis companies, cannabis businesses have boomed. Sales in the legal marijuana industry have grown from about $4.6 billion in 2014[20] to a projected $33 billion in 2023 (an amount that exceeds sales of chocolate, eggs, craft beer, topical pain relief, and even legal opioids).[21] Five public companies in the industry have market capitalizations exceeding $1 billion.[22] Moreover, much of this growth has been enabled by debt, even though many bankruptcy scholars would predict an industry without a compulsory bankruptcy system would struggle to find affordable financing. In 2018, less than 20% of capital raised by the industry consisted of debt, but by 2022 this figure increased to almost 60%.[23] The industry raised almost $8 billion in debt capital in 2021 and 2022 alone,[24] with debt composing 93% of capital raised by U.S. marijuana cultivation and retail companies in 2022.[25] State-legal marijuana is big, debt-fueled business—yet one, exceptionally, without access to bankruptcy.
This Article is built around novel, hand-collected data set consisting of almost 75,000 pages across 1,167 documents filed by publicly-listed companies in the legal cannabis industry with a market capitalization greater than or equal to $25 million. Despite being aware of the bankruptcy prohibition and the opportunities this opens for bankruptcy contracting, the vast majority of the industry’s publicly-filed material contracts do not contain any language that specifically accounts for this unique status. Many contracts note the illegality of the company’s business under federal law and therefore carve out compliance with federal drug law from standard, more general provisions requiring compliance with any applicable law. However, few contracts include tailored provisions addressing the inaccessibility of the Bankruptcy Code or alternatives to the Bankruptcy Code in its absence.
Of the small number of contracts that address the bankruptcy issue head-on, 37.6% of them are secured loan agreements. Contracts written by secured lenders may include bankruptcy-related terms because the secured lending parties are more sophisticated, and have greater financial and relational stakes, than other contract counterparties. Institutional lenders continue to shy away from the legal cannabis industry. This financing void has been filled largely by opportunistic private debt funds and sale-leaseback financing firms. Many such lenders use their heightened negotiating leverage and sophistication to obtain loans with strict covenants and control provisions. They are particularly incentivized to negotiate airtight documents because many of their marijuana industry loans are outsized, long-term bets on nascent companies in an uncertain industry, and some portion of these lenders may be unable to diversify away from industry or company risks.[26] For secured lenders underwriting the risk of a loan based in part on the collectability of its collateral upon insolvency, a more tailored contract could also lead to lower costs and might therefore be worth the costs entailed. As a result, some of these lenders contract for bankruptcy: they negotiate insolvency-specific remedies that vary from what might typically occur if a cannabis borrower were able to file for bankruptcy.
While bankruptcy contracting is quite rare, “bankruptcy structuring” is more common. Recall that bankruptcy contracting is any form of contracting that amends, modifies, or waives some or all of the provisions of the Bankruptcy Code that would otherwise apply to a company. Bankruptcy structuring, meanwhile, is a term used in this Article to refer to two potential strategies: “opt-out,” or “cannabis structuring.” Opting out means largely or entirely removing the risk of violating US drug law by refraining from establishing plant-touching operations in the US. Cannabis structuring refers to a small set of corporate law techniques used to isolate and mitigate the specific risks a creditor might face if a cannabis company became insolvent but was unable to use the Bankruptcy Code.[27] Cannabis structuring ensures valuable assets are kept together in blocs easily accessible to a small number of creditors. Overall, the bankruptcy structuring strategies of opt-out and cannabis-specific structuring are used in about 44.8% of documents. Opting out is the far more common strategy, but it may be of limited effectiveness because of the wide range of the federal drug laws and the prohibition on accessing bankruptcy courts for even marijuana-adjacent companies. It also is all but impossible for a marijuana grower or retailer to implement. The cannabis-specific structuring tools, meanwhile, are illuminating because they provide a potential second-best option for companies seeking some of the benefits of bankruptcy contracting without the informational and transactional costs entailed by full customization.
Finally, over the years there have been numerous academic suggestions for alternative forms of debt or equity issuance that could directly address insolvency risk without needing to rely on the Bankruptcy Code. However, no company examined in this Article uses any form of exotic securities. Securities issued in this industry are limited to standard and preferred equity, convertible debt, and standard secured and unsecured debt. Failing to use exotic securities could suggest that any potential benefits from issuing them might be outweighed by the reduced information costs for subsequent public market purchasers achieved through standardization.[28]
Collectively, the data presented in this Article suggests that, at least for the legal cannabis industry, bankruptcy contracting does not appear to play a major role in most transactions. However, its manifestation in secured lending contracts underscores the important role of secured lenders, and in particular non-traditional private capital and sale-leaseback financing, in riskier or more uncertain industries, as well as the limits on the protections afforded to them. This Article therefore contributes to the literature on bankruptcy contracting by providing rare empirical data on bankruptcy contracts in practice. As discussed above, the blanket ban on opting out of bankruptcy means that what little empirical study of the topic exists does not focus on the contracts themselves, for there rarely are such contracts. The author is aware of no other work that examines the specific contractual language used by parties that are empowered to waive their right to bankruptcy and thus able to make credible alternative contracts. The hope is that the empirical data and analysis provided in this Article may suggest places where bankruptcy’s mandatory rules could be fruitfully replaced by default rules. At the same time, an absence of bankruptcy contracting may suggest important roles for default rules. Perhaps the absence of contracting implies that a fixed rule—any clear rule consistently applied, with wide latitude as to particular content of that rule—is at times preferable to a customized deal. Uncertainty, rather than suboptimal rules, may sometimes be the bigger enemy of commerce.[29]
The remainder of this Article proceeds as follows. Part I examines the distinctive legal framework governing the legal marijuana industry. This begins with explaining the ways the federally-illegal but state-legal industry operates—who regulates it, how it raises funds, and what behaviors are seen as legally permissible (or not) in the industry. It turns next to an analysis of the current law with respect to bankruptcy filings in the industry, and briefly examines underused state- or international-law alternatives to bankruptcy and what might explain their limited uptake. With no satisfactory bankruptcy system available for the industry, bankruptcy contracting could present a beneficial alternative.
Part II presents this Article’s novel empirical contribution. It briefly describes the data set collection and coding process,[30] and provides summary statistic data on the industry and its contracts. It then uses this data to answer this Article’s motivating question: are companies drafting bankruptcy contracts, and if so, what provisions do they include? The data suggests contracts that expressly address potential bankruptcy-related issues or seek to embrace, modify, or reject particular provisions of the Bankruptcy Code are rare. Those bankruptcy contracts that are entered into are negotiated primarily with secured lenders—parties withheightened negotiating leverage and sophistication, greater financial stakes, and longer-term relationships with borrowers. This Part examines several examples of bankruptcy contract provisions and analyzes what parties might have hoped to gain from including them. It also looks at various structural mechanisms parties use more frequently than substantive contracting.
Part III steps back from the empirical data to a more theoretical perspective. It summarizes the rationale behind the current bankruptcy system, critiques of its mandatory regime, and why bankruptcy contracting is seen by some as an antidote to some of its problems. It also identifies key provisions of bankruptcy law that contracts might seek to replicate, modify, or eliminate in a world shorn of mandatory bankruptcy rules.
Finally, Part IV investigates the broader implications of this data. Normatively, it suggests this industry’s experience offers a potentially limited role for bankruptcy contracting. It also counsels that there may be multiple independent reasons—equity, efficiency, and certainty—to maintain bankruptcy’s mandatory nature in most circumstances. The Article then briefly concludes.
I. Marijuana’s Distinctive Legal Framework
This Part describes marijuana’s distinctive legal framework. It first explains that marijuana is legal under state law but illegal under federal law, and how this affects the structures these companies use to conduct business. It then turns to the consistent approach undertaken by the Department of Justice in opposing bankruptcy filings by cannabis companies and the ways that bankruptcy courts have largely (though not entirely) accepted these arguments. Finally, this Part briefly canvasses two insolvency resolution alternatives to federal bankruptcy law: state-law proceedings and international proceedings. It notes that these alternatives have remained mostly unused and suggests why the industry’s size and structure could explain their limited uptake.
A. Marijuana Businesses Grow Despite Legal Uncertainty
Marijuana is illegal under federal law, but legal under state law in many states. This creates a unique regulatory framework for marijuana, with implications for organizational, contract, and securities law, as well as for how marijuana companies are financed. This Section explores each in turn.
1. Dual Status: Illegal under Federal Law, but Legal under State Law
As a Schedule I substance under the Controlled Substances Act (the “CSA”), marijuana is recognized as a drug with “a high potential for abuse,” “no currently accepted medical use,” and “a lack of accepted safety for use of the drug . . . under medical supervision.”[31] It is illegal to possess, manufacture, distribute, or dispense marijuana (whether intrastate[32] or across state lines), or to possess with intent to do any of these.[33] The CSA makes it illegal for individuals or companies to “knowingly open, lease, rent, use, or maintain any place, whether permanently or temporarily, for the purpose of manufacturing, distributing, or using any controlled substance.”[34] Sales of items “primarily intended or designed for use” in manufacturing a controlled substance is prohibited,[35] as is making anything with reason to believe the item made will be used to manufacture a controlled substance.[36] Merely profiting from a violation of the CSA is similarly banned.[37] Violations can lead to fines, prison sentences, and criminal forfeiture.[38] Doctrines such as conspiracy, aiding and abetting, and accessory after the fact further expand the breadth of illegality to businesses that provide ancillary services to the marijuana industry.[39]
Cannabis has nevertheless been made legal in forty states and the District of Columbia for medical use, and twenty-three states and the District of Columbia for recreational use.[40] But cannabis businesses operating in these states do so at the sufferance of the federal government.[41] Policy has shifted greatly under different presidential administrations.[42] Meanwhile, any federal enforcement would be limited by the Joyce Amendment, a provision included in federal appropriations legislation since 2014 that prohibits the Department of Justice from using Congressionally-appropriated funds in bringing enforcement actions against state-legal cannabis companies.[43] This policy vacillation suggests that marijuana businesses continue to operate in an uncertain federal legal framework, despite state policy gains.
2. Legal Implications of Marijuana’s Dual Status
The dual status of cannabis—illegal on a federal level, but legal on a state level—increases the costs cannabis companies face while limiting their operational and financing options in several ways. First, interactions between federal and state laws increase regulatory uncertainty and thus the cost of operating in the industry. As a general matter, state law respects the contracts of approved marijuana companies where marijuana has been legalized. Despite potentially implicating aspects of federal law, companies in the industry regularly enter into purchase and sale agreements, secured lending agreements with property as collateral, intellectual property license agreements governing state-granted intellectual property protections, leases, employment agreements, and a host of other legal agreements. Breached contracts get resolved through courts, employees can bring wrongful termination suits, foreclosures are permitted, and many other cases are resolved as they would be in other industries. Thus, this is not an industry that could be said to operate in the “wild west”—there is plenty of order, and plenty of law, and, in most cases, legal results do not vary from the underlying legal entitlements merely because marijuana is involved.[44]
Nevertheless, state courts can vitiate contracts to engage in illegal activities as void against public policy. Although this problem has been mitigated somewhat in the handful of states that have passed laws to carve out state-legal violations of the CSA from this common-law doctrine, some businesses may still choose not to conduct business with marijuana companies.[45] Additionally, cannabis companies have no access to federal patent or trademark protections.[46] Directors and officers of cannabis companies could also be subject to suit for violating fiduciary duties to shareholders by engaging in federally illegal activity.[47] Even the interaction between two parts of federal law can increase costs for marijuana operators: cannabis companies are not permitted to take federal deductions for business expenses in furtherance of their federally illegal operations,[48] but are nonetheless taxed on their federal income. These and other complications likely impose substantial costs on cannabis operators.
Next, state-legal marijuana businesses are tightly regulated, increasing industry costs while reducing the number of entrants. For example, Massachusetts has a comprehensive regulatory framework regarding entry into the market. No cannabis business may operate without a license, which is granted on a discretionary basis.[49] Limits are placed on the number of licenses and square footage of grow sites that a given individual or entity may have.[50] Licenses must be renewed annually, with fees of up to $50,000/year/license.[51] Initial license applications require financial disclosures, criminal background checks, and escrowed funds to pay for dismantling of the business if necessary.[52] They also require the applicant to reach agreements with host communities, create a plan to address each of thirteen separate areas of concern for marijuana operations, and meet many other requirements.[53]
Regulations also vary widely from state to state[54] (and even by locality),[55] further increasing entry costs. Moreover, approval from the applicable regulator is typically required before marijuana licenses can be transferred to a third party, while some place additional restrictions on transfers or ban transfers entirely.[56] As a result, licenses are among the most costly, valuable, and illiquid assets a cannabis company holds.[57] Substantial expertise is required to effectively monetize marijuana company assets, making underwriting loans to the industry more difficult and reducing the avenues of capital available to it.
Third, cannabis companies are generally shut out from traditional sources of financing, limiting their ability to grow through normal financial channels. Only around 10% of the approximately 5,000 commercial banks in the US report having worked with cannabis companies in 2021,[58] and marijuana companies are generally shut out from credit card processing systems.[59] Finally, access to public equity is made more challenging because US stock exchanges, concerned they will be subject to prosecution, do not permit companies that “touch the plant” (i.e., growers, processors, or retailers) to list.[60]
3. Financial Implications of Marijuana’s Dual Status
Despite the challenges described in the prior subsection, financing remains available to cannabis industry participants. Cannabis companies have relied on two primary strategies to obtain public equity capital. The first is to simply not touch the plant in the US. Marijuana-related financing companies, companies that provide goods and services to marijuana companies, and even companies that sell cannabis in countries where doing so is federally legal, have all been able to list on US exchanges because they do not touch the plant in the US.[61] The second strategy is for plant-touching cannabis companies to go public in Canada (where marijuana is federally legal) and then obtain a secondary over-the-counter listing in the United States. This method reduces the potential availability of capital as compared to a more traditional listing and also increases compliance costs since listing requirements in both jurisdictions must be satisfied.[62]
Marijuana firms face limitations in attracting equity or operating capital. As a result, firms often turn to a number of unconventional financing options. Two important (though by no means exclusive) sources of debt capital are opportunistic private firms and sale-leaseback financiers. Opportunistic financing firms often focus specifically on the cannabis industry and invest through private equity or debt offerings.[63] Sale-leaseback financiers[64] offer another channel of debt capital by monetizing the real estate holdings of marijuana companies. By purchasing a marijuana company’s real property and immediately leasing it back to the company on a long-term basis at an above-market rent, the premium paid effectively acts as interest on a long-term loan.
Any attempt by financiers to take security interests in the assets of their marijuana borrowers presents issues because of the nature of the assets and the legal framework of marijuana companies. The exact means used, the special issues presented, and what this says about bankruptcy law are all explored in Parts III and IV. The complexity of operating suggests that for lenders to be successful in the space, they must have specialized knowledge not only of the nature of cannabis businesses and how best to underwrite their risks, but also the distinctive challenges faced when trying to collect on defaulted loans. The next Section reviews the caselaw on cannabis bankruptcies, analyzing the legal arguments and institutional forces that prevent cannabis companies from accessing bankruptcy protection.
B. Cannabis Companies Can Rarely Obtain Bankruptcy Relief
Courts have been remarkably consistent for more than ten years in prohibiting cannabis companies to obtain bankruptcy protection. The Department of Justice has also been universally opposed to marijuana bankruptcy petitions, despite the policy lurches that have typified the executive response to cannabis companies during this period. This Section examines the Department of Justice’s view on cannabis bankruptcy filings. It then turns to an examination of how courts have usually adopted this policy, while leaving small room for maneuvers that could overcome certain inequitable results for individuals but will likely prove of little help to most cannabis companies.
1. U.S. Government Policy and Court Adoption
The Department of Justice operates the U.S. Trustee Program, which is responsible for overseeing the bankruptcy system.[65] Trustees are permitted to intervene in bankruptcy cases to preserve the integrity of the bankruptcy system as a whole.[66] The U.S. Trustee has been a major opponent of bankruptcy filings by marijuana companies and individuals operating or profiting from marijuana businesses since at least 2011.[67] In 2017, the U.S. Trustee issued a memorandum explaining two reasons why it must oppose marijuana bankruptcies:
First, the bankruptcy system may not be used as an instrument in the ongoing commission of a crime and reorganization plans that permit or require continued illegal activity may not be confirmed. Second, bankruptcy trustees and other estate fiduciaries should not be required to administer assets if doing so would cause them to violate federal criminal law.[68]
The memorandum explains that any other position would cause parties to violate one or both of the U.S. Trustee’s rationales.[69] Continuing to operate the business would permit either the debtor-in-possession or a trustee to violate federal law. Meanwhile, even a trustee “who liquidated the fertilizer or equipment used to grow marijuana, who collected rent from a marijuana business tenant, or who sought to collect the profits of a marijuana investment” would breach the CSA.[70] As a result, the U.S. Trustee has challenged the bankruptcy filings of marijuana growers and retailers, marijuana-related businesses such as lessors to marijuana tenants,[71] sellers of hydroponic equipment that could be used to grow regular produce or marijuana,[72] companies that provide management services to marijuana sellers,[73] defunct marijuana businesses that might recover funds from active lawsuits,[74] and others.
Courts have largely deferred to arguments put forward by the U.S. Trustee seeking to prohibit cannabis bankruptcies. Reasons courts have denied bankruptcy filings generally fall within several interrelated categories: good faith,[75] feasibility,[76] trustee inability,[77] and unclean hands.[78] First, a bankruptcy court cannot approve a reorganization plan unless it is “proposed in good faith and not by any means forbidden by law.”[79] Since reorganizations of marijuana debtors would involve continuing violations of law, courts have held that it would be impossible to confirm a marijuana bankruptcy plan.[80] Next, bankruptcy’s feasibility requirement mandates that a plan of reorganization cannot be approved unless it is not likely to be followed by a liquidation or further reorganization.[81] Any plan a bankruptcy court approved that resulted in an operating cannabis business would inevitably continue to violate the CSA. Continued violations of the CSA, in turn, would necessarily lead to liquidation. Thus, it would be impossible for a company that derives its income from marijuana sales to propose a feasible plan.[82]
Cannabis companies are also denied bankruptcy relief because neither trustees nor debtors-in-possession are able to liquidate the assets of a cannabis company or continue the company’s operations without violating federal law.[83] Finally, courts dismiss bankruptcy petitions filed by cannabis companies under the doctrine of unclean hands. Bankruptcy courts are courts of equity.[84] Under the equitable doctrine of unclean hands, “a federal court should not, in an ordinary case, lend its judicial power to a plaintiff who seeks to invoke that power for the purpose of consummating a transaction in clear violation of federal law.”[85] Since cannabis is federally illegal, the argument goes, all cannabis debtors have unclean hands and courts of equity should deny bankruptcy relief to them (and, correspondingly, to any creditors who might seek assistance under the Bankruptcy Code).[86] In the view of the U.S. Trustee and courts, therefore, there is no shortage of reasons to deny marijuana companies bankruptcy relief.
Despite the success of the U.S. Trustee, several courts have, in well-reasoned opinions, analyzed the flaws and “creeping absurdit[ies]” of a blanket prohibition on marijuana bankruptcy filings.[87] Nevertheless, there has not been a single reported case overruling the prohibition on bankruptcy filings for companies actively involved in the marijuana industry. Bankruptcy courts have, at times, tried to circumvent this blanket prohibition, but the alternatives they offer are limited. Under current law, cannabis companies may be able to proceed in bankruptcy if they fully cease their cannabis-related operations, divest all marijuana-related assets, and forswear future cannabis-related ventures. Even under such strict guidelines, however, the U.S. Trustee would likely still object to a bankruptcy filing, and a judge might still dismiss the case. Perhaps more importantly, this narrow path would be impractical for an operating company whose sole, or even primary, business is marijuana-related.
Courts acknowledge the bright-line rule adopted by the U.S. Trustee, but at least some believe “[t]here may be cases where Chapter 11 relief is appropriate for an individual or non-individual entity directly engaged in a marijuana-related business.”[88] For example, bankruptcy courts are sometimes willing to permit a debtor, especially a sympathetic individual debtor, to propose a plan that does not rely on income received in violation of the CSA.[89]
At other times, courts find it possible for businesses that do not touch the plant and for which cannabis-related income is only a small part of their overall revenue to propose plans that do not rely on income that violated the CSA. Courts have approved bankruptcy filings of landlords that have rejected their leases to cannabis companies,[90] or have permitted debtors to resubmit plans that exclude any marijuana-sourced income from the funds to be used to make payments under a plan of reorganization.[91] Of course, this type of strategy would be of little use to a marijuana retailer or grower, who could not easily remove cannabis from its business entirely.
The difficulty of disentangling a business from marijuana-sourced income does not only arise for plant-touching businesses. For example, the chapter 11 case of Way to Grow concerned the bankruptcy filing a seller of indoor hydroponic and gardening equipment that did not itself sell marijuana, but deliberately targeted the marijuana industry.[92] In dismissing the bankruptcy case, the court observed that it would be extremely unlikely the company could structure a viable business that did not depend on marijuana customers.[93] Thus, results vary on a case-by-case basis, with a subjective judgment regarding the centrality of marijuana to the business, rather than a simple test of whether a company touches the plant or not, determining eligibility for bankruptcy.
Some recent decisions have gone further, arguing “the mere involvement of marijuana-related assets, income, or connections to the debtor, is not dispositive of whether a particular case is permitted to proceed.”[94] Others have challenged the U.S. Trustee’s inflexibility and their view on whether the ban on a plan being proposed “not by any means forbidden by law” refers to the plan’s contents or the means of its proposal.[95] Finally, a decision in early 2023 appeared to provide the most direct opening to date for cannabis companies to file for bankruptcy. In Hacienda, a plant-touching debtor previously operated a cannabis wholesale manufacturing and packaging business known as Lowell Farms, which ceased operations more than a year before its chapter 11 filing but still retained intellectual property.[96] Lowell Farms was sold to a publicly-listed but thinly-traded Canadian cannabis firm (“CanCo”) in exchange for equity in CanCo.[97] Following this transaction, the debtor filed for bankruptcy, proposing to pay creditors through the sale of the debtor’s stock in CanCo over time.[98]
The U.S. Trustee objected, but the court rejected their arguments. It reasoned there would be no ongoing sale of marijuana, and any remaining inventory could be disposed of by federal authorities.[99] Next, the court explained that Congress “did not adopt a ‘zero tolerance’ policy that requires dismissal of any bankruptcy case involving violation of the CSA . . . .”[100] Therefore, courts have discretion as to whether or not to dismiss marijuana cases.[101] Since “the only thing for the Debtor to do is to sell its stock in [CanCo] . . . and then use the proceeds to pay creditors,” the case was permitted to proceed.[102]
Although this case represents the most direct opportunity a cannabis company has had to use the bankruptcy system, it should provide little comfort for many marijuana businesses. First, the opinion was careful to note that “[i]llegal activity can be cause for dismissal in appropriate circumstances [under § 1112(b)’s dismissal for cause provision and] to preserve the integrity of the bankruptcy courts . . . .”[103] Rather than broad permission for cannabis companies to file for bankruptcy, Hacienda mostly held that cannabis bankruptcy petitions could be permitted—or disallowed—on a case-by-case basis.
Second, and more fundamentally, it appears the conditions that might need to be met to fall within the holding of Hacienda would be challenging for cannabis businesses to achieve. Hacienda seems to require a complete cessation of any cannabis business, destruction of any remaining marijuana inventory, and a sale to a non-US company. The first two parts could be value-destructive for any existing cannabis business and would likely make it impossible for the business to continue to operate. The third, which appears necessary to avoid liability under the CSA provisions that prevent the proceeds of marijuana sales to be used to pay creditors,[104] could also result in value destruction through requiring a forced sale on a quick timeline—precisely the fire-sale discount bankruptcy is meant to avoid.
Finally, it is not immediately clear that other courts would accept the logic of Hacienda. Each step of the transaction is tainted by association with marijuana proceeds, and thus plausibly violates the CSA. It is also difficult to see how this scheme would not somehow fall within the wide reach the aiding and abetting or conspiracy portions of the CSA. In short, if Hacienda offers salvation from bankruptcy excommunication, the rites needed to obtain such salvation may rarely be accessible to the typical cannabis company.[105]
C. Alternatives to Bankruptcy—State and Foreign
Federal bankruptcy protection may be unavailable for cannabis companies, but they are not wholly without options for managing insolvency. This Section briefly examines two alternatives—state-law insolvency procedures and use of the Bankruptcy Code’s international debtor provisions. Yet as this Section will explain, these are poor substitutes for many of the key benefits of the Bankruptcy Code: national jurisdiction, an automatic stay, asset sales free and clear of liens, rejection of contracts, and others.
1. State-Law Insolvency Procedures
Pioneering work by Edward Morrison has shown that the vast majority of business failures in the United States are not resolved through chapter 11, or any other part of the Bankruptcy Code. Instead, “[t]he vast majority of small businesses resolve distress under state law.”[106] For smaller companies (a) with fewer secured lenders, (b) where lenders can obtain reliable information about the value of their collateral, and (c) where insolvency is more likely to result from a gradual decline rather than a sudden shock, state law procedures might be more efficient than federal bankruptcy.[107]
Professor Morrison’s article focuses on assignments for the benefit of creditors (“ABCs”). In an ABC, a distressed company transfers its assets to an assignee, who then either liquidates the company or sells it as a going concern and distributes the proceeds to creditors, though state law varies widely in the duties and responsibilities of parties in ABCs.[108] Another popular state-law alternative to bankruptcy is the receivership, which unlike an ABC’s voluntary regime, is generally imposed on a defaulting debtor by a concerned lender or creditor.[109] Receivers obtain control of the business and replace existing management to either run the business as a going concern or liquidate it; however, the high costs of receiverships can often lead to liquidation.[110]
At first blush, these state law alternatives may appear to solve the cannabis company bankruptcy prohibition conundrum. For some cannabis companies with few or no secured lenders, smaller information asymmetries, and little risk of a sudden implosion, it could even seem as though state-law procedures might improve upon insolvency under the Bankruptcy Code (were it available). However, this fails to account for empirical, doctrinal, and jurisdictional aspects of major cannabis companies that make state-law proceedings a poor fit.
Initially, state-law procedures tend to be used by private companies that are far smaller than the major cannabis companies that are the focus of this Article, where large numbers of dispersed creditors and complex capital structures benefit from bankruptcy’s negotiation mechanisms. Professor Morrison’s work, by contrast, focuses small, privately-held businesses, with average annual sales of only $1.45 million.[111] In a similar vein, a study of all ABCs commenced in South Florida (Miami-Dade, Broward, and Palm Beach Counties) from 2012 to 2014 concluded that eighty percent of the debtors had liabilities under $5 million, sixty percent of these businesses were owned by a single individual or a married couple, and eighty percent were held by ten or fewer shareholders.[112] Moreover, unlike bankruptcy, ABCs may also be harder for publicly-owned corporations to initiate because under state law, corporations typically require approval of both the board and shareholders to commence.[113] Since receiverships typically displace existing management, larger companies tend to avoid them as well.[114] Empirically, therefore, ABCs and receivers would seem a poor fit for large, publicly-held marijuana businesses with substantial amounts of secured debt and assets scattered across the country.[115]
Next, the ability of marijuana companies to use state-law remedies is doctrinally questionable and varies from state to state. Washington state, for instance, has a comprehensive statute permitting ABCs and receiverships for marijuana business.[116] Moving south, Oregon also has a cannabis-specific statutory regime for receiverships (but not for ABCs), but one with powers and duties different from those offered in its neighbor across the Columbia River.[117] Several states expressly authorize marijuana ABCs, but only upon approval from the state’s cannabis regulatory authority,[118] and powers and duties vary by state.[119] Any receiver or assignee under state law would also, of course, risk forfeiture and criminal penalties under federal law. A fifty-state survey is beyond the scope of this Article, but suffice it to say that a multi-state cannabis company that sought to use state-law proceedings to address their insolvency would likely face substantial uncertainty, serious logistical hurdles, and high costs.
State law receiverships and ABCs would likely also be ineffective for major cannabis companies because these procedures lack the statutory authority and broad reach of the Bankruptcy Code. No state law proceeding can impose a nationwide automatic injunction on all collections efforts.[120] State law often includes remedies against fraudulent transfers,[121] but these vary from what are available on a federal level and from state to state.[122] State law preference protection is rare, making it less useful in cases where some creditors have been paid before others shortly before a bankruptcy.[123] One of bankruptcy’s primary advantages is that it allows assets to be sold free and clear of liens, which decreases uncertainty for buyers and limits the veto power of other creditors.[124] While some states permit assets to be sold free and clear of liens in a state receivership, this would only apply to liens on property within that state. State ABCs generally do not allow lenders that offer post-insolvency financing to obtain liens with priority over existing liens, and even if they did, such liens could only attach to property within that jurisdiction.[125] State-level proceedings also do not permit companies to reject unwanted executory contracts or leases, retain rights in intellectual property licenses, renegotiate union contracts, or exercise other powers that are only available under the Bankruptcy Code.[126] These and other unique aspects of the Bankruptcy Code[127] make state-level remedies a poor substitute for larger, more complex cannabis companies whose corporate structures, operations, and capital structures span across multiple jurisdictions and demand a unified solution to their insolvency issues.
2. Foreign Insolvency Procedures
One final alternative to bankruptcy that merits a brief mention because of the prevalence of Canadian-domiciled, US-operating entities in the marijuana industry is the possibility of using Chapter 15 of the US Bankruptcy Code. “Under chapter 15, courts in the United States freely recognize and enforce foreign insolvency proceedings of companies that have assets that are located in more than one country,” in order to facilitate cross-border insolvency and ultimately cross-border trade.[128] Thus, a company that files for bankruptcy in a foreign country can file a chapter 15 petition with a US bankruptcy court to obtain its assistance with the company’s US-domiciled assets.[129] Such assistance includes applying the automatic stay to US-based collateral, issuing necessary orders, granting injunctions against actions inconsistent with the foreign proceeding, and otherwise aiding the foreign proceeding.[130]
Importantly for cannabis companies, the structure of a chapter 15 proceeding defangs most of the U.S. Trustee’s arguments against permitting marijuana bankruptcies to proceed. Chapter 15 proceedings do not centralize the assets of a debtor in a fashion that requires a trustee to administer them.[131] In chapter 15 proceedings, no plan of reorganization is proposed, and no dismissal authority under § 1112 is applicable.[132] Thus, some commentators believe chapter 15 could be used as an alternative to chapter 11 relief for some Canadian-domiciled cannabis companies with significant US operations.[133] To date, however, there has been very little use of this strategy by Canadian-domiciled US cannabis companies.
While there may be many reasons why chapter 15 has not been used by many cannabis companies, four plausible explanations are suggested below. First, few companies are willing to be the first to experiment with this approach, since bankruptcy is often an expensive proposition and companies typically stray away from untested legal regimes.[134] Next, while chapter 15 may appear to be a good fit for marijuana bankruptcies because it does not contain a “for cause” dismissal standard, § 1506 permits courts to abstain from acting in a chapter 15 case if doing so would be “manifestly contrary to the public policy of the United States.” Although this exception is rarely applied in practice,[135] a cannabis chapter 15 might be just such a case. Even if a court did grant recognition, it could apply the public policy exception to particular aspects of relief requested, harming any chance of reorganization.[136] Third, cost is likely a factor: this process would require filings in two countries, almost certainly entailing additional legal and administrative costs. Finally, this strategy could only be used by companies with Canadian-domiciled operations. While it is possible the Canada-chapter 15 route could help many cannabis companies, and if this became a well-tested approach other companies might establish operations in Canada for these purposes, it would still likely remain an incomplete solution for bankruptcy relief for marijuana companies.[137]
This Part explored marijuana’s distinctive legal framework. Companies that operate in this industry face numerous challenges as a result of the dual status of marijuana—illegal on a federal level, while legal on a state level. This affects how cannabis companies are established, financed, and operated. Moreover, to prevent violations of the Controlled Substances Act from continuing under the aegis of bankruptcy courts, cannabis companies are foreclosed from accessing bankruptcy protection. Although some alternatives may exist under state or foreign law to address insolvency, these are imperfect remedies for large, publicly-traded cannabis companies with more complex capital structures. Nevertheless, the unique position of cannabis companies in the bankruptcy system creates opportunities for companies to make contracts around bankruptcy. The next Part explores whether companies in the cannabis industry actually engage in bankruptcy contracting by turning to a novel, hand-collected data set on the industry.
II. Cannabis Companies as a Testing Ground for Bankruptcy Contracting
As we have seen, cannabis companies are uniquely excluded from bankruptcy protection. While scholars have argued against the imposition of bankruptcy’s mandatory rules on all companies and in favor of permitting bankruptcy contracting, there has been little opportunity to test these arguments. Since cannabis companies are barred from accessing bankruptcy courts, no bankruptcy court would have an opportunity to invalidate a contract regarding insolvency for altering the rights that a contract counterparty would otherwise have under the Bankruptcy Code. In a sense, therefore, marijuana companies operate with a bankruptcy blank slate. As the door to the bankruptcy courthouse closes, another door—the door to contracting around bankruptcy—opens in its place. Cannabis companies therefore provide a quasi-natural experiment[138] through which to evaluate how companies respond when given the chance to contract around bankruptcy.
This Part briefly outlines the methods used to collect data on the marijuana industry and provides summary statistics on the industry and the Data Set,[139] and then investigates how cannabis companies actually use bankruptcy contracting. The data shows that cannabis companies know they cannot file for bankruptcy under current law but generally do not make specific changes to their contracts in light of this fact. Of the small number of contracts that do contain such changes, a large proportion of them include secured lenders as counterparties. Instead, cannabis companies more commonly rely on structural mechanisms to obtain favorable outcomes. Finally, contrary to the predictions or suggestions of some scholars, they never issue exotic securities that would use more objective criteria to dictate when they would be required to file for bankruptcy.
A. Data Collection Methods and Summary Statistics
This Section describes the methods used to collect and develop the database upon which the findings of this Article depend, as well as summary statistics regarding the nature of the companies examined and their contracts. Each is described in turn.
1. Data Collection Methods
The claims made in this Article are based upon a hand-collected data set (as more fully defined in Appendix A, the “Data Set”) consisting of 74,922 pages in 1,167 publicly-filed documents from thirty-four publicly-listed companies in the legal cannabis industry. To the knowledge of the author, this is the most comprehensive data set of the publicly-accessible legal documents for the marijuana industry ever compiled. First, a list of all publicly-listed cannabis companies with a market capitalization greater than or equal to USD $25 million was created. This list was cleaned to remove companies that were not relevant for purposes of this Article, such as companies with no US operations. The final list (the “Company List”) contained thirty-four companies.
All public documents of certain types filed between May 1, 2018, and May 10, 2023, for companies on the Company List were reviewed. Documents reviewed for each company included: 10-Ks (or equivalent if listed in Canada or a 10-K was not otherwise available), organizational documents, merger agreements, secured debt contracts, unsecured debt contracts, leases, and material contracts. Material contracts, in turn, were any contract that was publicly filed as part of a company’s 10-K or 8-K filings but which did not fit within any of the other preceding document categories, such as employment contracts or indemnity agreements.
Each document in the Data Set was reviewed by hand and coded according to set procedures. The goal of the coding was to answer two questions: how did companies alter their legal or capital structures to account for their ability to make bankruptcy contracts, and what kinds of substantive changes did they make to their contracts to account for their exclusion from the protections of the Bankruptcy Code. The data presented below is based on this analysis.
A fuller description of the methods used to build and code the Company List and the Data Set can be found in Appendix A.
2. Summary Statistics
a. Company List
The Company List contains thirty-four companies ranging from relatively small businesses to companies with multi-billion-dollar market capitalizations. Some continue to be extremely active in the marijuana market, while others appear to be companies situated toward the end of their life cycle. With one exception, all operate in multiple states.
Companies on the Company List can be generally broken down into three categories: cannabis operators (“Operators”), cannabis financiers (“Financiers”), and other cannabis-related companies (“Others”). The Company List contains twenty-four Operators. Operators consist of wholesale, retail, and vertically-integrated growers and dispensers. All but one of the Operators on the Company List are what the industry terms “multi-state operators,” meaning that they have cultivation and dispensary operations in multiple states. Since multi-state operators cannot ship inventory across state lines, they either operate their own grow facilities or purchase inventory from other growers within each state in which they operate. Finally, only Operators touch the plant, while the rest of the companies on the Company List do not. As directly plant-touching businesses, they would not be eligible for bankruptcy relief.
Next, Financiers consist of companies that specialize in providing financing to the marijuana industry. These companies originate loans either through sale-leaseback financing or through more traditional secured or unsecured financing channels. The five companies in this category are all cannabis-specific financiers, meaning they derive most or all of their revenue from providing capital to the cannabis industry. As a result, it appears unlikely the cannabis-specific financiers would be able to file for protection under the Bankruptcy Code.
Finally, five companies that do not fit within the prior categories, but are still on the Company List, are categorized as Other. Whether these Others would be eligible for bankruptcy relief would generally depend on the business models of each company, with the likelihood of access to bankruptcy inversely correlated with the extent to which the business depends on revenue from the cannabis industry. These Others either are not primarily cannabis companies but have made substantial investments in plant-touching cannabis companies in the US, are reliant upon marijuana sales for sales growth in their marijuana accessory business, or are developing marijuana-related intellectual property that does not at this time directly involve the sale of marijuana.
The table below provides a summary of the financial and operational statistics of the companies on the Company List:
Company Type | Count | Largest Market Cap ($mm) | Smallest Market Cap ($mm) | Average Market Cap ($mm) | Median Market Cap ($mm) |
Operators | 24 | $2,141.4 | $26.1 | $490.0 | $180.1 |
Financiers | 5 | $1,942.0 | $53.5 | $550.8 | $255.1 |
Other | 5 | $41,311.5 | $30.9 | $8,445.6 | $331.4 |
Total | 34 | N/A | N/A | $3,508.8 | $255.5 |
These statistics show that the typical size of listed cannabis companies far exceeds the size of companies typically handled under state-law insolvency resolution procedures. With a median market capitalization of over $255 million, the median company on the Company List is likely at least an order of magnitude larger than the companies examined by Professor Morrison, which had average annual sales of only $1.45 million.[140]
The average market capitalization of companies in the Other category is skewed due to one major outlier–Constellation Brands—a very large producer and marketer of various alcoholic beverages that made a multi-billion dollar investment in Canopy Growth, a large multi-state operator.[141] Without Constellation, the average market capitalization of this category drops from $8.45 billion to about $229 million (i.e., a reduction of more than 97%).
Companies on the Company List typically take one of several approaches when seeking funding from public equity markets: they file a primary listing in Canada and a secondary listing on over-the-counter markets in the US, they file on over-the-counter markets in the US only, or they file on major stock exchanges in the US. A small number of companies pursue a fourth option, discussed below. Only the first option is available for plant-touching companies. The table below enumerates the listing strategies undertaken by companies on the Company List:
Listing Type | Operators | Financiers | Other |
Primary Canadian, OTC US | 21 | 0 | 0 |
OTC US Only | 1 | 1 | 2 |
Major US Index | 0 | 4 | 2 |
Other | 2 | 0 | 1 |
Total | 24 | 5 | 5 |
The data demonstrates a largely bimodal distribution consistent with the regulatory environment for marijuana. Companies that service the marijuana industry turn first to the deeper capital markets available in the United States, while companies that touch the plant primarily list in Canada and offer shares on a secondary basis through over-the-counter markets. The two apparent exceptions to this are not really exceptions at all. Operators Tilray Brands, Inc. (third-largest Operator on the Company List, with a market capitalization of $1.7 billion) and Canopy Growth Corporation (sixth-largest Operator on the Company List, with a market capitalization of $694 million) have multiple listings, including on Canadian, US, German, French, and other exchanges. These companies eschew plant-touching operations in the US for building brand awareness and intellectual property in Canada with the goal of quickly dominating the US market if cannabis is ever legalized at the federal level. The creative structuring used by these companies, allowing them to get close to the plant without touching it, opens capital markets to them and likely makes a US bankruptcy filing more feasible as well. This structuring and its implications are discussed in more detail below.
Data on the Company List allows us to draw several conclusions. First, the market is likely too large to use state-law insolvency procedures, since, as described earlier, these systems are optimized for far smaller businesses with simpler capital structures and operations. Next, at least twenty-nine of the thirty-four companies on the Company List—the Operators and the Financiers—would almost certainly be ineligible to file for bankruptcy. Finally, public equity financing appears available for the industry, as shown by the existence and size of the publicly-listed companies. Unpublished results provide greater detail on the companies in the Company List, while more granular detail on the contracts used by these companies is provided in the next subsection.
b. Data Set
In the Data Set, a document, together with all amendments, is counted as a single document. Unpublished results from my analysis of the Data Set provide greater detail on the companies therein, with reference to their May 10, 2023, market capitalizations, their 2022 long-term debt figures, and the number of documents included in the Data Set for each company.
The next table offers a breakdown of the types of documents (collectively, the “Company Documents”) included and analyzed in the Data Set.
Document Type | Number of Documents | Percentage of Total |
Annual Reports | 120 | 10.3% |
Organizational Docs. | 104 | 8.9% |
Merger Agreements | 64 | 5.5% |
Secured Debt | 100 | 8.6% |
Unsecured Debt | 58 | 5.0% |
Leases | 38 | 3.3% |
Material Contracts | 683 | 58.5% |
Total | 1,167 | 100% |
Average long-term debt loads from 2019-2022 are provided below in thousands of dollars (not including the outlier Constellation Brands):[142]
2019 | 2020 | 2021 | 2022 | |
Average Long-Term Debt ($000 USD) | $154,670 | $147,604 | $332,440 | $377,318 |
Number of Reporting Companies | 15 | 20 | 29 | 27 |
In addition, the average (median) market capitalization of companies in the Data Set was $1,668.8 million ($223.5 million) as of May 9, 2023. After removing Constellation Brands, the single major outlier, the average (median) market capitalization of companies in the Data Set was $467.5 million ($220.2 million).[143] This generally demonstrates a growing industry of micro- and small-cap companies with increasing debt loads. It also suggests the absence of bankruptcy is not acting as a major deterrent for debt financing, market entry, or investment. Of course, we cannot be certain as to this latter point—how could we know whether companies might have been able to take on even more debt had they had access to the bankruptcy system? But two facts make this conclusion more plausible—one logical, one empirical. First, to state the obvious, the mere issuance of debt makes insolvency possible for any company. If lenders were seriously concerned that the absence of bankruptcy protection would make it impossible to make profitable loans, as a logical matter they would not lend billions of dollars to the industry. Second, an average debt load of $377.3 million on an average market capitalization of $467.5 million results in a book debt-to-market capitalization ratio of 80.7%. This is higher than the book debt-to-market capitalization value for public companies in industries that might be assumed to have similar or relevant characteristics, such as the alcoholic beverage (42.6%), agriculture (55.3%), food processing (43.3%), REIT (61.1%), retail (general) (49.8%), retail (grocery and food) (61.4%), and retail (specialty lines) (63.6%) industries.[144] The tobacco industry, meanwhile, has a book debt-to-market capitalization ratio of 115.1%.[145] Again, this data is suggestive of the fact that this industry appears to have substantial borrowing capacity in spite of its overall legal uncertainty and the specific absence of federal bankruptcy protection.
The data collected here further suggests difficulty in valuing and underwriting the risk of investing in and lending to cannabis companies. Market capitalizations may shift substantially, due to thin trading, secular undervaluation of cannabis companies because of its tenuous legal status, or other reasons.[146] To compare two similarly-valued and similarly-situated companies, as of May 9, 2023, Trulieve (an Operator with a primary listing in Canada and a secondary listing in the US) had a market capitalization of $1.051 billion, while Verano (another Operator with a primary listing in Canada and a secondary listing in the US) had a market capitalization of $1.089 billion. While the had similar market capitalizations, Trulieve had $1.262 billion in debt at the end of 2022, and Verano had less than half as much debt, $668 million. A more fundamental analysis of the difference between the two companies might elucidate reasons for this distinct treatment, but at first blush it appears that companies may be able to take on substantial debt loads that their market capitalizations may not support at least some of the time.
Finally, these summary statistics suggest that companies are not electing to eschew debt issuance simply because they are unable or unwilling to incorporate provisions reflecting the absence of bankruptcy protection for this industry.
B. Cannabis Company Contracts
This Section examines the contracts included in the Data Set to answer how cannabis companies are making ex ante changes to their legal or capital structures or substantively altering their contracts to account for the absence of bankruptcy protection and their ability to contract around bankruptcy. The data shows that cannabis companies (1) are largely aware of their unique bankruptcy status, but (2) generally do not tailor contractual provisions to capitalize on this fact, and (3) generally limit the tailoring that does exist to contracts with secured lenders. In addition, cannabis companies (4) more frequently rely on structural mechanisms to obtain favorable outcomes, and (5) never issue exotic securities to automate the bankruptcy process.
1. Awareness of Unique Bankruptcy Status
Companies in the Data Set all expressly describe their cannabis-related activities in their public filings. Of the 120 annual reports examined, only five fail to note that marijuana is federally illegal and that their activities could violate the CSA, and these companies all mention this risk in their most recent annual reports.[147] In fact, almost all companies that do not touch the plant in the United States and have made a primary listing in the United States still discuss the risks of CSA violations in their annual reports.
Seventy of 120 annual reports (58.3%), from eighteen of thirty-four companies (52.9%), in the Data Set expressly discuss the unavailability of bankruptcy in their industry. After accounting for companies that do not mention this risk in one annual report but mention it in another, twenty-three of thirty-four (67.7%) companies with annual reports in the Data Set expressly explain in at least one of their annual reports that bankruptcy is unavailable in the industry.
A typical discussion of the issue from the perspective of an Operator is provided in the 2021 10-K of Acreage Holdings, Inc., one of the largest multi-state operators. They explain that:
Because cannabis remains illegal under U.S. federal law, many courts have denied cannabis businesses bankruptcy protections, thus making it very difficult for lenders to recoup their investments in the cannabis industry in the event of a bankruptcy. If the Company . . . or any of the Subsidiaries were to experience a bankruptcy, there is no guarantee that U.S. federal bankruptcy protections would be available, which would have a material adverse effect.[148]
Financiers also recognize this risk and that they might be foreclosed from bankruptcy protection.[149] Some financiers also note that even if they could access bankruptcy protection, the disallowance of bankruptcy access for their tenant-borrowers would harm them as well. If their tenants became insolvent without access to bankruptcy’s tools to right their finances, this insolvency would “reduc[e] the probability that such tenant would be able to honor its lease obligations with us.”[150]
Eleven of thirty-four companies (32.4%) fail to expressly acknowledge that bankruptcy protection is unavailable to them in any of their annual filings. This group consists of seven out of twenty-four Operators (29.1%) and five of five Other companies (100%). The companies in the Other category are easiest to explain: none touches the plant in the US. For example, Bakhu describes itself as a company that has licensed technology to “mirror . . . [the] flavor, aroma, and CBD and THC potency qualities of the source plant cells without growing the plant. We do not now, and do not intend to, produce, transport, or sell cannabis or cannabinoids directly.”[151] Even without directly touching the plant, Bakhu is concerned enough about possible violations of the CSA that it included more than ten pages of potential risks related to marijuana’s legal status in its public filings. But it may genuinely believe that access to bankruptcy would not present an issue for it, and as a corollary that its bankruptcy contracts would not be enforceable. This is consistent with the observation that Bakhu takes no steps in any of its Company Documents to negotiate around insolvency.
Some companies, like Canopy and Crucial, do touch the plant, but not in the US. They devote many pages of their public filings to discussing the federally illegal nature of marijuana businesses and many correlated risks.[152] Nevertheless, they structure their business so they do not produce or sell any marijuana in the United States. Instead, they have grown by negotiating contingent purchase agreements, warrants, and other similar devices which will allow them to quickly enter the US marijuana market upon legalization.[153] Despite this hands-off approach, Canopy is one of the few companies that has actually negotiated specific contractual remedies upon insolvency, the particulars of which will be discussed in greater detail in the next subsection.
Finally, there are four plant-touching multi-state operators[154] that do not note the industry’s bankruptcy prohibition in their annual filings. This is true even though these companies expressly acknowledge the federally illegal nature of their businesses in their public filings and, for all but one,[155] in at least some of their contracts. The omission is puzzling, and likely represents a material risk factor that was not appropriately disclosed by these companies.
In short, 67.7% of companies in the Data Set expressly discuss the unavailability of bankruptcy in at least one of their annual filings. Seven of the remaining eleven companies (63.6%) do not discuss it for plausible reasons, while four companies (36.4%) do not discuss the issue without an apparent justification. It is also important to view how common it is for industry participants to describe the unavailability of bankruptcy in their public filings in context: the unavailability of bankruptcy for industry participants has been well-established and reported upon for years, and it appears implausible that a well-represented company with US-based plant-touching operations would not be aware of this issue. Therefore, we can conclude that most, if not all, companies are aware of this restriction, even if they failed to disclose it.
2. Limited Bankruptcy Contracting
Despite broad awareness and public acknowledgment of the unavailability of bankruptcy, companies in the industry generally do not tailor contractual provisions in their Company Documents to account for this fact. The table below shows the number and percentage of contracts that include, or do not include, contractual provisions of any kind that seek to modify or waive aspects of the Bankruptcy Code:[156]
Type | Count (%) of Total | Count (%) with Bankruptcy Contracting | Count (%) Without Bankruptcy Contracting |
Annual Reports | 120 (10.3%) | 0 (0%) | 120 (100%) |
Organizational Documents | 104 (8.9%) | 8 (7.7%) | 96 (92.3%) |
Merger Agreements | 64 (5.5%) | 7 (10.9%) | 57 (89.1%) |
Secured Debt | 100 (8.6%) | 29 (29.0%) | 71 (71.0%) |
Unsecured Debt | 58 (5.5%) | 0 (0%) | 58 (100%) |
Leases | 38 (3.3%) | 16 (42.1%) | 22 (57.9%) |
Material Contracts | 683 (58.5%) | 17 (2.5%) | 666 (97.5%) |
Total | 1,167 (100%) | 77 (6.6%) | 1090 (93.4%) |
Only 6.6% of Company Documents contain bankruptcy contracting of any kind. Nevertheless, this bankruptcy contracting is broadly dispersed among companies in the Data Set: 26 of 34 companies (76.4%) have at least one document that contains some kind of bankruptcy contracting. Interestingly, only five documents designated the use of a particular state-law insolvency procedure, perhaps reflecting a lack of knowledge of, or confidence in, such a process.[157] This data presents a puzzle for proponents of bankruptcy contracting, since companies should want to write bankruptcy contracts to obtain cheaper capital. Rationales for why bankruptcy contracting might theoretically be expected are provided in Part III, while reasons that could explain the empirical rarity of bankruptcy contracting are explored in Part IV.
3. Contracting is Concentrated in Agreements with Secured Lenders
Bankruptcy contracting is concentrated in secured debt agreements and leases, with smaller concentrations in organizational documents and merger agreements. Each of these is discussed in turn.[158]
a. Secured Debt Agreements
Bankruptcy contracting is present in 29% of the reviewed secured debt agreements. Since bankruptcy is not available for companies in the marijuana industry, terms in these agreements supplement, modify, or replace the state-law rules that might otherwise apply upon insolvency. Sometimes they replicate what would occur if protection under the Bankruptcy Code were available, while other times they create procedures more favorable to one party or another than what would occur under the Bankruptcy Code.
A credit agreement entered into among various members of the Acreage Holdings, Inc. corporate family as borrowers provides among the most aggressive examples of bankruptcy contracting.[159] Section 9.3 of this agreement provides that upon an Event of Default (such as failing to make timely loan payments or declaring insolvency), Lenders have the option to send Borrowers a notice (the “Sale Notice”) requiring Borrowers to auction one or more of their marijuana licenses and related real property.[160] Very similar to the current practice for sales under § 363, the auction must follow a pre-ordained process, requiring selection of an investment bank and strict timelines for completion of the auction.[161] The sale must include sufficient assets to repay all obligations owed to Lenders, “including any premiums, exit fees, penalties, and/or default interest . . . .”[162] If the sale of assets proposed by the borrower is insufficient to repay all amounts owed, then the sale must be for all assets which serve as collateral for the loan.[163] Finally, since the Lenders “would not have an adequate remedy at law” if the provisions of this section were breached, the Lenders are entitled to specific performance of the obligation.[164]
It is difficult to overstate how dramatic a departure Section 9.3 is from the typical operation of the Bankruptcy Code, both procedurally and substantively. Procedurally, the decision under the Bankruptcy Code to sell, or retain, property lies at the discretion of the debtor, not its creditors.[165] A sale would take place only following a hearing where other creditors could object because they believe the sales price is too low, the auction is not being run in good faith, the property being sold would be worth more to the debtor (and ultimately its creditors) if kept by the company, or many other possible reasons.[166] The sales timeline and process would typically be proposed by the debtor and approved by the court, with the goal of ensuring that the process is fair to all creditors. If a sale did not result in sufficient proceeds to repay all creditors (let alone one specific creditor), creditors would be unable to object. Finally, a specific performance right under a contract would be subject to the automatic stay, and a bankruptcy court would likely be unwilling to enforce it if doing so would harm other creditors.
The Lenders’ right to interest payments under the Acreage Holdings, Inc. agreement also substantively diverge from what they would be entitled to under the Bankruptcy Code. The Bankruptcy Code disallows claims for unmatured interest,[167] while the payment of premiums remains contentious and is often disallowed by courts.[168] Payment of interest on undersecured claims is also not permitted, and bankruptcy courts typically need to determine whether a claimant is oversecured or undersecured before authorizing certain interest payments.[169] The rate of interest paid to a claimant also varies, with interest paid at the statutory rate following a liquidation.[170] These procedural and substantive differences are exactly the types of things scholars predicted might be negotiated were bankruptcy contracting permitted. Similar language found in other secured credit agreements, which in bankruptcy would be subject to careful parsing and defenses under the Bankruptcy Code, would instead be far more likely to result in a payment obligation.[171]
Procedural and substantive differences for sales processes can be found in other agreements as well. For example, a Loan and Security Agreement entered into by Green Thumb Industries, Inc. provides that upon an Event of Default, Lender may require a public or private sale of Collateral, may bid for Collateral itself, and may “restrict the number of prospective bidders or purchasers, and require such bidders or purchasers to have certain qualifications” before allowing them to participate.[172] Bidding qualifications, minimum bid amounts, and other restrictions on who may bid in an auction are commonplace in bankruptcy; however, they are typically determined by the debtor, not by the secured creditor. Sometimes secured lenders that provide debtor-in-possession financing to a company already in bankruptcy will be able to negotiate with the debtor to set such terms as a condition of providing such financing, but here the terms of the auction are being set by the lender unilaterally and well before insolvency. These differences further demonstrate how lenders can use the absence of bankruptcy protection to improve their recoveries from borrowers in this industry.
Another notable contractual provision is included in a credit agreement entered into by Acreage and collateralized by intellectual property rights.[173] Upon an Event of Default, in addition to traditional acceleration remedies, the agreement sets forth an elaborate procedure for selling intellectual property assets at a set price within a fixed time period to a pre-designated third party.[174] If the pre-designated third party does not consummate the purchase, then “the Agent shall . . . direct the Borrower to cause [parent company Acreage Holdings, Inc.] to offer for sale” a fixed number of shares in Acreage Holdings, Inc., with proceeds of the sale being used to repay the loan according to a predetermined priority waterfall.[175] If either of these are insufficient to repay the obligations, then Lenders may exercise their rights with respect to certain intellectual property and the pledged shares of other subsidiaries.[176] This provision appears to strike a balance between strict lender control and the potential downsides of the types of withdrawal rights discussed by Professors Baird and Casey.[177] It also modifies the careful balance between the rights of intellectual property licensors and licensees established under the Bankruptcy Code, granting this secured lender rights at the expense of other creditors that they would not have under the Bankruptcy Code.[178]
A less drastic, but still material, divergence from the Bankruptcy Code can be found in a credit agreement by AFC Gamma, Inc., as Borrower. Section 8.12 of the agreement authorizes the Agent, upon instructions from the Lenders, to credit bid or purchase all or any portion of the collateral that secures the loan at any sale conducted under the Bankruptcy Code, under the UCC, or as part of any other sale proceeding.[179] “Obligations with respect to contingent or unliquidated claims [will be] estimated” for a credit bid or purchase.[180] But if doing so would “impair or unduly delay” any credit bid or purchase, “then such claims shall be disregarded, not credit bid, and not entitled to any interest in the Collateral that is the subject of such credit bid or purchase . . . .”[181] Similar language is found in other agreements as well.[182]
Bankruptcy judges are statutorily authorized to estimate the value of contingent or unliquidated claims if failing to do so would “unduly delay the administration of the case.”[183] They do so in accordance with the terms of the statute and precedent, with courts barred from estimating claims that are unenforceable and ensuring that no party receives more than it is owed.[184] The language in Section 8.12 closely mirrors the language of the Bankruptcy Code, but instead empowers the Agent and the Lenders to take on a role for themselves that is normally assigned to the bankruptcy judge under the Bankruptcy Code. Without the limitations imposed under the Bankruptcy Code or precedent, and with the estimation process run by an interested party rather than a neutral judge, it would seem possible that the process would not be run in the interests of preserving value for all creditors.
Terms in other contracts authorize lenders to have a receiver appointed by a court to operate the business and dispose of assets,[185] require an equity raise to repay a loan in full,[186] treat licenses differently from how they would be treated in bankruptcy,[187] require marijuana to be disposed of upon repossession in accordance with marijuana laws in the state where the property is located,[188] permit the appointment of a receiver with a state marijuana license to continue business operations,[189] and to take other actions that would be stayed or prohibited under the Bankruptcy Code. Inclusion of these types of provisions appears to indicate that some lenders negotiated different rights than they would be entitled to under the normal bankruptcy regime.
Although aspects of contracts that were altered by parties to affect the substantive rights that might apply in the event of insolvency, it is notable that parties rarely appear to use cannabis-related reasons to select any particular jurisdiction as the governing law for their secured loans. Governing law provisions are easily altered and generally accepted by courts, and as discussed section I.C.1. supra, some jurisdictions have more sophisticated marijuana regulatory regimes than others.[190] Nevertheless, the data shows that of the 100 secured loan contracts reviewed, forty-four of them were governed by New York law. The second-most commonly used jurisdictions were Illinois and British Columbia, Canada, with thirteen agreements each; Ontario, Canada, rounded out the most-used jurisdictions with twelve agreements. Only Delaware and Florida had five agreements each, with a smattering of other jurisdictions with fewer than five agreements each making up the rest. It appears unlikely that New York law was selected because of expertise in cannabis regulation generally or insolvency resolution in particular for several reasons. First, New York’s cannabis regulatory regime has been slow to commence and is mired in controversy.[191] Additionally, many of the contracts entered into designating New York law as the governing law were entered into before recreational marijuana was legalized in New York in March 2021, and almost all secured debt contracts in the Data Set were entered into before the first legal dispensary opened in December 2022.[192] Second, and perhaps more fundamentally, New York law is commonly used as the jurisdiction of choice for agreements of all stripes, so its use as the applicable law for cannabis financing law is unsurprising.[193] New York’s status as a go-to choice of law jurisdiction and its limited cannabis legal infrastructure during the relevant period suggests clustering in New York (and, to a lesser extent, Illinois[194]) has little to do with marijuana considerations. Choosing these jurisdictions, rather than ones with more sophisticated marijuana regulatory regimes, could be seen as a missed opportunity for easy bankruptcy contracting.
b. Leases
42.1% of the leases in the Data Set contain some kind of bankruptcy tailoring. Since cannabis companies use sale-leaseback financing as a major source of financing, these documents are often treated by parties more like parts of secured loan agreements than as standard leases. As such, parties sometimes negotiate special rights in bankruptcy to ensure their lease (i.e., loan) payments continue even though such payments would be strictly limited under the Bankruptcy Code.[195]
Fourteen of the sixteen leases that contained substantive bankruptcy provisions of the type discussed in this article were entered into with subsidiaries of IIPR, the largest sale-leaseback financier in the cannabis industry, as landlord/lender.[196]
Almost all IIPR leases provide that a trustee or debtor-in-possession in bankruptcy, “or another person with similar rights, duties and powers under Applicable Laws,” must follow certain steps to cure any defaults under the leases.[197] A bankruptcy judge would traditionally require “adequate assurance” payments to cure a defaulted lease, which amount would be set by a judge but which would typically be based on a balancing of the facts and circumstances of a case.[198] However, under the leases:
such adequate assurance shall include any or all of the following, as designated by Landlord in its sole and absolute discretion: (i) those acts specified in the Bankruptcy Code or other Applicable Laws as included within the meaning of “adequate assurance,” even if this Lease does not concern a facility described in such Applicable Laws; (ii) a prompt cash payment to compensate Landlord for any monetary defaults or actual damages arising directly from a breach of this Lease; (iii) a cash deposit in an amount at least equal to the then-current amount of the Security Deposit; or (iv) the assumption or assignment of all of Tenant’s interest and obligations under this Lease.[199]
Here, the Landlord is arrogating to itself the power to determine what constitutes “adequate assurance,” and doing so even if these payments would not be required under bankruptcy law. It also gives the Landlord the ability to require the debtor to assume or assign all obligations under the lease, a power which belongs only to the debtor under the Bankruptcy Code.[200] Unfavorable leases can typically be shed by a debtor in order to permit it to restructure its business, while favorable ones can be kept or assigned to a third party, even if there is a default and even if a landlord would prefer the lease be abandoned. The provision in the IIPR lease instead shifts the optionality to the Landlord, all while ensuring that the Landlord’s lease/loan payments continue unabated. This shift places the Landlord in a far more favorable position than they would be under the Bankruptcy Code. Nevertheless, the almost exclusive use by IIPR of provisions like this could suggest that larger or more sophisticated landlords, such as IIPR, are more likely to use bankruptcy contracting than smaller landlords.[201]
IIPR leases contain other provisions that would be unenforceable under the Bankruptcy Code. For example, one lease provides that upon default, in addition to amounts due under the lease, Landlord is entitled to liquidated damages equal to, at Landlord’s option, (1) the present value of rent payments due under the lease discounted by an agreed-upon rate, or (2) twelve months of rent.[202] Under the Bankruptcy Code, damages for defaulted and rejected leases are capped at “the rent reserved by such lease, without acceleration, for the greater of one year, or 15 percent, not to exceed three years, of the remaining term of such lease . . . .”[203] Default acceleration provisions are commonplace in all leases despite the limitations placed on them in the Bankruptcy Code, but a liquidated damage provision is far rarer (and was not present in the other leases examined). This latter provision grants IIPR greater certainty in its recovery upon default and thus could also be an example of contracting for better outcomes than would be available under the Bankruptcy Code.
c. Organizational Documents and Merger Agreements
7.7% of organizational documents and 10.9% of merger agreements in the Data Set contain some form of contractual tailoring to adjust for bankruptcy purposes.
Some organizational agreements provide substitutes for bankruptcy’s estimation mechanism. For example, upon dissolution of High Street Capital Partners, LLC, if the manager or its appointee (the “liquidator”) determines that “an immediate sale of part or all of the Company’s assets would be impractical or would cause undue loss,” the liquidator can instead distribute assets in kind.[204] (The Limited Partnership Agreement for GreenAcreage Operating Partnership contains similar language.[205]) Any distribution in kind would be valued according to a determination of fair market value—“the amount which the Company would receive in an all-cash sale of such asset in an arms-length transaction with a willing unaffiliated third party, with neither party having any compulsion to buy or sell . . . .”[206] If there is any dispute as to the fair market value, the LLC agreement sets forth an independent appraisal-based dispute resolution mechanism.[207] Rather than an imperfect judicial valuation as provided for under the Bankruptcy Code, or a market test that might misvalue assets due to regulatory uncertainty, the company relies on an estimate of a market of willing buyers and sellers to value assets.
The merger agreement between Canopy Growth and Acreage Holdings also includes substantive terms that differ from the rules of the Bankruptcy Code. Pursuant to an intellectual property license entered into as part of the merger, a subsidiary of Canopy licensed intellectual property to Acreage. The intellectual property license provides that “if Licensor, Canopy and/or their estate shall become subject to any bankruptcy . . . all rights and licenses granted to Licensee hereunder will continue . . . and will not be affected [by the bankruptcy], including by Licensors’ or Canopy’s rejection of this Agreement.”[208] However, if the licensee files for bankruptcy, the license agreement automatically terminates.[209] This partially replicates the treatment under the Bankruptcy Code. Under the Bankruptcy Code, a license that is rejected by a bankrupt licensor can be maintained by the licensee.[210] However, a license cannot automatically be terminated upon a bankruptcy filing by a licensee, as this would violate the prohibition on ipso facto clauses in bankruptcy.[211] Thus, this provision gives Canopy, as a licensor, more rights than it would have under the Bankruptcy Code. By contrast, license agreements entered into by Medicine Man Technologies, Inc. provide for termination if either party files for bankruptcy or initiates an assignment for the benefit of creditors, displaying very different preferences than those that would be foisted upon the parties if the Bankruptcy Code applied.[212]
This subsection 3 examined in detail the contractual tailoring achieved by secured lenders, lessors, and others in the legal cannabis industry. While somewhat rare, these changes favor secured lenders and lessors, who effectively act as secured lenders. The next subsection analyzes the structural mechanisms imposed on marijuana businesses in light of the freedom to contract in bankruptcy and the unavailability of bankruptcy protection.
4. Structural Mechanisms are More Common
As demonstrated in the prior subsection, bankruptcy contracting can be used by parties to achieve outcomes that diverge from what would typically be required under bankruptcy’s mandatory rules. But only about 6.6% of contracts within the Data Set contain such tailored terms. This subsection analyzes the ways that companies more commonly use structural mechanisms, rather than precise contractual terms, to ensure that their rights are protected when bankruptcy is unavailable.
Bankruptcy structuring refers to two distinct approaches a company could use to address the fact that bankruptcy protection is unavailable to the cannabis industry. One approach that a company could use to largely, if not entirely, sidestep the bankruptcy issue is by choosing not to touch the plant in the United States. Thirteen of thirty-four (35.3%) of companies in the Data Set do not touch the plant in the United States (an “opt-out” strategy). Nevertheless, as described in Part I, this approach is not foolproof for at least two reasons. First, the reach of the bankruptcy prohibition could extend beyond only plant-touching companies to companies like the seller in Way to Grow whose businesses largely depend on the cannabis industry.[213] Second, non-plant touching companies—particularly financiers—need to concern themselves with the availability of bankruptcy for their plant-touching contract counterparties. As such, many non-plant touching companies in the Data Set still spend significant portions of their Annual Reports discussing the risks facing marijuana companies.
A second approach would be to use the tools of corporate and organizational law to limit the unpredictable consequences of a bankruptcy filing. Strategies commonly used include seeking pledges of equity interests in affiliates and/or parent company guaranties. In each case, these strategies reduce or eliminate potential collective action problems and ensure additional assets are available to repay loans. At least some of these strategies are common in the marijuana industry: for example, pledges of the ownership interests of the borrower and related entities were found in 67% of the secured loan agreements reviewed.[214] Similarly, parent guaranties were present in 71% of the secured loan agreements reviewed. But since these strategies are so commonly used across industries, relying on them might present a misleading picture of the prevalence of structuring mechanisms as a response to the unavailability of bankruptcy in the marijuana industry, rather than simply a typical secured lending tool.
Therefore, for purposes of this subsection, a document is counted as including a structural mechanism if it (a) was entered into by a non-plant touching company (“non-plant touching”), or (b) contains one or more cannabis-specific structures (“cannabis-specific structuring”). Cannabis-specific structuring could include using separate legal entities for each cannabis license or property, special treatment of marijuana fixtures or inventory, or other strategies. A complete list of the types of structuring considered to be cannabis-specific structuring can be found in Appendix A.
The table below shows the number and percentage of contracts that include any kind of structuring (i.e., contracts that are non-plant touching, or contain cannabis-specific structuring), along with a breakdown by category:
Type | Count (%) of Total | Count (%) Non-Plant Touching | Count (%) Cannabis-Specific Structuring[215] | Count (%) without Any Structuring |
Annual Reports | 120 (10.3%) | 44 (36.7%) | 22 (18.3%) | 54 (45.0%) |
Organizational Documents | 104 (8.9%) | 87 (83.7%) | 3 (2.9%) | 14 (13.5%) |
Merger Agreements | 64 (5.5%) | 14 (21.9%) | 0 (0.0%) | 50 (78.1%) |
Secured Debt | 100 (8.6%) | 21 (21.0%) | 45 (45.0%) | 34 (34.0%) |
Unsecured Debt | 58 (5.5%) | 34 (58.6%) | 0 (0.0%) | 24 (41.4%) |
Leases | 38 (3.3%) | 10 (26.3%) | 6 (15.8%) | 22 (57.9%) |
Material Contracts | 683 (58.5%) | 237 (34.7%) | 0 (0.0%) | 446 (65.3%) |
Total | 1,167 (100%) | 447 (38.3%) | 76 (6.5%) | 644 (55.2%) |
The data demonstrates that 55.2% of contracts contain no structuring of any kind, meaning that a little less than half (44.8%) of contracts contain some form of bankruptcy structuring. Simply trying to “opt out” of the dual status issues faced by marijuana companies appears to be the most common strategy for almost all types of contracts examined. Recall that thirteen of thirty-four (35.3%) companies in the Data Set used this opt-out strategy. Thus, by classifying all their contracts as having opted out, we see that more than half of the coded documents are written on this basis. About one in twenty (6.5%) agreements contained some form of cannabis-specific structuring, with these terms again being concentrated in secured debt agreements (45.0%) and leases (15.8%), along with a smaller concentration in organizational documents.[216] Each of these will be explored below.
Opting out is an easy enough strategy to implement. A company can simply decide not to engage in plant-touching activities until the dual status issue is solved. A prime example of this strategy is Canopy Growth Corp. Its annual report explains that the company “will only conduct business activities related to growing or processing marijuana in jurisdictions where it is federally permissible to do so.”[217] This extends even to transactions where Canopy has obtained ownership interests or options to purchase ownership interests in US-based operators: in such situations, the company has “taken steps to insulate [itself] from all economic and voting interests until such time as there are changes to the federal laws of the United States related to cannabis related activities.”[218] Although this strategy is not foolproof, it is sufficient to permit several companies that use this strategy to conduct primary listings on US stock exchanges.[219]
That said, even non-plant touching companies can face troubles related to bankruptcy. For example, AFC Gamma, Inc., a financier, dedicates several pages of its annual report to the multiple forbearances, amendments, and defaults, including an attempted bankruptcy filing under Canadian law, it has struggled with in connection with one of its borrowers.[220] In brief, this strategy is blunt, likely forecloses business opportunities for operators, and is only partially effective.
Cannabis-specific bankruptcy structuring is much less common, and found in only 6.5% of agreements in the Data Set. Acreage Holdings, Inc.,[221] Goodness Growth Holdings, Inc.,[222] Planet 13 Holdings Inc.,[223] The Cannabist, Inc.[224] and Trulieve Cannabis Corp.[225] each create separate entities for each state in which they own property to hold all property within that state. Cresco Labs Inc.[226] and Glass House Brands Inc.,[227] meanwhile, create individual entities for each property. Presumably, either strategy could make it easier for a lender to take assets within each jurisdiction without needing to worry about the marijuana laws or regulations of other jurisdictions. Either, but particularly the latter, would also likely make the seized assets more valuable, by keeping the most important and specialized assets together in a single bloc. Other operators scatter ownership of licenses and properties across state lines in a number of different ways, which may reflect less concern for bankruptcy-related structuring than ease of integration into a larger corporate structure following waves of mergers and acquisitions. Similarly, only Acreage Holdings, Inc.,[228] Ascend Wellness Holdings, Inc.,[229] and MedMen Enterprises Inc.[230] pre-authorize all-asset sales to facilitate liquidation in case of insolvency. These provisions make it easier for secured lenders to seize assets or proceeds from asset sales, increasing the likelihood they will recover greater amounts in case of insolvency and thus lowering borrower credit costs.
An example of structuring provisions related to the special treatment of inventory can be seen in leases entered into between Ascend Wellness, Inc. and IIPR. Section 9.5 of this lease[231] (which language is mirrored in other Ascend leases)[232] provides the landlord with special powers to deal with abandoned marijuana inventory. As with many commercial leases, if Tenant fails to remove its property from the Premises prior to the termination of the lease (including through acceleration as a result of a Tenant bankruptcy filing or other default), Landlord may remove the property.[233] However, if the property is marijuana, then “subject to approval from Massachusetts cannabis regulatory authorities as required for certain items, [Landlord may] sell such property . . . for such prices as Landlord may obtain . . . .”[234] Trulieve’s landlord can act similarly, in accordance with Section 381.986 of the Florida Statutes, which governs medical marijuana.[235] Despite the possibility such provisions could expose the landlord to federal liability, it makes clear that the landlord will not wait for the cooperation of the tenant to dispose of cannabis inventory if the tenant files for bankruptcy. It also acts to remove potential conflict between the landlord and other creditors of the tenant who might try to obtain the inventory or sales proceeds. By purporting to treat marijuana as no different from any other personal property, these leases rely on well-established doctrines of property and contract law to remove some degree of uncertainty from the landlord’s ability to recover in case of tenant insolvency.
Finally, secured lenders use cannabis-specific bankruptcy-related structuring in forty-five secured debt contracts (45%). A breakdown of the strategies used by secured lenders is included below:[236]
Collateral Includes License Count (%) | Collateral Includes Marijuana Count (%) | Collateral Includes Real Estate Count (%) |
43 (95.6%) | 35 (77.8%) | 40 (88.9%) |
These structuring tactics are meant to permit a lender to keep the most valuable parts of a marijuana company—licenses, inventory, and real estate—together. In all likelihood, seizing a license would require state approval, if permitted at all. Nevertheless, a license remains one of the most valuable assets held by a marijuana company, and therefore is an asset a secured lender would want to ensure it had a strong claim on in case of default or insolvency.[237] Marijuana inventory is also quite valuable, but perhaps presents even greater difficulties for lenders, for whom mere possession of the inventory would cause them to run afoul of federal law. The status of a lien placed on inventory under the UCC is less clear, since mere possession of marijuana (or even its proceeds) is a federal (and if not licensed, a state) crime. Finally, real estate, through well-established foreclosure law, is the most easily seized asset, especially if cannabis inventory is removed. It also is the most readily adaptable for another user, making it more likely to serve as a source of recovery for the lender. Finally, if a lender were able to keep two, or even all three, of these assets together, they could far more easily maintain the company as a going concern, whether for themselves or a third party, if they so chose. This can explain why these strategies are fairly common among secured lenders in the Data Set.
The data presented in this subsection demonstrates that bankruptcy structuring is more common than bankruptcy contracting. 523 contracts in the Data Set (44.8%) contain some form of bankruptcy structuring: 447 documents (85.5% of the 523-document subset, and 38.3% of all documents) rely on the opt-out strategy, while 76 documents (14.5% of the 523-document subset, and 6.5% of all documents) use one or another form of cannabis-specific structuring. These figures, however, should be considered in light of some important caveats. It was not infrequently challenging to ascertain whether lenders claimed a security interest in certain useful blocks of assets. Many underlying security agreements were inaccessible, so the precise terms needed to be inferred from the related loan agreements. Unless it was absolutely clear from the accessible documents that the inventory, license, and/or real estate was included in the collateral package, a document was not coded as such. As a result, the data presented above likely represents an undercount of the actual use of cannabis-specific bankruptcy structuring. This means that the data presented in this subsection should not be taken with the same level of precision as the other data presented in this Part. Nevertheless, the existence of bankruptcy structuring, and its greater prevalence than bankruptcy contracting, should be apparent.
This subsection presented data on the prevalence of bankruptcy structuring, showing it to be fairly common and far more regular than bankruptcy contracting. The next subsection addresses a prediction from the bankruptcy contracting literature: would companies, if able, choose to automate parts of the bankruptcy process through the issuance of exotic equity? The data presented in the next subsection answers this question in the negative.
5. Never Issue Exotic Securities
As described briefly in the next Part, scholars have predicted that if companies were permitted to contract around bankruptcy’s mandatory rules, they might choose to use their very capital structure to automate the insolvency process. These proposals are premised upon the theory that a simpler insolvency process could lead to higher recoveries for creditors, who would, in turn, pass on this benefit in the form of lower interest rates.[238]
One approach starts from the common view that bankruptcy exists, in part, to prevent value destruction caused by the liquidation of financially insolvent, but otherwise economically viable firms.[239] Rather than impose the “ubiquitous cost” of bankruptcy on all firms, a firm’s capital structure could offer clues as to whether a firm is viable or not.[240] To achieve this, firms could issue “Chameleon Equity.”[241] Such a firm would not issue any debt[242]; instead, it would issue a special form of preferred equity, where a default would automatically convert this preferred equity to common equity and eliminate any lower-priority equity. This would accomplish a quicker and cheaper reorganization than under the Bankruptcy Code.[243]
Another way to accomplish the same goal—allocation without an expensive valuation process—could be through the issuance of option contracts to senior creditors, junior creditors, and equityholders. If structured properly, these options could ensure that each party receives exactly what it would be entitled to regardless of the ultimate value of the company.[244] A final way to automate parts of the bankruptcy process concerns the decision of whether to liquidate or reorganize. Reorganization is only appropriate when a firm is financially distressed but economically viable—i.e., it is temporarily unable to pay creditors but otherwise remains a potentially profitable business. But it can be difficult to discover whether a firm is economically viable in any given case. Nevertheless, it may be that, in general, the viability of firms in a certain industry is strongly correlated with certain objective economic indicators, such as the consumer price index or the price of certain common inputs.[245] So, a firm could issue equity or other contracts that require liquidation upon insolvency if a relevant indicator (in the case of marijuana companies, this could be the average price of legal marijuana in a given market) performs poorly. Alternatively, if the firm is insolvent while the relevant indicator is performing well, the contract could call for reorganization. In either case, the decision to liquidate or reorganized is automated based on objective, easy-to-ascertain factors.
Such unique, bankruptcy-automating terms would need to be clearly disclosed to investors under securities laws. The presence of such equity or contracts would need to be included in a company’s 10-K filings, and the terms of these options or equity conversion arrangements would separately need to be disclosed as well.
There is no contract or filing for any company in the Data Set that indicates the existence of any such exotic equity or option arrangement. Moreover, all firms limit themselves to issuing secured and unsecured debt, standard and preferred equity, and convertible debt. This indicates that not only are the three proposals discussed above not used by any cannabis companies, but also that no other similar proposal involving options or conversion rights that would automate aspects of the bankruptcy process are used. Possible explanations for this complete absence are suggested in Part IV.
6. Conclusions on Bankruptcy Contracting
This Part presented novel, hand-collected data on the bankruptcy contracting undertaken by marijuana companies in the real world. Perhaps surprisingly, even though these companies know that traditional chapter 11 protection is almost entirely unavailable for them, they largely do not write contracts that account for this fact. Nevertheless, a key exception to the lack of bankruptcy contracting is found in secured lending and leases that are part of larger sale-leaseback financing transactions. Although still somewhat uncommon, bankruptcy contracting is found four times more frequently in secured financing agreements and six times more frequently in leases than in the general set of examined contracts. Furthermore, while bankruptcy contracting is rare, bankruptcy structuring is far more common, with a little less than half of examined documents containing one or more forms of bankruptcy structuring. Finally, while bankruptcy contracting is rare and bankruptcy structuring more common, issuance of exotic securities is nonexistent. These results suggest a reexamination of the theory justifying the existence of the Bankruptcy Code may be warranted, its mandatory nature (outside of the marijuana industry), and contractualist critiques of these foundational principles. The next Part analyzes this theory and why the results presented in this Part might be surprising, while the final Part offers some explanations for these results.
III. Bankruptcy Contracting as an Alternative to a Mandatory Bankruptcy Regime
The first Part of this Article explored why bankruptcy is unavailable to the legal marijuana industry. The industry’s unique inability to access the bankruptcy system raises questions for supporters of bankruptcy’s mandatory rules while presenting opportunities for those who support a regime based on bankruptcy contracting. The prior Part, meanwhile, made the empirical case that bankruptcy contracting is rare despite its availability in the industry. This Part investigates the typical understanding of why access to bankruptcy is mandatory, criticisms of this view, and suggestions regarding how companies might write bankruptcy contracts to implement tailored, more efficient solutions to insolvency.
A. The Creditors’ Bargain and Arguments Against Bankruptcy Contracting
Bankruptcy is unlike many other parts of corporate law. Corporate law mostly consists of default rules that parties can contract around if they prefer a different outcome. Mandatory rules, which parties cannot contract around for public policy reasons, are much rarer.[246] Bankruptcy, however, primarily consists of mandatory rules—parties cannot waive their right to file for bankruptcy or otherwise generally contract around parts of bankruptcy’s framework. This Section gives an overview of the justifications for bankruptcy’s mandatory system.
Bankruptcy is a mandatory regime because it is meant to force all parties with a financial interest in the debtor—lenders, vendors, employees, tort victims, governments, and others—to the negotiating table while preventing those parties from unfairly taking advantage of one another.[247] If parties could opt out of bankruptcy, theory suggests that they would race one another to the courthouse to levy on a debtor’s assets.[248] Slower creditors would be unfairly left without recovery, while value may be wasted if assets could be worth more kept together rather than sold piecemeal.[249] Creditor monitoring costs would increase, because each creditor would need to guard against the strategic behavior of all other creditors.[250] These aspects, in turn, would increase the costs to lenders of providing credit while decreasing their potential recoveries in the event of borrower distress, ultimately leading to less credit being provided at higher cost.[251]
The bankruptcy system exists, it is argued, to avoid this parade of horribles. Scholars conceptualize bankruptcy as the outcome of a hypothetical bargain among creditors to solve this collective action problem—a set of rules that would apply to all parties to a bankruptcy proceeding. Thus, in this classical “Creditors’ Bargain” formulation, bankruptcy is a compulsory system because “allow[ing] the debtor to contract with other creditors on an opt-out basis would destroy the advantages of a collective proceeding.”[252] Bankruptcy therefore consists of a system of mandatory rules regarding the imposition of a stay on collections, specified treatment for different types of claimants, powers that debtors can exercise to bring value back to the company, and a voting system to ensure that creditors of all kinds are informed about and have a say in a reorganization plan.
Other reasons for a mandatory bankruptcy system have been put forward by scholars concerned about how certain classes of creditors would be treated in a default rule regime. Theorists worry that a system where some parties could opt out would harm tort victims,[253] small creditors, unsophisticated parties, and certain types of governmental claimants.[254] These “maladjusting creditors” might lack the time, inclination, sophistication, or wherewithal to make bankruptcy contracts with a counterparty.[255] Courts, meanwhile, reason that if opting out of bankruptcy law were permitted, “astute creditors would routinely require their debtors to waive” access to bankruptcy, thus rendering the law effectively moot.[256]
Bankruptcy’s mandatory regime has convinced many theorists, and there is little judicial or legislative support for a default, rather than mandatory, bankruptcy system. In short, with limited exceptions, bankruptcy contracting has not caught on. Nevertheless, there are powerful arguments for permitting bankruptcy contracting across industries, along with considered predictions about the types of changes we might expect in a regime that permitted bankruptcy contracting. These are explored in the next Section.
B. Arguments in Favor of Bankruptcy Contracting
Although the explanation for a mandatory bankruptcy system summarized in the prior Section has become commonly accepted,[257] it is not without its detractors. Practice also demonstrates demand for contractual alternatives, with contract being used to chip away at parts of the existing bankruptcy system.[258] Many scholars have suggested that it is odd, inefficient, and costly to impose a single bankruptcy system on all debtors and creditors while the rest of corporate law is built upon freedom of contract. These critics instead propose that at least some parties be permitted to negotiate insolvency remedies by contract. Using private ordering,[259] they could replicate, modify, or even abandon some or all of the current Bankruptcy Code, tailoring its provisions to their particular needs. Arguments in favor of bankruptcy contracting are explored in this Section.
1. Weak Form—Bankruptcy as a Menu
Proponents of bankruptcy contracting fall into several camps. One set of critics sees bankruptcy’s dependence on nonwaiveable provisions as incongruous with the rest of corporate law. Bankruptcy’s mandatory rules lead to less efficient outcomes than would be available if parties could contract around them.[260] A default rule, on the other hand, is one that would apply only if parties do not decide on a different rule. Default rules provide parties with more flexibility to tailor their arrangements to their specific situations, and only come into force if parties are silent on the particular issue to which the default rule applies. As a result, parties are generally expected to negotiate to a better outcome for them than would be provided by a default rule so long as the transaction costs of doing so are not too high.[261] Without this flexibility, parties are forced to use an expensive, time-consuming, one-size-fits-all process with no analogue in any other part of corporate law.
Professor Rasmussen proposes that companies could reach a more efficient solution by selecting, at the time of their formation, from a menu of corporate reorganization options in the event of insolvency.[262] The choices available would be to commit to: (1) never use federal bankruptcy law, (2) only file for liquidation under chapter 7, (3) only use chapter 11, (4) only use a modified proceeding where all creditors except the financing creditor are stayed, or (5) create a fully-customized insolvency regime.[263] In theory, creditors and equityholders might prefer one or another of these choices under certain circumstances. For instance, never using federal bankruptcy law would enable equityholders to obtain lower interest rates from secured creditors because secured creditors could more quickly and easily repossess collateral using state law foreclosure procedures.[264] A chapter 7-only firm might benefit unsecured creditors who could rely on an orderly liquidation to ensure they will obtain a pro rata share of a company’s remaining value and forestall a wasteful race to the courthouse.[265] Finally, a custom regime could unlock even more efficiencies for parties willing to negotiate for it.[266]
The purpose of the menu is to provide a broader set of options to debtors and creditors than is currently available, along with a credible mechanism to ensure they do not strategically change their choice to selectively harm creditors.[267] This form of critique does not say that bankruptcy does not solve problems, or that contract is the ideal solution for every economic actor. Instead, it merely suggests that parties should have a broader set of options in the name of efficiency.
2. Middle Form—Markets Evolve
Other scholars take a different approach to their criticisms of the bankruptcy system. They start with the Creditors’ Bargain theory,[268] but suggest that markets have changed in important ways that makes this theory obsolete in some respects. As a result, some of bankruptcy’s mandatory rules, as justified by the Creditors’ Bargain theory, might be better circumvented by mutual agreement between the parties.
Some critics believe the collective action problem bankruptcy is meant to solve is overblown. Smaller businesses or those with simple capital structures may have only a handful, or even just one, dominant creditor. As discussed above, smaller businesses often resolve financial distress without bankruptcy’s supposedly essential mandatory rules, instead using consensual workouts or state-law procedures.[269] Coordination problems for larger, publicly-owned firms are also much smaller now than they were in the past. Over the past two decades, distressed private investment firms have been able to acquire controlling positions in widely-held distressed firms through purchases of their debt or other claims against them on secondary markets.[270] The increasing power of senior secured lenders means there is less need for lenders to coordinate with one another, since “the most senior creditors call the shots.”[271] As long as this is true, bankruptcy is not needed to prevent a race to the courthouse or force parties to the negotiating table; therefore, other rules during insolvency might achieve better outcomes.[272]
Some recent developments in capital markets suggest that control of distressed companies is swinging back from secured lenders to private equity owners. Even though ownership may be more widely dispersed on public markets, since control is concentrated in the hands of private equity owners coordination problems may be less important. Moreover, private equity owners may try to avoid bankruptcy because its mandatory rules would wipe out a private equity firm’s substantial investment in a portfolio company, eliminate its advisory fees, harm its market reputation, and expose it to lawsuits for prebankruptcy actions.[273] Instead, private equity sponsors can use the threat of a drawn-out bankruptcy case, which could diminish firm value and therefore harm creditors,[274] in exchange for creditor agreement to grant them value in direct contravention of bankruptcy’s mandatory requirements.[275] Thus, even though under bankruptcy’s rules the private equity sponsor should retain no equity in the company and should remain liable for certain prebankruptcy actions, clever private equity firms are sometimes able to obtain releases from liability[276] and retain equity in the firm instead.[277] If markets have transformed so there is no collective action problem to solve, or if it can be efficiently solved by contract, then why have a compulsory, yet inefficient, bankruptcy system?
Bankruptcy practice has also evolved to take advantage of the fact that “[b]ankruptcy operates on legal entities, not on firms.”[278] Creditors cannot force companies to opt out of bankruptcy by writing a contract that excludes an important asset from a bankruptcy proceeding. But, creditors can achieve much the same thing by requiring a company to place a key asset in a separate legal entity within a corporate group, and then keep that entity out of bankruptcy (and out of the hands of other creditors) if the group suffers financial distress.[279] This type of “withdrawal right,” which in this case would take place prior to a bankruptcy filing by one or more of the other entities within a corporate group, might sound like exactly the circumstances that bankruptcy’s mandatory procedures are meant to avoid—indeed, “[t]here is nothing mandatory about rules like the automatic stay when assets can be partitioned off into legal entities that never enter bankruptcy.”[280]
But, a withdrawal right can solve the collective action problem that bankruptcy is supposedly needed to solve. If a firm’s key asset (without which the firm would need to shut down) is placed in a separate entity owned by only one investor, then there is no collective action problem upon insolvency. Instead, the withdrawal right operates as “an acid test of whether the firm should be saved.”[281] If the firm has going-concern value, the parties will negotiate to keep the firm together; if not, the withdrawal right will be exercised and the firm will shut down, all at a cost far below that of a bankruptcy proceeding.[282] Withdrawal rights impose costs on other creditors, and if too many withdrawal rights are given then collective action problems can be reintroduced.[283] But, withdrawal rights offer a potentially more efficient alternative to bankruptcy’s mandatory rules under some circumstances.[284] Withdrawal rights demonstrate another way markets have evolved to provide, in limited circumstances, a response to financial distress that differs from that supposedly demanded by the Creditors’ Bargain.
Finally, the international shipping industry provides an idiosyncratic example of how private ordering solutions can arise in particular markets in the absence of public law mandates. Ships are owned and operate in multiple jurisdictions and on the high seas, and visit ports where corruption and inefficiency may lead to suboptimal outcomes for creditors if insolvency resolution was based on public law alone.[285] Since depending on a single nation’s bankruptcy laws may be impractical or unduly costly, this industry has developed institutions for resolving financial distress through “a nexus of private, decentralized, differentiated and competitive market institutions.”[286] Primary among these for purposes of this Article are the “double mortgage,” governing law, and forum selection clauses.[287]
First, a double mortgage gives a lender a mortgage on a ship as well as a security interest in the shares of the special-purpose entity that is the registered owner of the ship.[288] So, if distress strikes the holding company while a particular ship is on the high seas, a lender can simply seize ownership of the shares of the special-purpose entity that owns the ship, becoming the record owner of the ship without any court process.[289] Lenders also often require that ships be registered in certain trusted jurisdictions because while a ship could be registered in any jurisdiction some have poor reputations for protecting lenders’ property rights.[290] Similarly, loan agreements often specify choice of law and arbitration venue in favored jurisdictions to ensure expertise and reduce expense.[291] Collectively, these contractual and legal innovations replace bankruptcy law in this industry, suggesting that under some circumstances it may be possible to avoid the parade of horribles that justifies the Creditors’ Bargain theory.
3. Strong Form—Replace Bankruptcy with Contract
A final group of critics claims that bankruptcy law should be eliminated entirely and replaced by contractual arrangements, auctions, and other similar structures.[292] Replacing bankruptcy with these other mechanisms could improve outcomes in several ways. First, if there is only one or a small number of creditors, bankruptcy law—specifically the ability of management to decide whether to liquidate or reorganize, or delay undertaking either[293]—often serves to shift value from creditors to management, rather than helping the business. Eliminating bankruptcy law, the argument goes, would improve recoveries for creditors and therefore lower interest costs for borrowers.
To understand why, recall that managers are often major equityholders in the companies they run, and obtain substantial private benefits from their positions—salaries, perks, reputational benefits, and others.[294] Moreover, consider that liquidations and reorganizations harm management in two ways. Liquidations that fail to repay senior creditors in full necessarily wipe out equityholders,[295] while reorganizations often dilute any preexisting equityholders. In both liquidations and reorganizations, management will typically be replaced. With all of this in mind, management may wish to delay a liquidation or reorganization to maintain their private benefits while also betting that the company’s fortunes may improve in the meantime and their equity will once again have value (this is the so-called option value of equity).[296] Delay harms creditors because assets dissipate in value over time and because reorganization is expensive.[297] But this harm is asymmetric, because equity in an insolvent company is already worthless. Proponents of bankruptcy contracting argue that replacing these rules with a contract providing for quick foreclosure and an abbreviated auction could improve creditor recoveries and thus lead to a lower cost of credit for debtors.[298]
Alternatively, the bankruptcy system could be redesigned to give junior creditors and equityholders a stake in the reorganized company to compensate them for the option value of their equity. This so-called relative priority system would give some value to junior creditors and equityholders because it is possible a company’s fortunes could turn around and simply wiping out junior claimants is unfair (why should they bear the entire loss?) and inefficient (it could lead to the delay and waste discussed in the prior paragraph). Relative priority has some prominent supporters, but contravenes current bankruptcy law’s insistence on absolute priority when parties do not consent to different treatment.[299] Senior and junior claimants could potentially negotiate for customized forms of relative priority and implement them by contract rather than legislatively.
Finally, if entire firms can readily be sold at auction, then the market can determine what the going concern is worth and there may be no need for bankruptcy’s costly and time-consuming forced negotiation process.[300] A bankruptcy judge might play an important role if she can help parties overcome bargaining frictions that might otherwise prevent them from reaching a mutually beneficial deal. But large and efficient markets for distressed companies and their debt reduces the need for judicial intervention.[301] Over the years there have been numerous proposals to replace the reorganization process with an auction of some or all of the firm,[302] or flotation of shares or options to purchase the firm.[303] More exotic solutions include urging companies to issue preferred equity that automatically converts to common equity,[304] grant options that permit the purchase of shares in a reorganized company for a fixed price,[305] permit signals from the wider industry or broader economy to determine whether a company will liquidate or reorganize,[306] and others. Collectively, these proposals suggest that bankruptcy’s mandatory rules can be inefficient for some potential debtors, and that sophisticated parties could, and would, prefer to use contractual solutions to automate some aspects of insolvency resolution.
C. Predictions About Bankruptcy Contracting
In a world where bankruptcy contracting was permitted, we might expect companies to act differently from how they currently do across a variety of different measures. This Section identifies and classifies these changes, tying them to the theory explored in the prior Section and suggesting the types of contracts that might have been predicted to be present, but rarely were, in the legal cannabis industry. The list provided here cannot be anything but incomplete because a true freedom of contracting regime would allow for an unlimited amount of customization. Instead, it aims to catalogue two broad types of changes one might expect if the restrictions on bankruptcy contracting were loosened: changes legal or capital structure prior to examining any particular contracts, and substantive changes to a company’s contracts.
1. Ex Ante Changes to a Company’s Legal or Capital Structure
If companies could more freely contract around bankruptcy, we might expect them to publicize this ability and the broad contours of any bankruptcy contracts they made. Even proponents of bankruptcy contracting agree that “[f]irms cannot be allowed unlimited freedom to amend” their general choice of which bankruptcy rights to use, because counterparties will change their behavior in reliance upon the bankruptcy choices an individual firm makes.[307] It would therefore be reasonable to expect companies to explain the absence of bankruptcy protection, how this enables them to make bankruptcy contracts, and the importance of such contracting to their business.[308] We might expect material contracts, charters, and public disclosures to indicate the type of bankruptcy system that a company would choose to use upon insolvency. Over time, custom provisions might evolve into a relatively small set of potential alternatives for efficiency reasons.
We could also predict that the absence of bankruptcy protection, or the need to make bankruptcy contracts, might impact the supply of, and demand for, debt. Of course, it would be difficult to demonstrate that a given company’s decision not to borrow was based upon the unavailability of bankruptcy. But debt is not essential for corporate growth, and companies can both grow and remove insolvency risk by funding themselves entirely through equity.[309] Therefore, if debt was an important feature of most companies in the marijuana industry, or if debt incurrence increased over time, this could suggest that neither companies nor their lenders are put off by the unavailability of bankruptcy.
We might further conclude that looser restrictions on bankruptcy contracting could affect the types of debt that cannabis companies incur. Secured debt might be more common than unsecured debt, since the existence of identified collateral could reduce concerns about a race to the courthouse.[310] Looser restrictions might also impact the structure companies use to borrow. For example, if lenders believe it will be difficult to foreclose on properties in multiple states without bankruptcy’s automatic stay, they could try to implement structural mechanisms to ease the foreclosure process. These might look like the double mortgages used in the international shipping industry described supra,[311] or like mezzanine loans common in real estate transactions. As described above, these structures permit quicker foreclosure with reduced court intervention.[312] More generally, companies might engage in asset partitioning to keep useful blocks of assets (for instance, using a single entity to hold all licenses or property in a given state) together and thus increase potential recoveries for a foreclosing party.[313]
Finally, even severe critics of bankruptcy’s mandatory rules tend to agree that the ability to sell property quickly, and free and clear of liens held by creditors, is an essential part of insolvency resolution.[314] Without the powers under the Bankruptcy Code, a cannabis company would likely need to obtain assent from all shareholders prior to engaging in a sale of substantially all assets.[315] As a result, cannabis companies might try to obtain pre-authorization for an all-asset sale in their organizational documents.
2. Substantive Changes to a Company’s Contracts
In addition to changes cannabis companies might make to their legal or capital structure if able to contract around bankruptcy, they might also make changes to contracts themselves. First, we could examine their contracts to see whether parties try to mirror certain rights they might have in bankruptcy. Scholars have disagreed over the need for a non-opt-out automatic stay,[316] the usefulness of the absolute priority rule (perhaps bankruptcy’s most fundamental mandatory rule),[317] and whether bankruptcy functions as an effective valuation mechanism,[318] among a host of other issues. The importance of these terms might be reflected in their presence, or absence, in contracts when bankruptcy contracting is permitted.
Contractual terms could demonstrate how concerned parties are about collective action problems—do they try to contractually create some version of an automatic stay or replicate some version of bankruptcy’s class-based voting mechanisms (such as through an intercreditor agreement)?[319] Lenders could try to obtain detailed information rights or third-party appraisal rights—the types of things the Bankruptcy Code would, perhaps imperfectly, provide.[320] If so, this might indicate that such information is important to remaining informed about the value of their collateral and thus aid in deciding when liquidation is warranted.[321] Alternatively, do secured financiers successfully get borrowers to waive any right they may have to a stay on foreclosure?[322] If lenders instead demand auctions upon borrower insolvency, this could suggest a general distaste for the reorganization process.[323] Junior creditors or equityholders, meanwhile, might try to negotiate deviations from absolute priority in exchange for agreement to move quickly through distress.[324] These or other potential contractual incorporation of, or deviation from, bankruptcy’s mandatory rules could shed light on what parties themselves believe are the most important parts of the Bankruptcy Code.
Next, as discussed above, there are several state-law alternatives to bankruptcy, only some of which have been explicitly authorized for use by cannabis companies. Nevertheless, some companies or their creditors might prefer the advance selection of one procedure or another (such as a marijuana-specific state-law receivership) to reduce uncertainty, obtain the particular protections available under a specified procedure, or for any other reason. Although many contracts across all industries contain governing law and venue provisions, if some jurisdictions have made more efforts to legitimize marijuana businesses than others, parties might choose governing law provisions that reflect expertise in this domain.[325] Provisions dictating the use of a particular governing law, venue, or distress procedure might, therefore, be expected to be a common feature of contracts in the marijuana industry.
Finally, the possibility that companies might innovate in uncommon, or heretofore unexpected, ways should be considered as well. The variety of forms these could adopt is too large to catalog here. The point, however, is that if bankruptcy contracting is possible for cannabis companies, over a large enough sample we should expect to see some range of contractual innovation take place.
IV. THE ROLE OF BANKRUPTCY CONTRACTING AND MANDATORY BANKRUPTCY LAW
Part II demonstrated the key findings of this Article. Overall, it shows that in the one industry where bankruptcy contracting is possible, it is quite rare. The bankruptcy contracting that does exist is primarily concentrated in one type of contract—secured financings, whether by loan or by sale-leaseback. Marijuana companies more often alter their capital structures or operations to account for their prohibition of bankruptcy filings, but even these types of changes are uncommon. The cannabis-specific changes made are also quite rare. Finally, proposals for using bankruptcy contracting to automate the most difficult, costly, or time-consuming aspects of bankruptcy are entirely absent from the publicly-available documents in this industry. Informed by the data uncovered in Part II and how it diverges from the theoretical predictions of Part III, this Part aims to offer explanations for why bankruptcy contracting is concentrated in certain types of documents but otherwise generally rare, while providing a normative view of the uses and drawbacks of bankruptcy contracting. In short, the data suggests that bankruptcy contracting might be beneficial in some circumstances, but there may be higher costs associated with a regime in which bankruptcy contracting is permissible than theorists might have expected.
A. The (Ir)rationality of Bankruptcy Contracting
Why isn’t bankruptcy contracting more common in the legal marijuana industry? And why is the contracting that does exist concentrated among secured lenders? One likely explanation for both questions arises from the informational and transactional costs imposed by a bankruptcy contracting regime. For most contract counterparties, trying to create a tailored bankruptcy regime may simply be too expensive and the risks addressed by these rules too remote to make negotiation worthwhile. For entrepreneurs, most contract counterparties, and maladjusting creditors, the game simply isn’t worth the candle.
First, consider the cost-benefit analysis of entrepreneurs who start cannabis companies. On the cost side, there is reason to believe that entrepreneurs in the industry may possess a higher-than-typical risk tolerance. Operators in the industry must be comfortable with legal uncertainty. Insurance is often unavailable, difficult, or costly to obtain.[326] The regulatory landscape evolves rapidly and is subject to major shifts from administration to administration. In addition, leaders in this industry take on personal risk that is almost unheard of in other industries. For example, one company notes that its “employees, directors, officers, managers and/or investors could face detention, denial of entry or lifetime bans from the United States for their business associations” with it.[327] In short, entrepreneurs in this industry, perhaps more than others, may require an optimistic disposition in order to enter the market in the first place. This optimism in turn could systematically result in a heavy discounting of bankruptcy risks. Entrepreneurs embarking on a new venture rarely want to consider the particulars of how to resolve the failure of their new venture, and with marijuana entrepreneurs this may be especially true. Viewing bankruptcy as unlikely, or having a higher-than-typical risk tolerance, could lead an entrepreneur to conclude that the cost of negotiating a bankruptcy contract is not worth any benefit that might accrue as a result.
At the same time, the benefits of bankruptcy contracting to entrepreneur-equityholders are also uncertain because of their low priority in bankruptcy and the nature of any negotiation process. Under normal circumstances, equityholders would expect to recover last (if at all), be removed from management, and likely be wiped out in a bankruptcy. This would likely be true whether the Bankruptcy Code applied or not, since as a general principle repaying debt takes priority over redeeming equity. Knowing this, entrepreneur-equityholders could try to negotiate a provision to retain them or make an explicit payment to them in the event of insolvency. However, doing so might be challenging: it would require them to ask their lender, the party that would pay them or retain them, to do so if the business fails before the business even commences. Additionally, they would be negotiating from a position of weakness as compared to a normal company, since bankruptcy’s automatic stay—the key holdup tool by which payments to managers are typically extracted—would be unavailable. Such a negotiation might be possible and could succeed, but its absence from the data implies that entrepreneur-equityholders are not likely to be a primary source of demand for bankruptcy contracting. Thus, from both the cost and the benefit side it may be that it is more rational for entrepreneurs, and perhaps especially marijuana entrepreneurs, to avoid bankruptcy contracting.
Next, consider parties like employees, trade creditors, and other unsecured creditors. These groups will have a wide range of negotiating leverage and sophistication regarding the unique risks of bankruptcy for cannabis companies. For some, any given contract with a cannabis company may be one of a very large number of contracts the company has with all its other counterparties, or may concern relatively small dollar amounts. Such creditors may not be willing or able to negotiate custom bankruptcy terms in their contracts. Additionally, the cannabis companies themselves might not want to negotiate such contracts because of the high transactional cost of doing so with each potential creditor. Moreover, in order to negotiate an effective bankruptcy contract, both the marijuana company and its creditor would need to understand the bankruptcy terms of all other contracts entered into by the cannabis company. Without this understanding, a counterparty might believe it is entitled to a certain priority, only to find out that some other party negotiated a contract granting that party priority instead. A further difficulty arises since contracts can vary infinitely, so understanding a company’s particular priority within a complex capital structure could be daunting.[328] For them, even if bankruptcy contracting is possible, it likely is too costly, in terms of the information, time, and legal fees required, to make sense for many parties to regularly undertake. This could be true even of sophisticated parties able to understand the potential benefits and drawbacks of bankruptcy contracting—they might decide that the benefits would be diffuse, unlikely to be realized, and not worth the cost.
Finally, consider maladjusting creditors. Professors Warren and Westbrook claimed that bankruptcy contracting would disadvantage tort victims, small creditors, unsophisticated parties, and governmental claimants who are proscribed by law from engaging in risk-based price discrimination. Some of these maladjusting creditors overlap with the claimants described in the prior paragraph. But the difference here is that there is no expectation of sophistication, or even of contractual awareness, with most maladjusting creditors. Tort victims, in particular, have commonly been understood by contractualists to be a class that requires protection via a default set of bankruptcy rules. This was true even before recent mass tort bankruptcies have highlighted the aggressive, and potentially troubling, bankruptcy tactics companies can use to reduce their litigation exposure. If sophisticated parties that might stand to benefit from bankruptcy contracting rarely engage in it, then it seems even less likely that maladjusting creditors would do so. Indeed, not writing bankruptcy contracts seems rational for these parties, and could be one major reason why only 2.5% of material contracts within the Data Set (of which a subset would be made with maladjusting creditors) contain any form of tailoring. A bankruptcy system that forces maladjusting creditors to write bankruptcy contracts likely imposes costs on them that exceed any benefits they might obtain through contracting. This would most probably shift any benefits from contracting to the cannabis company at the expense of the maladjusting creditors. Over time, this could increase inequities between sophisticated and maladjusting parties in the bankruptcy system.
Yet the fact is that bankruptcy contracting is not ignored completely by industry participants. Instead, contracting is concentrated among secured lenders. This concentration lends further credence to the information and transaction cost hypothesis presented in this Section. Secured lending is generally understood to be beneficial for companies because secured lenders can monitor borrowers more closely than other outsiders (and sometimes even more closely than insiders like corporate boards).[329] This monitoring ensures discipline, reduces conflicts of interest, and increases enterprise value.[330] The importance of secured lenders as monitoring parties in the marijuana industry may be heightened due to the regulatory uncertainty and the absence of most traditional sources of institutional capital. Marijuana companies, in particular, might substantially benefit from the steadying hand and watchful eye of a secured lender.
Secured lenders, in turn, are incentivized to negotiate tailored bankruptcy terms in their contracts because of their long-term, high-stakes bets in this uncertain industry. Secured lenders to this industry often specialize in the unique aspects of providing financing for cannabis companies, both because the industry is complex and because there is very limited appetite from more traditional lenders to enter the market. In addition, many secured loans obtained by cannabis companies contain all-asset liens or otherwise constitute the most important part of the capital structures of cannabis companies.[331] This, in turn, works to dissuade other lenders from providing capital to the borrower, limiting the complexity the secured lender might face in trying to determine its exact priority with respect to any given asset.[332]
Despite this putative security, however, as noted above and described in detail below, a secured lender may be severely limited in its ability to foreclose on much of the most valuable collateral on which it has a lien.[333] Thus, secured lenders take major risks in providing financing to these companies. As such, lenders are motivated to negotiate strict, complex, customized provisions relating to bankruptcy, as well as to ensure that they monitor and enforce these provisions. Secured lenders take advantage of the freedom to contract around bankruptcy because they believe the benefits of negotiating insolvency-specific remedies justifies its cost and complexity. Consonant with this statement, it is notable that IIPR, the largest sale-leaseback financier in the industry, included several bankruptcy-related substantive and structural rights in its leases, while almost no other lessor did.
Of course, sometimes it is still too difficult or costly to negotiate bankruptcy contracts. Lenders or other contract counterparties may hope for more certainty in the case of insolvency, but are unwilling or unable to negotiate the types of custom terms that the largest and most sophisticated secured lenders bargain for. In such cases, reliance upon a relatively small number of well-tested mechanisms founded in general principles of corporate law, which operate the in the same manner as they would for more traditional companies, could serve as a second-best approach. Such mechanisms include ensuring that the most valuable assets are held together in blocs that should be easier to seize, taking security interests on the most valuable assets, using distinct legal entities to hold the most valuable assets, and pre-authorizing a sale of all assets. Combining assets in easy-to-seize blocs is reminiscent of the international shipping industry’s double mortgage approach, which operates predictably across a wide range of potential circumstances to reduce the time and frictions required to complete a foreclosure. Although this approach may appear second-best when compared to bespoke contracting, this standardized approach may actually be optimal after accounting for information and transaction costs.
The foregoing analysis, based on informational and transactional costs, provides a possible explanation for the relatively low prevalence of tailored bankruptcy contracting and capital and operational structuring, as well as why the latter is more common than the former. If bankruptcy contracting is seen as too costly, then parties will not engage in it; at the same time, if structuring is cheaper and provides some of the same benefits, parties may choose to use structuring instead. Following this analysis leads to proposals to improve the use of bankruptcy contracting, whether in this industry or in a hypothetical bankruptcy system more amenable to contracting. The first would be to implement a mandatory, standardized disclosure requirement for any bankruptcy contracts. This would enable all interested parties to more quickly understand what rights would already be spoken for through prior bankruptcy contracts, while also making public the common types of rights that parties contract for. A second way to improve bankruptcy contracting would be to try to limit, either by law, contract, or custom, the varieties of bankruptcy contracts that are possible. If only a fixed, small number of different types of bankruptcy contracts or structures would be permitted—akin to property law’s numerus clausus principle that limits the different types of property ownership available[334]—information costs would be reduced and transactions could proceed more smoothly. Finally, there should be clear and balanced default rules for parties who choose, rationally or otherwise, not to make bankruptcy contracts. Without well-established default rules, the risk of inequitable outcomes for many types of creditors is too great.
Disclosure and standardization are interrelated and have been recognized as important by scholars in the bankruptcy contracting and other contexts.[335] Together, these strategies would reduce informational and transactional costs, thus making it more likely that greater numbers of secured lenders and other parties might use bankruptcy contracting. However, it would also reduce the costs imposed by granting too much freedom. Creating a choice architecture that facilitates mostly optimal choices under most circumstances should benefit those who make bankruptcy contracts as well as those who do not.[336] Disclosure and standardization in turn should reduce insolvency costs and thus lower companies’ cost of capital. By requiring better disclosure, more standardization of the tailoring and structuring strategies that might be used in this industry, and clear default rules, some of the possible inequities that bankruptcy contracting can introduce may be reduced as well. Even if bankruptcy contracting never extends beyond the legal cannabis industry, making these changes would still improve outcomes for marijuana debtors and creditors.
B. The Purposes of (Imperfect) Bankruptcy Rules
The motivating premise behind the contractualist critique of bankruptcy is that a well-functioning bankruptcy system, by maximizing the insolvency-state returns of creditors, lowers a firm’s financing costs. If mandatory bankruptcy rules stand in the way of this goal, then parties should be able to alter or reject them in a manner that does not harm other creditors. The marijuana industry may have many reasons why its cost of capital remains higher than in other industries, making it impossible to use marijuana companies to test this premise directly.[337] Nevertheless, we can look to the data and the structure of the industry overall to consider whether permitting free contracting might be less important for a company’s cost of capital than other legal interventions. This Section describes how a focus on the specific bankruptcy rules that could be improved upon through contract, rather than on the importance of clear and consistent rules more generally, may be missing the forest for the (marijuana) trees.
To understand why this might be, recall the discussion of the marijuana industry’s legal framework from Part I. The dual status of marijuana—legal at a state level, but federally illegal—imposes substantial uncertainty on the industry. State action to form distinctive state markets and achieve political goals that may be unrelated to maximizing the efficiency of the market can also cause fragmentation and increase costs.[338] Tax, intellectual property, regulatory, and other issues all conspire to make it very expensive to run a cannabis company, and many cannabis companies remain unprofitable despite growing sales.[339]
Marijuana’s dual legal status, and the legal uncertainty it generates, contributes to a cost of capital for the industry that far exceeds that faced by other industries.[340] One way it directly affects the cost of capital is by reducing the certainty of liens placed on company assets.[341] A marijuana company’s most valuable assets would typically include its inventory, licenses, and real estate. But enforcing liens on each of these assets presents challenges. Inventory cannot be repossessed without running afoul of federal law, potentially endangering the lender’s other business interests. Without proper licensure, doing so would violate state law as well. Thus, prominent marijuana-focused law firms do not view repossession of inventory as a viable option.[342]
Next, consider licenses. Generally, liens can be placed on general intangible assets (such as cannabis licenses) even if foreclosure would be barred because of government restrictions on alienability.[343] But a rational lender would discount the value of this collateral to the extent of the difficulty of foreclosing on any collateral so restricted. With the ability to foreclose on a borrower’s most valuable asset depending on government discretion, financing costs will increase.
Finally, cannabis real estate presents several lien issues. If the secured party is not licensed, foreclosing on an active cannabis retail or grow facility could place the lender in immediate violation of state law. Federal law regarding seizure of property used in furtherance of violations of the CSA would also apply. Perhaps these issues could be avoided if the foreclosing lender agreed to use the property for non-marijuana purposes or to lease it to a party that would do so. But this makes underwriting for a lender incredibly difficult, as it is likely that marijuana operations at many properties would constitute the most profitable use possible at that location. For example, a cannabis grow facility prohibited from growing cannabis will generate far less revenue as a non-cannabis greenhouse facility, if it could even be used as such. Without marijuana use, many such facilities would simply be large warehouses in unprofitable locations that could not otherwise justify their cost.[344] A well-located retail facility might better retain value, but in many cases the valuations upon which financing is underwritten might only be achievable on high-margin products like marijuana. The uncertainty of foreclosing on secured collateral, whether inventory, licenses, or real estate, presents a major hurdle for lowering capital costs in the industry. Lien uncertainty might also make the value of a debtor’s ability under the Bankruptcy Code to sell assets free and clear of all liens more apparent to those in the marijuana industry.[345]
High credit costs in the cannabis industry may, at least partially, result from uncertainty that creditors face regarding the amount they stand to recover in case of insolvency. Yet this uncertainty, in turn, does not stem from the inclusion or removal of any one particular bankruptcy rule. It instead arises from the overall uncertainty faced by industry participants across a whole host of dimensions—liens, policy enforcement, tax, intellectual property, contract, and, yes, insolvency. All these uncertainties interact with and compound upon one another: for example, if a company’s intellectual property turns out to be unenforceable, the company’s creditors cannot properly value the intellectual property for collateral purposes even if the company could guarantee that creditors would obtain a recovery on the collateralized assets, which of course the company cannot do. Moreover, the data showing that cannabis companies and their counterparties rarely tailor bankruptcy terms in their contracts suggests that all involved understand that tailored contractual terms would likely only lead to marginal improvements in collectability and thus credit costs. Fiddling with the precise timeline within which a bankruptcy sale must be made, or the exact ways a license would be treated in insolvency, is of profoundly secondary importance compared to the major sources of uncertainty faced by industry participants. An industry with so much uncertainty will ineluctably face higher credit costs, all else equal.
The Creditors’ Bargain theory explains that bankruptcy law should consist of the hypothetical deal that creditors would strike if not faced with collective action problems. Contractualists, in response, take the Creditors’ Bargain theory to its logical conclusion. They argue it would be better to permit parties to implement an actual, customized bargain rather than a hypothetical, one-size-fits-all solution. On the other side, there are those like Warren and Westbrook who are skeptical of the law-and-economics approach of the Creditors’ Bargain theory generally, and who reject contractualism as naïve. They explain that bankruptcy law is needed to promote public welfare goals such as the protection of maladjusting creditors and parties like communities or municipalities that would be affected by the second-order effects of major bankruptcies.[346] But the implied precondition that makes disagreement among these commentators possible is the existence of a large body of settled and predictable debtor-creditor law (whether in the Bankruptcy Code or in the various state contract, secured transactions, and other laws that support its operation). It is only because stable debtor-creditor law exists that we can consider how it could be optimized, as the experience of marijuana companies underscores.
This realization—that a mostly fixed debtor-creditor law is the foundation upon which the contractualist critique (and its detractors) stands—suggests another possible purpose of bankruptcy law writ large. Uncertainty, rather than any particular rule, may be the greater roadblock to smooth commerce and value-promoting distributions. Bankruptcy law and its state-law adjuncts, merely by providing a stable and well-settled set of rules, reduces uncertainty and thus makes the other goals possible. In this conceptualization, the comparison is not between one set of rules and another, or whether sophisticated parties can or should override a rule that does not work perfectly for them, or even whether some parties might be harmed more than those who might gain from bankruptcy contracting. Instead, the comparison is between a system where some consistent set of rules governs and one where it does not. Uncertainty helps explain the higher capital costs faced by cannabis companies. Companies that rationally choose to contract around bankruptcy are sometimes able to obtain better deal terms for themselves, but also benefit by circumscribing the range of potential outcomes that could occur upon insolvency. Meanwhile, the desire to reduce uncertainty in a simple manner could explain the larger number of companies that turn to a few well-established structuring tools for transactions instead of writing bankruptcy contracts.
In this paradigm of consistent versus inconsistent rules, the specifics of the rule matter less than the existence of rules that are understood by, and predictably enforced against, the parties.[347] Increased transactional and informational costs make the presence of some rule, whether optimal or not, essential.[348] As a result, perhaps any movement beyond the bankruptcy state of nature will be an improvement. By contract, a freer bankruptcy regime could counterintuitively increase credit costs, rather than decreasing them, by injecting still more uncertainty into the company’s affairs. The fact that some parties might write bankruptcy contracts while others might not would increase the informational costs to all parties, who would therefore need to consider whether any other contract contains bankruptcy terms that would influence their own contracts.
Perhaps the increased complexity involved when parties transact under unsettled rules offers a partial explanation of why only about one in twenty contracts in the Data Set contains any form of bankruptcy contracting. The problem of having no insolvency rules is so massive that, absent incorporating the entirety of the Bankruptcy Code and all related state law by contract into each agreement, making an agreement about how one particular asset is handled upon insolvency may be missing the forest for a single tree. Indeed, it may be worse than that, since (to extend the analogy) the single tree may be an invasive species that actually adds to problems faced by creditors at large. To return to the most dramatic example of bankruptcy contracting discussed above,[349] the credit agreement entered into by Acreage Holdings, Inc. required marijuana licenses to be auctioned according to a particular process, with lenders having more control over the process and receiving greater returns than they would under the Bankruptcy Code. Would a court enforce these provisions? Would enforcement vary by state or asset type? How should other creditors react under this uncertainty? Could other creditors even discover these provisions or the exact assets to which they applied? And if a key asset is removed from under the noses of other creditors, would they have any recourse? These and a multitude of other questions arise from a single provision in a single contract for a company that likely enters into hundreds or thousands of agreements. Without well-established answers, it may well be that contract counterparties simply decide that taking the risk on a novel contractual provision is not worth the cost.
Perhaps as a result of the complexities added by inconsistent rules, the data demonstrates that parties mostly do not write bankruptcy contracts even when able. The existence of a bankruptcy system that cannot be modified or waived might therefore, at least in part, be justified by consideration of the alternative: no stable bankruptcy law, a general unwillingness to write bankruptcy contracts, insufficient disclosure of contracts that are written, higher informational and transactional costs, inequitable outcomes for many sophisticated and all maladjusting creditors, and potentially higher credit costs as a result.
Conclusion
Marijuana companies cannot file for bankruptcy. This status—unique among American businesses—offers them the opportunity to write contracts that would override the otherwise mandatory bankruptcy rules that apply to all other firms. This Article leveraged a novel, hand-collected data set consisting of almost 75,000 pages across 1,167 contracts and filings publicly disclosed by thirty-four companies in the cannabis industry to examine whether, and how, the marijuana industry writes bankruptcy contracts. It found that despite widespread knowledge of the unavailability of bankruptcy, bankruptcy contracts were generally rare, with only around 6.6% of contracts including any bankruptcy terms. The bankruptcy contracting that did exist was concentrated among secured lenders, where a around one-third of contracts contained tailored bankruptcy provisions. Shaping operations and capital structures to account for the absence of bankruptcy protection was more common, with around 45% of contracts including such strategies. However, most of these documents relied on an opt-out strategy, with more than 85% of contracts with structuring pursuing an opt-out strategy. Cannabis-specific structuring techniques were only used in about 14.5% of contracts with structuring (or 6.5% of all contracts). Finally, companies never issued exotic securities to automate the bankruptcy process. This suggests that bankruptcy contracting may have a limited role to play, but that the costs of bankruptcy contracting, and uncertainty more generally, may be underappreciated by proponents of a bankruptcy contracting regime.
Appendix A—Database Creation and Coding Notes
Parts II and III of this Article are based upon the creation and hand-coding of a novel database on the publicly-available documents for listed cannabis companies. This database consists of 1,167 unique documents, totaling approximately 74,922 pages of legal documents disclosed by thirty-four companies in the legal cannabis industry. The author created this database for purposes of this Article, and is aware of no similar database elsewhere. This Appendix explains how this database was created and coded.
A. Database Creation
Using S&P Capital IQ Pro, a list of all companies that met the following criteria as of May 10, 2023 was created[350]:
- Business Description Includes: “cannabis” OR “marijuana”; OR
- Long Business Description Includes: “cannabis” OR “marijuana”; OR
- Topic Tags Includes: “cannabis” OR “marijuana”; AND
- Geography In: United States, Canada; AND
- Market Capitalization ($M) >= 25
A screenshot of the filter used is provided below:
This generated a list of 113 companies. I then cleaned the list to remove any companies that were included in this list but were otherwise not relevant for the questions examined in this Article. Since the goal of creating this database was to focus on companies that were most directly involved in the cannabis industry, certain companies were removed that could plausibly be considered CSA-violating marijuana businesses,[351] but which likely would not be barred from accessing US bankruptcy courts as a result of this activity. In furtherance of this goal, nine categories of companies were removed from the list, as explained below:
- No US Operations with Cannabis as a Product: approximately twenty-three companies were removed from the list under this criterion. This criterion was meant to address two distinct, but related, reasons why a given company would not be relevant to the questions examined in this Article:
- Several companies on this list simply did not have operations in the United States. With little need to access US bankruptcy courts, and in light of the general hostility of US bankruptcy courts toward cannabis businesses, it is unlikely that such a company would attempt to file for bankruptcy in the US.
- Also included in this tabulation is companies that maintain only federally-legal cannabis-related operations in the US. The 2018 Agriculture Bill legalized the manufacture and sale of products derived from marijuana or that include low levels of cannabidiol.[352] These include hemp-based products, products containing CBD, and other similar products. Because these products are not federally illegal, they do not raise the bankruptcy issues discussed in this Article.
- SPACs: approximately seven companies were removed from the list under this criterion. Special-purpose acquisition companies, or SPACs, raise funds and list on stock exchanges with the intention of acquiring existing private businesses. The SPACs removed under this criterion describe their intention to purchase companies in the cannabis industry, but to date have not done so, and so do not implicate the issues raised in this Article.
- ETFs: approximately seventeen stock listings were removed from the list under this criterion. These exchange-traded funds target investments in companies with varying degrees of connection to in the cannabis industry, ranging from investments in producers or retailers to investments in upstream input providers. ETFs are removed from the list for three reasons. First, ETFs are a step removed from the cannabis companies themselves, and therefore analyzing their contracts would provide less information on the contractual responses to the absence of bankruptcy than analysis of the active companies themselves. Second, analyzing ETFs would almost certainly lead to double-counting, since ETFs likely invest in the same companies that remain on the list. Finally, although equity holders in a cannabis bankruptcy would almost certainly be affected by the absence of bankruptcy protection, ETFs do not negotiate the terms of their investment in equities before making the investments—instead, they simply purchase equities within their mandate for the market price.
- Testing Companies: approximately four companies were removed from the list under this criterion. These companies, rather than trading in cannabis, create products meant to test for the presence of cannabis (say, though the use of a blood test or breath analysis). As such, they do not implicate the issues raised by potential insolvency in the cannabis industry without the protections of the Bankruptcy Code.
- Pharmaceutical Companies: approximately eight companies were removed from the list under this criterion. Many of these companies do not yet have a cannabis-related product to sell, as they are in clinical-stage trials. In addition, none of them require cannabis licensing at a state level because they do not intend to sell cannabis to consumers.
- One company was removed under this criterion because it received approval from the federal Drug Enforcement Agency to produce and sell federally-legal cannabis commercially for research and manufacturing purposes to, inter alia, DEA-registered pharmaceutical companies. As a result, it would likely not face the issues raised in this Article.
- Dual-Use Input Suppliers: approximately 12 companies were removed from the list under this criterion. These companies sell products that are necessary for the manufacture and sale of cannabis and other agricultural products, such as hydroponic supplies, fertilizer, disinfection materials, and packaging. Case law discussed in the body of the Article indicates that courts might look skeptically at so-called dual-use products and thus plausibly bar these companies from filing for bankruptcy. Nevertheless, the particular facts of these cases mattered greatly: when bankruptcy relief was denied, this was because sales to cannabis companies were not an ancillary part of the business, but instead the prime reason for the existence of the business.[353] Thus, companies like Scott Miracle-Gro, for which cannabis is an ancillary business, were removed.
- No Marijuana Tech Companies: four companies were removed from this list under this criterion. These companies offer technologies like payment processing, data analytics, and e-commerce portals for non-marijuana paraphernalia for cannabis operators. Similar to the so-called dual-use input suppliers discussed directly above, these companies might be prohibited from filing for bankruptcy, but also might be able to file for bankruptcy depending on the facts and circumstances of their particular operations. Because these companies do not directly sell marijuana, they were not included.
- No Banks Holding Companies: three companies were removed from this list under this criterion. While the lion’s share of traditional financial institutions have shied away from involvement in the legal cannabis industry for fear of running afoul of federal anti-money laundering and other laws, a handful of smaller banks serve the industry. Despite material ties to the industry, these companies were excluded for two reasons. First, cannabis operations made up only a part of the operations of each bank, meaning that it is not clear whether the prohibition on bankruptcy filings would apply to them (or that they would not be able to evade the prohibition through divestment of the offending parts of the business). Second, and more fundamentally, banks receive unique treatment under the Bankruptcy Code, with bank holding companies being permitted to file for bankruptcy while financial distress at operational divisions of banks is resolved by the FDIC.[354] Since it is not clear whether the Bankruptcy Code would play a role in any given bank contract, these companies were excluded to reduce analytical confusion.
- No US Listing: one company was removed from the list under this criterion. Although this company appears to have some operations in the US, and lists its headquarters as being in the US, the company is only publicly listed on the Korean Securities Dealer Automated Quotations (KOSDAQ) market. As such, no publicly-available English-language documents were available consistent with US securities laws.
After removing these companies, there were thirty-five companies remaining on the list (the “Initial Company List”).
B. Building the Data Set
Once the Initial Company List was constructed, the next step was to build a data set (the “Data Set”) based on the documents made publicly available by the companies on the Initial Company List. To do this, I used Bloomberg Law to search for every public filing by each company on the Initial Company List that mentions the word “cannabis” or “marijuana” posted on the SEC’s Electronic Data Gathering, Analysis, and Retrieval system (“EDGAR,” and searches on EDGAR using Bloomberg Law, “Bloomberg EDGAR”). I reviewed every result (including both the filings and all exhibits) for the five-year period between May 1, 2018 and May 10, 2023. When documents were limited or unavailable on Bloomberg EDGAR for Canadian-listed operators,[355] I also cross-referenced the documents included on Bloomberg EDGAR with documents included on Canada’s System for Electronic Document Analysis and Retrieval Plus platform (“SEDARPlus”) for companies listed in Canada.
After removing one company which returned no documents on a search of Bloomberg EDGAR, the final list of companies (the “Company List”) consisted of thirty-four (34) companies. [356]
I downloaded the following categories of results, which I used to code the following information[357]:
Category | Coded Information (more detail given below) | Notes |
10-K or others[358] | Listing jurisdictions Marijuana in business description[359] Marijuana risk factors[360] Reliance upon trade secrets due to the unavailability of federal trademark and/or patent protection[361] Unionized employees[362] Bankruptcy risk factors[363] Limited reporter under JOBS Act Corporate structure Is there a separate entity for each property? Is there a separate entity for each license? Is there a single entity for all licenses in a state? Is there a single entity for all property in a state? Exotic securities[364] Total long-term debt | 10-Ks can include many different types of exhibits. Some of these are pertinent to the questions in this Article, such as management’s discussion and analysis, loan agreements, leases, employment contracts, indemnities, consulting agreements, and others. |
Organizational Documents[365] | State of incorporation Do the organizational documents provide for a preauthorized sale of all assets? Marijuana in business description Marijuana risk factors Enumerated bankruptcy process Bankruptcy-specific rights[366] Authorization for state-level procedures Exotic contract provisions | Only Articles where the entity is a US entity; if the entity is a Canadian entity, then it would be governed by Canadian law and would not implicate the issues addressed in this Article. |
Merger Agreements[367] | Marijuana risk factors Marijuana-specific provisions[368] Bankruptcy risk factors Bankruptcy-specific rights[369] State-level procedures Involvement of US companies Governing law | Merger agreements contain numerous disclosures, and sometimes contain financing arrangements that would be relevant to the questions addressed in this Article. While many occur under Canadian law and primarily involve Canadian companies, US companies are often involved enough that their bankruptcy might have a material impact on the transaction. |
Secured Debt | Borrower domicile Governing law Does the collateral include the license?[370] Does the collateral include marijuana?[371] Does the collateral include intellectual property? Does the collateral include cash? Are there mezzanine-style pledges of interests in upper-tier or related entities? Is there a parent guaranty? Number of financial creditors Multiple layers of creditors?[372] Marijuana risk factors Marijuana-specific provisions Bankruptcy risk factors Bankruptcy-specific rights Can the lender exercise significant control?[373] Does the lender have audit rights?[374] | Secured debt comes in many forms. Some of the types of secured debt examined include secured convertible notes, construction loans, pledges, and others. If a license is held by an entity which pledges its ownership interests in favor of a lender, this is treated as though the license is also pledged as collateral (since the reasonable expectation of the lender would be that the license would transfer along with the ownership interests, subject to the same transfer restrictions that would apply to a transfer of the license itself). |
Unsecured Debt | Borrower domicile Governing law Does the loan agreement explicitly state its repayment priority (i.e., senior or subordinated)? If so, is it senior or subordinated? Existence of secured vs. unsecured debt Are there multiple tiers or layers of borrowers or guarantors?[375] Number of financial creditors Multiple layers of creditors? Bankruptcy-specific rights | Unsecured debt comes in many forms. Some of the types of unsecured debt examined include unsecured promissory notes, convertible promissory notes, and debt paid by share or warrant issuance. |
Leases | Does the collateral include the license? Does the collateral include marijuana? Does the collateral include cash? Special treatment of marijuana-related fixtures? Special treatment of on-site inventory? Are there mezzanine-style pledges of interests in upper-tier or related entities? Are there lease guarantors? Marijuana risk factors Marijuana-specific provisions Bankruptcy risk factors Bankruptcy-specific rights Audit rights | Because cannabis companies often face difficulties obtaining traditional financing, sale-leaseback arrangements are a common and important method used to raise capital. As a result, leases can contain loan-like terms, including pledges of collateral, guaranties, financial and operational covenants, and others. |
Material Contracts | Marijuana risk factors Marijuana-specific provisions Bankruptcy risk factors Bankruptcy-specific rights Governing law | Inclusion of a contract as an attachment to a 10-K, or inclusion as part of an 8-K filing, indicates that it is material. |
A relatively small number of documents (fewer than twenty) were removed after examination because they expressly disclaimed any cannabis-related operations in the United States.[376] These typically related to companies, such as SNDL Inc., which have substantial cannabis retail operations in Canada but no retail operations in the United States.[377] In the case of SNDL, Inc., the company also has a joint venture through which it makes loans to cannabis companies worldwide, including in the United States. Documents for this company, such as secured credit documents memorializing loans made to non-US companies that held no cannabis-related assets in the United States, were therefore excluded.[378] Nevertheless, the companies associated with these documents were still analyzed if they had documents that were still valuable for other parts of the analysis.
In total, the Data Set contained 1,167 documents (with all amendments to a document being counted with the original document as a single document), composed of 74,922 pages.
C. Exclusions from the Data Set
Although every publicly-available document was examined, several types of documents were consistently removed from the data set. After a time, these documents were excluded entirely (and therefore were not reviewed once they were identified as one of the document types described below).
These excluded documents, the corresponding forms used when filed with the SEC, and reasons for their exclusion, are described below:
- 8-K Press Releases: as these releases did not contain actual legal documents, they were ignored.
- 10-Q: 10-Q information is already included in the corresponding 10-K, in even greater detail.
- 10A: Registrations of securities will cover the same general issues, descriptions of securities, and risk factors as 10-Ks.
- DEF-14A, 14A: proxy statements are irrelevant to the issues in the Article.
- 6-K: duplicative of 8-K for foreign issuers.
- 20-F: a restatement of 10-K.
- D: Notice of Exempt Offering of Securities.
- Beneficial ownership information is ultimately irrelevant to the issues examined in this Article, and therefore was excluded. These forms include:
- 3: initial statement of beneficial ownership.
- 4: changes in beneficial ownership.
- 13D: beneficial ownership report.
- 13G: movement of ownership from an individual to a trust.
- S-8: the allocations under stock incentive plans are irrelevant to the issues examined in this Article.
- SD: specialized disclosure form.
- 424B3, 424B7, S-1, S-1/A, S-3: prospectus; but, attachments to prospectuses were reviewed, as these sometimes contained material documents not easily accessible elsewhere.
- Comment Letters to/from SEC Staff: correspondence between the SEC and the respective issuers relating to securities laws requirements were outside of the scope of the Article.
- Miscellaneous: items such as letters of intent were excluded if a corresponding document fully memorializing the agreement reached in the letter of intent could not be located.
D. Coding for Bankruptcy Contracting
Contracts that included one or more express provisions that incorporated, altered, or rejected procedures that would typically be implemented under the Bankruptcy Code would be coded as a form of bankruptcy contracting. Such rights might include:
- Mirrors of rights under the Bankruptcy Code, such as:
- An automatic stay.
- Delays in payments of interest or use of cash collateral.
- Any collective action mechanism that could mirror what might be included in a restructuring support agreement or the voting rights present in the Bankruptcy Code.
- Any examination or audit rights.
- Express deviations from rights under the Bankruptcy Code, such as:
- Replacing the sales process under section 363.
- Changing rights under section 365.
- Permitting parties to estimate the value of claims.
- Appointment of an administrator or receiver.
- Implementations of state-level procedures, such as:
- Appointment of a state-approved receiver.
- Statements regarding governing law in the event of insolvency.
- Barring or facilitating ABCs.
- Any more customized, more exotic form of bankruptcy contracting that may be present.
E. Coding for Structuring
Contracts that included bankruptcy structuring provisions would be coded as such if they (a) were entered into by non-plant touching companies, or (b) contained one or more of the following cannabis-specific structures:
- Separate legal entities for each cannabis property.
- Separate legal entities for each cannabis license.
- Using a single entity to own all properties in a given state.
- Using a single entity to hold all licenses in a given state.
- A preauthorized all-asset sale.
- Any special treatment of marijuana fixtures or inventory in leased space.
- Any other procedures for keeping useful blocks of assets together, such as by including at least two of the following items as security:
- Real property
- State licenses
- Inventory
* * *