Indiana Law Review Defending Purchase Money Security Interests Under Article 9 of the UCC From Professor Buckley Paul M. Shupack* I. Introduction At least since Karl Marx, economists have moved from theory to practice with what, in hindsight, appears to have been unseemly haste. * Those who analyze law from the view of economics have tended to share this trait. Professor Francis H. Buckley, in his article, The Bank- ruptcy Priority Puzzle,^ recommends immediate changes in our legal practice in light of the insights resulting from his theoretical inquiry. This Article argues that no persuasive reasons exist to accept the changes Professor Buckley proposes for Article 9 of the Uniform Commercial Code (*'UCC"). His theory does not translate into practice as easily as he appears to believe it does. Professor Buckley's argument, developed in counterpoint to a lit- erature which asks why the law permits any secured lending,^ questions any statutory regulation of secured debt. He observes that **management is the decisionmaking body best able to determine the particular firm's value-maximizing secured debt policy.'"* He then asks why the UCC imposes what he calls mandatory barriers to management discretion in creating secured transactions. While he defends some of these mandatory rules, he finds inadequate the arguments conventionally used to justify * Professor of Law, Cardozo School of Law, Yeshiva University; A.B. Columbia, 1961; J.D. University of Chicago, 1970. I wish to thank my colleagues WilHam Bratton and David Carlson for their help with this article. 1. "The philosophers have only interpreted the world in various ways; the point, however, is to change it." K. Marx, Theses on Feuerbach {reprinted in 2 K. Marx & F. Engels, Selected Works at 405 (1962). 2. Buckley, The Bankruptcy Priority Puzzle, 72 Va. L. Rev. 1393 (1986). 3. This literature includes Jackson & Kronman, Secured Financing and Priorities Among Creditors, 88 Yale L.J. 1143 (1979); Jackson & Schwartz, Vacuum of Fact or Vacuous Theory: A Reply to Professor Kripke, 133 U. Pa. L. Rev. 987 (1985); Kripke, Law and Economics: Measuring the Economic Efficiency of Commercial Law in a Vacuum of Fact, 133 U. Pa. L. Rev. 929 (1985); Levmore, Monitors and Freeriders in Commercial and Corporate Settings, 92 Yale L. J. 49 (1982); Schwartz, Security Interests and Bank- ruptcy Priorities: A Review of Current Theories, 10 J. Legal Stud. 1 (1981); Schwartz, The Continuing Puzzle of Secured Debt, 37 Vand. L. Rev. 1051 (1984); Scott, A Relational Theory of Secured Financing, 86 Colum. L. Rev. 901 (1986); White, Efficiency Justifications for Personal Property Security, 37 Vand. L. Rev. 473 (1984); Shupack, Solving the Puzzle of Secured Transactions, 41 Rutgers L. Rev. 1067 (1989). 4. Buckley, supra note 2, at 1469. 777 778 INDIANA LAW REVIEW [Vol. 22:3 Others. Among the rules he finds most open to question are those creating superpriority for purchase money security interests (*TMSr'). Under the UCC, a first-in-time^ secured lender who wants to use the debtor's after-acquired property as further security for a loan cannot be certain that the debtor's after-acquired property will be effectively available as collateral. Despite the terms of any contract between the debtor and a first-in-time secured creditor, the debtor may acquire that property subject to a PMSI, to which the UCC awards a priority ahead of the first-in- time secured party. Professor Buckley argues that this inability of cred- itors and debtors to bargain away the consequences of the statutory scheme imposes unnecessary costs on first-in-time secured creditors and their debtors. Against Professor Buckley's conclusion that the phenomenon of PMSI superpriority cannot be justified on efficiency grounds, this Article argues that the existing legal order can be rationally defended. In this type of Kaldor-Hicks efficiency discussion, the absence of data prevents the possibility of proof.^ What follows is very much in the spirit of Professor Buckley's article. It offers reasons to believe that if data did exist, it would show that PMSIs would be efficient, even using Professor Buckley's special test for what he calls mandatory rules, though it also takes issue with the adequacy of that test to determine any question of public policy. This Article shows that enough debtors and secured creditors with after-acquired property clauses in their loan agreements will want to allow for PMSI superpriority that the probable transaction costs saved by the existing arrangements will outweigh the probable costs to debtors and secured parties resulting from these restrictions. That demonstration alone makes Professor Buckley's argument somewhat implausible. Moreover, Professor Buckley does not consider the costs that his alternative regime of free contract would impose on both first-in-time and second-in-time secured lenders. The first-in-time secured lenders would find that they had added drafting and collateral realization costs under a free contract regime compared to those costs under the UCC. Under the UCC, first-in-time secured lenders potentially bear costs of 5. By first-in-time, I mean the secured lender who first makes a claim in a security agreement to collateral that qualifies as after-acquired property. Until the debtor has rights in the collateral, no secured party can have a claim to the collateral, and, as a result, the issue that this Article addresses, if described precisely, is to which of two simultaneously- attaching security interests does the law give priority. 6. For a transaction to be efficient in a Kaldor-Hicks sense, it must be one in which the utility gains of the winners exceed the utihty losses of the losers, but there is no need for the winners to compensate the losers. See, e.g., A. Feldman, Welfare Economics and Social Choice Theory, 142-44 (1980). 1989] SECURITY INTERESTS 779 certain types of debtor dishonesty. Professor Buckley notes that his regime of free contract will eUminate these risks to first-in-time secured lenders, but he does not mention that these risks are simply transferred to second-in-time secured lenders. There is no saving here, and in fact there is some loss because second-in-time lenders are less able to protect themselves against these risks. Once this more complex accounting is done, Professor Buckley's argument for the relative efficiency of a free contract regime becomes increasingly implausible. //. Professor Buckley's Test for **Mandatory Rules'* When Professor Buckley analyzes what he calls mandatory legal rules in the UCC, he introduces a novel test by which these rules may be judged. In his view, to justify these mandatory legal rules, it is not enough to show that the effect of this type of rule creates efficient behavior. One must also show that * 'transaction costs or other barriers would prevent the parties from devising optimal priority rules in non- mandatory regimes.''^ Professor Buckley's example of a ** mandatory'* rule is that part of UCC Section 9-307(1) which states, **A buyer in the ordinary course of business . . . takes free of a security interest created by his seller even though the security interest is perfected and even though the buyer knows of its existence."^ This subsection gives buyers in the ordinary course of business rights superior to those of the seller's secured lenders in any goods bought from the seller. Professor Buckley is satisfied that this statutory provision is fully justified by his novel efficiency test. Under that test, it is not enough that the parties, if left to their own devices, want to reach a result equivalent to the one reached by UCC Section 9-307(1). An additional element must be satisfied. If Section 9-307(1) did not exist, the parties would find it difficult and expensive to reach its result. Without man- datory superpriority, the screening process that buyers would have to undertake, e.g., the review of the seller's financing documents, "seems extremely inefficient. Mandatory superpriority rights provide a simple solution to the problem."^ It is common to observe that legal rules, and especially contract rules, have the effect of providing ready-made patterns for action. These ready-made patterns create a savings for those who use them, modest costs for those who contract out of them, and potentially major costs for those who should have contracted out of them, but failed to do 7. Buckley, supra note 2, at 1452. 8. U.C.C. § 9-307(1) (1972). 9. Buckley, supra note 2, at 1454. 780 INDIANA LAW REVIEW [Vol. 22:3 SO. The difference between **many*' and "all" justifies preserving the possibility of freedom of contract even when one is certain that the ready-made pattern is one that most would prefer. Should the legislature require the pattern, the requirement would prevent those who did not want it from contracting out of it. Professor Buckley argues that before a legislature requires a rule which results in efficient behavior, it ought to satisfy itself that some barrier exists to prevent the contracting parties from reaching the result they are assumed to want. There are two problems with this otherwise attractive test. First, to the extent that a legislature does not make a rule * 'mandatory," it may well end up imposing costs on contracting parties. To the extent these costs exist, they must be weighed against the costs imposed on those who dissent from the mandatory rule. Second, the idea of a mandatory rule is itself somewhat problematic, especially in light of the way Buckley uses his own term. If the meaning of a term may be understood from its use, then we can understand Buckley's concept of a **mandatory" rule through an examination of Section 9-307(1), which he has declared to be a good mandatory rule. While Section 9-307(1) gives superpriority to a buyer in the ordinary course of business, any buyer who would qualify as a buyer in the ordinary course of business could, by expHcit contract, waive his right to take goods free of the security interest created by his seller. If a generous buyer did so, courts would uphold the consequences of such a waiver for buyer, seller, and seller's secured party (absent of course the defenses of unconscionability and the Hke which would be suggested by such an altruistic act in the midst of an ordinary commercial transaction). Calling this type of rule mandatory uses that word in a sense that differs from its ordinary usage. When Professor Buckley uses the term **mandatory," he would have us look not to the relationship of buyer and seller, but rather to the seller and his secured lender. He would have us ask what the effect is under the law if a seller, whose business made his buyers "buyers in the ordinary course of business," were to contract with his secured party to seek a waiver of the effect of UCC Section 9-307(1) on each sales transaction. Again, nothing in the law would prohibit the parties from doing so or from ordering their affairs in accordance with this type of contractual system. The seller may find few buyers willing to purchase goods agreeing to this type of waiver, but if the seller can find them and the buyers consent, then the law is satisfied. Problems arise only when the seller breaches his contract with the secured lender and sells goods without obtaining the waiver called for under the contract with the secured lender. In this event. Section 9- 307(1) will deprive the secured party of any remedy against the goods. At this point, however, the secured creditor still would have a claim 1989] SECURITY INTERESTS 781 against the seller based on breach of contract, a legal remedy which concededly she neither wanted nor contracted for. To understand what Professor Buckley means by a mandatory legal rule, one must ask whether the so-called mandatory legal rule has the effect of making contractual provisions unreliable in their effect in the event one of the parties breaches the contract. The rule creating man- datory superpriority for buyers in the ordinary course of business makes unreliable any secured party's attempt to provide by contract for security interests that survive sale by her debtor of the goods serving as collateral. The law does not forbid parties from so contracting. Nor does the law leave the creditor suffering from breach of such a contract without remedy for that breach. The law does, however, leave the aggrieved party without the contracted-for remedy, thus imposing costs that the aggrieved party will find unacceptable. There is thus no difference in kind between an ordinary background rule of law, filling the gaps in the absence of agreement of the parties, and a rule of law which is '^mandatory." A '^mandatory" rule, in Buckley's usage is simply one that creates a result which imposes unacceptable costs on those who dissent from the rule. III. Professor Buckley and the Purchase Money Security Interest Using his test. Professor Buckley questions those provisions of Article 9 of the UCC law that create mandatory superpriority for PMSIs.*^ Under the UCC, a lender who takes a security interest in the debtor's after-acquired collateral cannot prevent the debtor from creating a secured lender second-in-time, whose interest in the after-acquired collateral will be prior in right to that of the secured creditor who is first-in-time. By having PMSI status, a second secured lender can effectively remove the benefit of the first secured party's after-acquired property clause. In Professor Buckley's view, '*[b]ecause security interests in after- acquired property give rise to readily apparent efficiencies, the exception for PMSI superpriority appears to require a countervailing allocational justification."'^ Professor Buckley then examines the *'four basic the- ories" defending PMSI superpriority, and finds each wanting. A secured creditor with an after-acquired property clause has a situational mo- nopoly, but, says Professor Buckley, if the law did not create PMSI superpriority, '*any inefficiencies associated with an issue of after-ac- quired property interests [could be] cured ex ante in the first loan, which could itself provide for PMSI superpriority if it were efficient to do 10. Id. at 1462. 11. Id. 782 INDIANA LAW REVIEW [Vol. 22:3 so."^^ The second justification for PMSI superpriority, which, as Buckley observes, is a variant of the first argument, notes that if a secured creditor with an effective after-acquired property clause is certain that the debtor's existing line of business will generate revenues sufficient to pay off the loan, then the secured creditor will have disincentives to permit the debtor to enter into high risk, high return new ventures. Buckley does not agree that the situation generates these sorts of dis- incentives because, as he notes, the first creditor can always make a second loan at an interest rate different from that of the first loan.'^ To the extent that first-to-file creditors do take advantage of their situational monopoly and engage in strategic behavior, ex ante contracting by the debtor reserving the right to create PMSI superpriority again operates as an effective prevention of the evil the statutory PMSI su- perpriority is supposed to cure.*'* The third and fourth arguments that support statutory PMSI su- perpriority presuppose secured creditors with significant differences among themselves concerning their information about the debtor or the collateral. The logic of this theory leads to the conclusion that an after-acquired property clause benefitting such a monitoring secured creditor leads to a * 'sapping" of *'the most plausible monitor's incentive to police the debtor's behavior,"'^ which in turn can be controlled by PMSIs. Buckley observes that if the problem is the loss of incentive to monitor by the original secured party, that incentive problem can be solved directly by recreating pre-UCC rules which conditioned the effectiveness of after- acquired property clauses on the creditor's monitoring of the after- acquired property.'^ The other branch of justifications based on the creditor information differentials also contains the explanation that Professor Buckley finds most plausible as to why the law might favor PMSIs. To the extent that creditor monitoring of particular items of collateral guards against debtor misbehavior, and to the extent that creditor capacity to monitor is a function of the precise collateral involved, then the PMSI, with its capacity to tie a specific creditor to specific collateral, aids in bringing about that assignment at low cost. (Similarly, to the extent that creditors 12. Id. at 1463. 13. Id. at 1464. 14. Id. at 1464 n.l58. 15. Id. at 1464-65. 16. Id. at 1465. Buckley's account uncovers yet another obscurity in the monitoring theory of secured transactions. If the justification for security interests is at bottom the benefit the monitor performs for herself and the general creditors, it is not at all obvious how the general creditors are helped by monitors whose advantage lies in the collateral rather than in the debtor. 1989] SECURITY INTERESTS 783 have differing knowledge of the value of assets on default, it is likely that sellers of the collateral will also be in a better position to value the collateral in the event of default) The PMSI again operates to bring about an easy assignment of the "right" collateral to the ** right" creditor. To both of these points, Buckley gives his familiar answer—to the extent that such regimes do exist, then if allowed to do so by law, debtors would agree ex ante with their first creditors to retain the right to create these sorts of PMSIs.'^ IV. Defending PMSI Superpriority from Professor Buckley > With respect to three of the four objections he notes. Professor Buckley observes that if the parties wanted PMSI superpriority, and if the law did not already give it to them, they could create it by contract. The repetition in his analysis of the phrase **if the debtor wants PMSI superpriority, he could contract for it ex ante'' should stand as warning that the analysis cannot be complete. So long as the analysis concerns reality, where contracting is itself a costly affair, the recognition that parties will repeatedly contract ex ante to bring about a particular result must suggest the possibility that a legal rule which brings about exactly that result is one that will not only be efficient on Kaldor-Hicks principles, but will also reveal a problem with Professor Buckley's additional test. If enough people want the result required by a rule, then the commonplace Kaldor-Hicks efficiency account for having a rule at all 17. Id. at 1465-66. In his discussion of the possible advantages of the UCC's system of PMSI superpriority, Professor Buckley notes that creditor advantage with respect to collateral could create efficiencies by decreasing screening costs for all creditors. He then goes on to say, in an uncharacteristically opaque passage: Moreover, it is not at all clear that PMSI superpriority decreases screening costs. PMSI superpriority may compound uncertainties in the valuation of the debtor's assets. It requires the debtor and the after-acquired property financer to estimate not merely the anticipated value of the firm on default but also the risk that anticipated collateral may be lost to subsequent PMSI lenders. In this way PMSI superpriority may actually increase screening costs and thus the cost of credit. Id. at 1466. The passage fails to tell the reader against what measuring point there is increase and decrease. Presumably the increases and decreases are to be measured against a regime of free contract. Even if it were true that screening cost savings created by mandatory PMSI superpriority were outweighed by the screening costs, it does not necessarily follow that the cost of credit would "actually increase." This passage concludes a discussion in which Professor Buckley has described three other reasons to believe that mandatory PMSI superpriority might decrease the cost of credit. He has said that as to two of them, they plausibly suggest reasons to believe that PMSI superpriority does in fact reduce the price of credit. Thus even if screening costs were increased by mandatory superpriority, the reductions in costs of credit described in the other two theories might well mean that mandatory PMSI superpriority nonetheless does reduce the price of credit. 784 INDIANA LAW REVIEW [Vol. 22:3 comes into play. The saved transactions costs for those desiring the result will outweigh the costs of those who now need to contract out of the result. If, however, the rule is a mandatory rule, the costs of contracting out of the rule are presumed to be so great that, as a practical matter, everyone is stuck with the rule. Instead of having the inconvenience and cost of contracting out of a rule's effect, those not wanting the rule now have the full cost of that unwanted effect. It is this additional cost that must be weighed against the savings to those contracting parties who, in the absence of mandatory rule, would have to provide ex ante for its result. Even so, there remains the possibility that the convenience of the many will, at some level, provide a Kaldor- Hicks justification for a mandatory rule. This result would occur if the costs the many would incur to contract into the rule's result would exceed the costs to the dissenters of having to follow the rule. It is precisely this calculation that Professor Buckley's special test does not permit. Yet there does not appear to be any reason, other than an a priori preference for freedom of contract, no matter what its costs to members of society, for excluding this type of calculation from a Kaldor- Hicks efficiency analysis. In the absence of data, weighing these two sets of values against each other is admittedly difficult. If, however, it can be shown that hardly any contracting party is likely to want a result other than that called for by the mandatory rule, and if it can be shown in addition that those few who are burdened by the rule would not consider that burden to be a heavy one, then these values suggest that Kaldor-Hicks criteria could be satisfied by a mandatory rule. If it can also be shown that the absence of the mandatory rule creates its own burdens, and especially if these burdens are likely to be heavy to the many who would prefer the result required by the mandatory rule, then it becomes ex- tremely plausible to believe that it is efficient to let that result be a consequence of a mandatory rule. A. Virtually No Debtor Wants Out of PMSI Superpriority Professor Buckley asserts that the efficiencies of security interests in after-acquired property are readily apparent. The efficiencies are readily apparent if one is concerned with the relations among a debtor, a secured party and the general creditors. Those efficiencies are not, however, apparent in the relation among debtor, the first-in-time secured party and second-in-time secured party. Analysis suggests that there are very few debtors who would freely consent to eliminating by contract their capacity to create PMSI superpriority. Behind the UCC's after-acquired property rule lies the assumption that once a secured creditor takes a security interest in property, the 1989] SECURITY INTERESTS 785 debtor's productive activity will produce revenues which could be used either to repay the creditor or buy more property. The after-acquired property allows the debtor to offer substitutes for the original collateral without having to go through the formality of paying off the original loan and taking out a new loan. Thus, within the UCC, the debtor's use of his own funds is made the equivalent of using a new advance from the first secured lender.'* The transactional efficiencies arising by means of the automatic perfection of the secured party's interest in the after-acquired collateral are obvious. An enormous amount of paperwork tracing the flow of funds from the debtor to the creditor and back to the debtor, is eliminated.'^ PMSI superpriority arises only in circumstances in which the source of funds for purchasing the after-acquired property is not the debtor, but a second secured creditor. The question, then, is what are the circumstances that would cause a debtor to choose to go to a second secured lender rather than either returning to the first secured lender or using his own assets to replace the security originally provided to the first lender. While it is possible that the second creditor could give the debtor a better deal than the first creditor, it is by no means obvious how this circumstance would come about. Both first and second secured creditors sell their funds in extremely competitive capital markets, and presumably have similar investment opportunitites. Thus, it is unlikely that there is something about second secured creditors that give them inherently cheaper money. If there were, one would wonder why they were not first secured creditors. It is possible that the advantage of the second secured creditor is simply that she is second-in-time. To the extent that the first secured creditor has search or investigative costs on which the second secured creditor can free-ride, that saving could account for the capacity of second secured creditors to do better by debtors who have granted security 18. To the extent that the source of funds is an unsecured creditor, so long as ostensible title issues can be assumed to be solved, for purposes of this discussion there is no difference between a debtor using his own funds or borrowing them from a knowledgeable general creditor and paying the appropriate risk premium. 19. Whether this transactional efficiency is also Kaldor-Hicks efficient cannot be said with certainty unless one assumes continuous full employment. The claim of efficiency does not take into account the welfare of the small army of clerks that the adoption of the UCC left unemployed. A further Kaldor-Hicks analysis would have to ask whether, with the increase in volume of financing activity that the UCC created, enough of this group became reemployed. If so, then further analysis would require evaluating alternative life-time income flows of these clerks, with or without the UCC, measured against the gains to financial institutions, as well as the presumed reduction in prices paid by ordinary consumers as a consequence of a lower price of credit. 786 INDIANA LAW REVIEW [Vol. 22:3 interests in their after-acquired property. The UCC provides ample op- portunitites for second secured creditors to take advantage of first secured creditors in this way. Thus, one would expect that in loan contracts now entered into under the UCC there would appear some type of protection for the first secured creditor against this risk. These defensive measures to costs that mandatory PMSI superpriority would appear to impose on the first secured lender are Ukely to include prepayment penalties or negative pledge agreements. Professor Scott, whose theory of secured transactions predicted that these types of clauses would appear whenever creditors created blanket Hens (including after-acquired property clauses), went out to look for them and could not find them. Their absence, as he freely admitted, put the validity of his theory into question. ^^ Their absence also makes this explanation of second-in-time secured creditor's advantage unlikely. If these search and investigative costs must t>e incurred by any creditor, and, if it is assumed that the second secured creditor does not take advantage of the first secured creditor's information, then the capacity of the second secured creditor to give the debtor a better deal than the first secured creditor becomes even more mysterious. When the debtor seeks a new loan, the first-in-time creditor's sunk costs ought to enable her to make that loan to the debtor at rates lower than any second-in-time creditor, since she need not incur costs to acquire in- formation she already has. Yet the reality remains that second-in-time secured creditors can and do give better deals to debtors than do first-in-time secured creditors. The principal source of the second-in-time secured creditor's advantage must lie in the collateral itself. The second-in-time secured creditor, when judging the value on default of the collateral she finances, must place a higher value on that collateral than does the first-in-time secured creditor. Evidence within the UCC permits the conclusion that the drafters beUeved that financing sellers had special capacities to realize value in the event they repossessed the goods they sold, and they systematically provided statutory advantages for those sellers. There is also substantial evidence that the drafters had good reason for their belief. B. PMSI Superpriority within the UCC 1. The Statutory Provisions.—The very complexity of the provisions within the UCC creating PMSI superpriority suggests some of the reasons for having PMSI superpriority. In the contest between later PMSI holders 20. Scott, supra note 3, at 951, Scott also reports that the equivalent of mortgage points, a device familiar to anyone who has mortgaged the family home, is also virtually unknown in personal property security. 1989] SECURITY INTERESTS 787 and first-in-time secured creditors with after-acquired property rights, both Article 2 and Article 9 have something to say. Within Article 9, section 9-312 is the principal arbiter. If the collateral is inventory, subsection (3) gives superpriority to the PMSI lender, provided she gives proper notice to the first-in-time secured party whose interest in the same collateral has been created by an after acquired property clause. ^^ In collateral **other than inventory," subsection (4) gives superpriority to the PMSI, provided the PMSI lender perfects within 10 days of the date when the debtor ''receives possession" of the collateral. ^^ The rights that Sections 2-702 and 2-705 give to sellers of goods have the effect of creating a PMSI in goods sold. The rights under Section 2-705 are superior to those of a first-in-time secured creditor with rights to after- acquired property, while PMSI rights under Section 2-702 are subor- dinated to a secured creditor with rights in after-acquired property. ^^ 21. U.C.C. § 9-312(3) states: A perfected purchase money security interest in inventory has priority over a conflicting security interest in the same inventory and also has priority in identifiable cash proceeds received on or before the delivery of the inventory to a buyer if (a) the purchase money security interest is perfected at the time the debtor receives possession of the inventory; and (b) the purchase money secured party gives notification in writing to the holder of the conflicting security interest if the holder had filed a financing statement covering the same types of inventory (i) before the date of the filing made by the purchase money secured party, or (ii) before the beginning of the 21 day period where the purchase money security interest is temporarily perfected without filing or possession (subsection (5) of Section 9-304); and (c) the holder of the conflicting security interest receives the notification within five years before the debtor receives possession of the inventory; and (d) the notification states that the person giving the notice has or expects to acquire a purchase money security interest in inventory of the debtor, describing such inventory by item or type. 22. U.C.C. § 9-312(4) reads: A purchase money security interest in collateral other than inventory has priority over a conflicting security interest in the same collateral or its proceeds if the purchase money security interest is perfected at the time the debtor receives possession of the collateral or within ten days thereafter. 23. Under section 2-702, a seller has a right against an insolvent buyer to reclaim goods "upon demand made within ten days after the receipt." But this right of reclamation is "subject to the rights of a buyer in the ordinary course of business or other good faith purchaser under this Article (Section 2-403)." A holder of a perfected security interest in the same goods would qualify as a "good faith purchaser under this Article." See Gilmore, The Good Faith Purchase Idea and the Uniform Commercial Code: Confessions of a Repentant Draftsman, 15 Ga. L. Rev. 605, 616 ff. (1981). Under section 2-705, the seller may stop delivery of "carload, truckload, planeload or larger shipments of express or freight when the buyer repudiates or fails to make a payment due before delivery or if for any other reason the seller has a right to withhold or reclaim the goods." This 788 INDIANA LAW REVIEW [Vol. 22:3 PMSI lenders do not have to accept the superpriority the statute offers them. In the case of inventory, a PMSI lender can simply choose not to give the required notice, and in any event, for every sort of collateral. Section 9-316 authorizes *'any person entitled to priority'' to subordinate her priority by agreement. Despite the expansive wording of Section 9-312(4), PMSI super- priority rights under this subsection are effectively limited to tangible collateral. In his discussion of Section 9-312(4), Grant Gilmore char- acterizes his discussion of the possibility of PMSI interests arising in general intangibles as being **almost on a hypothetical level. "^'^ Gilmore's characterization, if anything, overstates the possibility of applying Section 9-312(4) to intangibles, since he did not consider the effect of the requirement that the PMSI lender perfect before or within 10 days of when the debtor '* receives possession" of the collateral. The metaphysical difficulties involved in receiving possession of anything intangible are formidable.^^ When a secured lender rehes on priority rights to after-acquired intangibles, her expectations may be defeated by making those rights tangible and then having the tangible version of those rights acquired by a good faith purchaser. If, for example, a lender rehes on a pool of continuously renewing accounts receivable, a debtor could defeat the creditor's interest by having his customers embody their obligations in negotiable instruments. The debtor could then sell these instruments to a person who qualifies as a holder in due course, leaving the original secured creditor without assets she legitimately expected to be available to her. Consistent with his views concerning mandatory PMSI super- priority, Professor Buckley would abolish negotiability, at least to the extent it has the consequence of subordinating after-acquired property financers. The real problem is, of course, the risk of debtor dishonesty.^^ section does not offer any general protection to the secured party as a good faith purchaser. See A. Farnsworth & J. Hunnold, Commercial Law, Cases and Materials 720-50 (4th ed. 1985). 24. 2 G. Gilmore, Security Interests in Personal Property § 29.5 (1965). 25. Gilmore limits his discussion of this requirement in the statute to a one sentence cross reference to the parallel language in section 9-312(3). See id. at 799. With respect to inventory, section 9-312(3)'s reference to "receives possession" creates no mysteries. 26. Cf. Buckley, supra note 2, at 1467. A debtor who under the UCC would convert accounts receivable into tangible paper, would, under Professor Buckley's alternative regime, be likely to forge records of accounts receivable. Nor is it so certain that because "the market for negotiable instruments . . . has largely disappeared," negotiability has lost its importance for financers. Id. at 1469. One use for negotiability in commercial financing is in bankers acceptances. The Federal Reserve reports that for the six months of April to September of 1986, the volume of bankers acceptances was approximately $12 billion per month, of which approximately $2 billion per month consisted of drafts 1989] SECURITY INTERESTS 789 2. One Policy Rationale for the UCC Provision.—There can be no doubt that creditors, when examining a debtor's tangible property, con- sider the costs of repossessing and selling that property. Creditors will assign different values to the property. If a creditor has property assigned to her, one component of that valuation has to be the creditor's own capacity to resell the goods. If all other things are equal, a rule of law that assigns the property to the highest valuing creditor will do only what a debtor would choose to do himself.^^ One creditor who is highly likely to have advantages over other creditors in disposing of property will be the creditor who first sold it to the debtor on credit. There is evidence that this type of repossessing seller can on resale sometimes realize an amount in excess of the original contract price. ^^ That seller is Ukely to have her greatest advantage if the goods are still in the same form and packaged for transport in the same way as that seller had left them. Not surprisingly. Article 2 creates a PMSI in favor of reclaiming sellers and reserves its strongest security interest for the creditor-seller of goods who stops them while they are in transport. purchased by the holding bank. 73 Fed. Res. Bull. A23 (Jan. 1987). While this $2 billion is admittedly a small fraction of the commercial paper and bankers acceptance market, which the same source puts at close to $400 billion per month, it does seem an overstatement to say that a $2 billion market has "largely" disappeared. It is also worth noting that a substantial number of existing financial practices assume the existence of negotiable instruments. Among the most important of these practices is the modern check clearing system. This is not to say that check clearing could be accomplished without negotiability, but only to say that the current system does depend on negotiability in important ways. See Rogers, The Irrelevance of Negotiable Instruments Concepts in the Law of the Check-Based Payment System, 65 Tex. L. Rev. 929 (1987). Thus, even if Professor Buckley's intuitions were correct and the "case for holder in due course superpriority" were "speculative" enough to justify the conclusion that if we were planning the world anew, we would not create this superpriority, the transition costs that would be incurred to abolish it from our world do act as an argument against this reform. Given our starting point, it is not enough to show that having no holder in due course superpriority would be a better state of affairs than having it. One must show that the improvement would be worth enough to pay the costs of the transition as well. 27. The highest valuing creditor will not (except for certain accounts receivable transactions) be the economists' familiar highest valuing user of a particular good. Since voluntary creditors are by definition individuals who have entered into transactions with the debtor, had they been the highest valuing user of the collateral, they would have bought it from the debtor. The qualification that transaction costs could have prevented such a sale is always present and possibly true. However, since the creditor and debtor were already engaged in a loan transaction in which the debtor had informed the creditor of the existence of the items standing as collateral, it is hard to imagine how transaction costs are likely to inhibit significant percentages of transactions that have reached this point of information exchange. 28. See Shuchman, Profit on Default: An Archival Study of Automobile Repos- session and Resaley 22 Stan. L. Rev. 20 passim (1969). 790 INDIANA LAW REVIEW [Vol. 22:3 This analysis is sufficient to explain why the UCC creates PMSIs arising by reason of law under Article 2. The sellers benefitted by these provsisions are likely to be in a better position than any other creditors to dispose of the goods to best advantage. While creditors as a class would prefer to have the goods, even at the price of an additional claimant in the seller's bankruptcy, the overall wealth of this particular unhappy community is most probably increased by creditors generally foregoing their claim to goods and having the original seller dispose of the goods for the best price possible. The economic justification for that imputed creditor preference is supported by a story made familiar by agency costs analysis. ^^ The implicit efficiency account concerning sellers becomes less com- pelling the more the seller's goods are commingled with the debtor's general assets. There is still reason to beHeve that the seller will have advantages over general creditors in disposing of the goods, but that advantage is not so probable if the competitor for the goods is a secured creditor who has chosen to create an interest in the same assets. Here a priori economic analysis has little to say, other than that the outcome depends on the relative advantages of the two secured creditors in disposing of the goods. The inconclusiveness of the economic analysis does, however, suggest why Section 2-702 has remained a major subject for debate. ^^ Under Article 9 of the UCC, any secured creditor can achieve superpriority over secured creditors by following the procedures set forth in Sections 9-312(3) or 9-312(4). The creditor who obtains superpriority under either of these subsections is likely to be a creditor who can anticipate that her valuation of the collateral will exceed that which would be given by the secured party of record. The statutory arrangements permit the debtor to take advantage of the higher valuing creditor's opinion. The superpriority provisions permit this transaction without disturbing the pre-existing credit relationship.^^ 29. The standard agency cost story applies here. The seller, so long as he captures the entire profit on resale, will have an incentive to make the investment that yields the maximum price. To the extent the seller shares those profits with others, his incentive to make investment necessary to obtain the maximum price will be correspondingly reduced. See generally, A. Barnea, R. Haugen, L. Senbet, Agency Problems and Financial Contracting (1985), and especially Chapter III, The Nature of Agency Problems. 30. See J. White & R. Summers, Uniform Commercial Code, 1027-28 (2d. ed. 1980). 31. Why a second secured creditor would value his collateral more highly than the first secured creditor may be appreciated in the context of specific cases. One example is the case of Brown and Williamson Tobacco Corp. v. First Nat'l Bank of Blue Island, 504 F.2d 998 (7th Cir. 1974). The debtor was a cigarette wholesaler who bought and resold cigarettes made by most of the major tobacco companies. The branded cigarettes 1989] SECURITY INTERESTS 791 The plausible perceptions about relations of secured creditors to collateral that underlie the provisions in both Article 2 and Article 9 suggest strongly that in a regime of free contract virtually all debtors would insist on PMSI superpriority. To show that PMSI superpriority would be extremely common does not, however, mean that it ought to be a mandatory rule. To make that judgment. Professor Buckley's attack on PMSI superpriority requires an examination of the costs to those contracting parties who do not wish to have PMSI superpriority and compare that cost to the relative drafting burdens under regimes with either mandatory PMSI superpriority or free contract. The question remains whether anything sensible can be said regarding the question of who should bear the drafting^ burden—the PMSI lender wanting superpriority or the first-in-time secured party wanting superior rights in after-acquired property. C. Weighing the Costs of the Drafting Burden 1. Silence as Consent to What?—Mandatory PMSI superpriority, although wanted by most contracting parties, may not be wanted by everyone. Whether the benefits to the large number of people wanting the statutory contract do outweigh the unavoidable costs to those not wanting that contract is very much an empirical question. The data to answer it does not, and in all probability cannot, exist. There are, however, a few considerations which, in the absence of data, suggest that if the data did exist, the data would support arrangements similar to those now provided by the UCC. The question at issue is what happens to contracting parties if they fail to agree explicitly to create PMSI superiority. The default rule in a regime of free contract would create a result that most contracting parties would not want. The great majority of contracting parties in such a regime would have provided for PMSI superiority by contracts containing standard language. There will, however, always be a few people who, because they do not consult a lawyer or because they consult an incompetent lawyer, end up with contracts that do not contain the standard language. This circumstance creates what Soia Mentschikoff once called the *'poor schnook" problem. These people will have contracts that fail to express their intent. Some judges will view the failure to include the standard language as conclusively expressing the parties' intent to contract contrary to the norm. A sympathetic court, however, might well observe both the common occurrence of the language and the absence in the case at hand of the special circumstances that usually obviously had a special value after repossession to the original manufacturer that they did not have to the other manufacturers. 792 INDIANA LAW REVIEW [Vol. 22:3 explain for the absence of the language. The court could then to choose to impose as a matter of law the legal consequences of the usual language. This process (often observable when the default rule does not match the ordinary practice) creates a mandatory regime piecemeal. Since it is piecemeal, neither the default rule nor its opposite can be relied upon with anything approaching certainty.^^ Since Professor Buckley's free contract regime's default rule would be contrary to the desires of most contracting parties, it would be especially vulnerable to the '*poor schnook" problem. 2. Search Costs and the Drafting Burden.—Professor Buckley finds reduction in search costs to be a sufficient justification for mandatory rules within Article 9, especially those which require filing for perfection of security interests." Assuming that Professor Buckley is correct in his conclusion that the mandatory fiHng provisions of Article 9 meet his test for vaUdity, these same provisions, to some extent, justify mandatory PMSI superpriority. Debtors and secured creditors find PMSI super- priority simplifies features of the mandatory fiHng system. Professor Buckley's failure to connect the two parts of his argument weakens the case he makes against mandatory PMSI superpriority. No mandatory search system necessarily reduces search costs. The system will do so if it has identified the information creditors will want to know with some accuracy, and does not burden those creditors with unnecessary information. In addition, the system must not impose sub- stantial costs on those individuals who must provide the data. The UCC's notice fihng system is built on the assumption that creditors are interested in classes of potential collateral and that the creditor will view notice that at least some of a debtor's property of a certain class is potentially subject to a secured claim as useful information. This method of in- formation categorization makes dealing with a floating lien on all of the debtor's property, whether existing or after-acquired, somewhat dif- ficult.^'* The burden on a creditor desiring to create such a floating lien, however, is not very great. 32. See, e.g., A. Trollope, PmNEAS Redux, CH. LX *Two Days Before the Trial" Trollope introduces the character of Chaffanbrass, the best criminal defense barrister in England. In this passage, he is in conversation with another attorney. Chaffanbrass says, "Caveat emptor is the only motto going, and the worst proverb that ever came from dishonest stoney-hearted Rome. With such a motto as that to guide us no man dare trust his brother. Caveat lex—and let the man who cheats cheat at his peril." "You'd give the law a great deal to do." "Much less than at present. What does your Caveat emptor come to? That every seller tries to pick the eyes out of the head of the purchaser. Sooner or later the law must interfere, and Caveat emptor falls to the ground. 33. See Buckley, supra note 2, at 1453-56. 34. See, e.g., cases noted in J. White & R. Summers, Uniform Commercial Code 1989] SECURITY INTERESTS 793 If the assumption concerning what creditors want to learn from a search is correct, then the notice provisions in the UCC succeed in reducing search costs. A quick look at a financing statement informs interested creditors whether further investigation will be needed. If the system is efficient, the answer will often be no. Because the UCC encourages thinking about collateral in broad categories, it also has the effect of encouraging both debtors and secured parties to create claims of security interests that go beyond what truly concerns the secured creditor. A secured creditor interested in only one part of a debtor's inventory will, by reason of U.C.C. Section 9-402, give notice to the world that she has an interest in the debtor's inventory. While it is true that the security agreement can limit the creditor's claim, the legal chmate in existence at the time the UCC was written, and which still exists, would encourage a cautious creditor to make expansive claims in the security agreement as well. The UCC does not give creditors and debtors any substantial motive for drafting security agreements narrowly. Since security agreements may be amended, no second-in-time secured party would comfortably rely on a security agreement which makes a narrow claim to part of a debtor's inventory where the financing statement asserts the first-in-time creditor has rights to "inventory." A sensible second-in-time secured creditor without PMSI superpriority would want to communicate with the first-in-time secured creditor to enter into an express agreement limiting the first creditor's claim. ^^ To the extent that efficient creditor search strategies encourage broad descriptions of the claims to collateral, both in the security agreements and in the financing statements, these efficient strategies will result in apparently overreaching claims by the first-in-time secured creditor. Man- datory PMSI superpriority counterbalances this type of overreaching drafting. If there were no mandatory PMSIs, then first-in-time secured lenders and their debtors would have substantial reasons to write much more precise descriptions, both in filings and in security agreements. More precise descriptions in security agreements and financing statements would be especially needed for inventory security arrangements. Because 910 n.50 (2d. ed. 1980). The point is not that these cases are correct, but simply that the existence of these cases mean the risk to secured parties with overbroad descriptions in their security agreements is real. 35. Under Professor Buckley's alternative system, any second-in-time secured cred- itor relying on an arrangement between the debtor and a first-in-time secured creditor which prevents the first-in-time secured creditor from claiming superpriority in after- acquired property would also have similar fears concerning amendment of that contract. Any cautious PMSI lender claiming superpriority under that alternative system would want to contract with the first-in-time secured party. Thus, under that system, a transaction that the UCC accomplishes with one contract would require two contracts. 794 INDIANA LAW REVIEW [Vol. 22:3 debtors and creditors often expect that the original seller will replace the original inventory as it turns over, these arrangements are likely to require after-acquired property clauses. When, however, the debtor has multiple sources for his inventory, (i.e., where the debtor is a retailer selling a broad range of products each supplied by a different financing manufacturer) the need for the debtor's business flexibility combined with the continuing creditors' need for after-acquired protection against competing secured parties will necessarily lead debtors and creditors to draft security agreements with limited and precise descriptions of the collateral pledged to each creditor. These more precise descriptions would, in a perverse way, have the effect of increasing the search costs of any second-in-time secured creditor who wished to enter into what today would be a mandatory PMSI superpriority transaction. Unless the first-in-time secured creditor had, by contract, permitted the creation of PMSI superpriority, the second- in-time creditor would have to determine the precise scope of the first- in-time secured party's claim to the collateral. That investigation alone would have some cost, as well as always creating the residual risk of error by the second-in-time secured creditor. Nor would these more limited descriptions of collateral necessarily be helpful to the first-in- time secured creditor. Since the debtor now would have an incentive to insist on narrowly drafted grants of security interests, the first-in-time secured creditor would have an increased risk that after default, property she thought was subject to the security interest would be determined by a court not to be.^^ Compared to the UCC, Professor Buckley's alternative regime carries with it a high probability of greater search costs for second-in-time secured lenders and, at least for inventory first-in-time secured lenders, a higher probability of risk of catastrophic losses due to hostile court interpretations of Hmiting language. To believe that Professor Buckley's alternative regime would be preferable to the UCC's structure, one has to believe that the sorts of costs that this regime would create for both first and second-in-time secured parties would be less than the value to debtors of having the ability to create, with certainty, rights of first- in-time secured parties to after-acquired property. In the absence of data, no belief can be unequivocably declared unreasonable. In light of the risks that one can identify in Professor Buckley's alternative regime, and in hght of the difficulty in identifying the circumstances in which 36. Considering the frequency with which this rather unpleasant surprise occurs under the UCC, where the debtor does not have the same motive to create narrow grants of security interests, this risk of finding no security for what had been made as a secured loan does appear to be more than academic. 1989] SECURITY INTERESTS 795 debtors would want to contract with first-in-time secured creditors without at the same time reserving the right to create PMSI superpriority, it does appear reasonable to conclude that the burden of proof still remains on Professor Buckley. D. The Risk of Debtor Dishonesty Professor Buckley considers and rejects the argument that, under existing law, debtors who wish to offer first-in-time secured creditors the benefit of after-acquired property clauses could do so by contract. He notes that even though *'an after-acquired property loan agreement might declare the grant of a subsequent PMSI interest to be an event of default, once made and registered, the PMSI interest is effective as against the first lender. "^^ Debtor dishonesty also creates troubling issues for his proposed alternative to existing law. In his preferred regime of free contract, the risk of debtor dishonesty falls entirely on the second-in-time PMSI lender. A dishonest debtor could fraudulently create a security agreement in which the first-in-time secured creditor appears to give the debtor rights to create PMSI su- perpriority. Public recording by the first-in-time secured creditor of notice that the debtor had rights to create PMSI superpriority would reduce opportunities for this sort of debtor misbehavior, but not completely eliminate it. A dishonest debtor could still enter into a recorded security agreement forbidding PMSI superpriority and then forge a document amending the recorded instrument to permit him to enter into PMSI superpriority credit arrangements. Under existing law, first-in-time secured creditors know they cannot totally rely on having a monopoly on after-acquired property. They can, however, expect to have both rights to after-acquired property and substantial protection against debtor dishonesty. This protection may be found in the structure of the transaction. Since UCC Section 9-312(3) requires a would-be PMSI lender to give notice to the prior secured party in order to obtain PMSI superpriority, the first-in-time secured creditor can use that notification to force a default and immediate repayment, having taken the (reasonable under all circumstances) pre- caution of not having the debt exceed the amount of the security. With respect to equipment and other tangible non-inventory security, a first- in-time secured creditor fearful of debtor dishonesty can arrange the payment schedule so that the amount outstanding is at all times equal to no more than a realistic depreciation schedule for the item standing as security. ^^ 37. Buckley, supra note 2, at 1465 n.l63. 38. Any secured creditor, whether under the existing or any proposed alternative regime, would of course have to monitor to be sure that the original collateral had not been disposed of in untraceable ways. 796 INDIANA LAW REVIEW [Vol. 22:3 Under the free contract regime, the second-in-time secured creditor who has been tricked into beUeving she has superpriority rights has no obvious means of protecting herself in the structure of her transaction. Instead, the second-in-time secured creditor must find her protection by entering into direct communications with the first-in-time secured creditor because no second-in-time creditor can safely rely on whatever documents either the debtor or public records produce. The second-in-time secured creditor will have to ask the first-in-time secured creditor to describe her rights to the debtor's after-acquired property. In every case where the UCC requires one communication, there would be two communi- cations under Professor Buckley's alternative system. In addition, second-in-time secured creditors will have to enter into two-way communications under circumstances where the UCC requires none. This communication would confirm that the first-in-time secured party has no rights and does not intend to have rights. Even where first-in-time secured parties intend to have rights in after-acquired prop- erty, debtors will likely respond by insisting on narrowly drawn security agreements. In these cases, cautious second-in-time secured parties must be sure that the first-in-time secured party's claim is subject to the limits stated in the documents possessed by the debtor. First-in-time secured creditors are not likely to admit to limited claims without involving lawyers in the communication process, which of course increases the costs of communication. When the concern is debtor dishonesty, the question becomes whether the costs to the second-in-time secured creditors whose debtors want PMSI superpriority under a free contract regime will exceed the costs under existing statutes to those first-in-time secured creditors who want after-acquired property superpriority. Again, in the absence of data no certain conclusions are possible. It is possible, however, to identify the interests which would have to be compared. Under Professor Buckley's alternative system, second-in-time secured lenders would have to invest to acquire information sufficient to avoid the total disaster which would result if a dishonest debtor were wrongfully to convince a creditor that he had the power to create PMSI superpriority. Under the UCC, sus- picious first-in-time secured creditors will incur the costs of restructuring the transaction to avoid being vulnerable to debtor dishonesty. Since these steps in restructuring the transaction have the result of keeping the value of the security below that of the debt, it is likely that first- in-time secured parties will take these steps even without the fear of debtor dishonesty. Even when first-in-time creditors would not take these sorts of steps, it is reasonable to believe that self-help will be less costly to secured creditors than would be the cost of acquiring the information that insures against disaster. If this reasonable belief is true, then the risks created by debtor dishonesty are less burdensome under the UCC then they would be under the regime of free contract. 1989] SECURITY INTERESTS 797 IV. Conclusion Professor Buckley offers a seductively attractive proposition to crit- icize the UCC's rules concerning PMSI superpriority. When he says that mandatory rules need special justification, he sounds as if he favors freedom of contract over regulation. What he makes us forget for a while is that any legal order, by providing for an outcome in the absence of agreement to the contrary, places burdens on contracting parties. All Professor Buckley would have us do is substitute one rule for another. The rule he proposes has the advantage that if the contracting parties do not Hke it, they can freely contract around it. His alternative rule, however, has two principal disadvantages. Nearly everyone would choose to contract around it, and no one who has contracted around it could be certain that if the debtor does breach his contract, the law would be able to provide an adequate remedy to the second-in-time secured party, the party presumably benefited by a contract term creating PMSI superpriority. The case for the efficiency of Professor Buckley's alternative system is not especially plausible. While he is correct to suggest that the UCC does not take into account the negative side to what its drafters saw as transaction cost-reducing mechanisms, a further examination of its provisions suggests that there is reason to believe that these costs are not great. There is even stronger reason to believe that the very few secured creditors who are disadvantaged by mandatory PMSI super- priority would find these costs less than the costs imposed on all secured creditors contracting in the absence of these statutory rules. So long as Kaldor-Hicks criteria provide the basis for a judgment about efficiency, the numbers of persons benefited by the present arrangements, the degree of their benefit, the near-absence of anyone wanting an alternative regime combined with the costs to persons injured by breach of contracts needed in Professor Buckley's proposed alternative regime, all suggest that PMSI superpriority under the UCC is preferable to this proposed alternative.