Friday, December 18, 2015
Well-Reasoned "True Sale" Opinion from Middle District of Pennsylvania Bankruptcy Court
In the December 2015 ABI Journal, I read an article by James Gadsden discussing the recent decision of United States Bankruptcy Judge Mary France in the Middle District of Pennsylvania, In re Dryden Advisory Services LLC, 534 B.R. 612 (Bankr. M.D. Pa. 2015). upholding a factoring agreement governed by New York law against an argument by the debtor-in-possession that the arrangement was a disguised financing arrangement such that the factored receivables were property of the estate. This having been an area that I often had to grapple with, in the sense of reviewing, negotiating and signing off on “true sale” opinions to support the securitization practice, and there being a dearth of modern opinions addressing the “true sale” question, I read the article with interest. I had met Judge France on a case in Harrisburg when she was in charge of the local office of the U.S. Trustee for the Region, and she had impressed me as having greater than customary business sense and common sense for one in that position (would she had been in charge of the Wilmington office instead), so my interest in the opinion was enhanced because she wrote it. This was a difficult case, and the result is debatable, but I think she analyzed it with precision and sophistication.
The debtor was in the business of pursuing tax refunds and other reductions for businesses, and was paid on commission. Cash flow was lumpy and frequently sluggish. Among its liquidity strategies was a factoring arrangement governed by New York law. The principal relevant terms of that agreement were:
* Factor was under no obligation to factor any particular account, but had discretion to accept and reject the ones Debtor proposed.
* Factor took an initial 3.5% discount on the face amount of each invoice. If the account remained outstanding after 30 days, Factor applied an additional 1.75% of face discount. Factor repeated that discount every 15 days thereafter until collection.
* To cover the discounts and other risks, Factor only advanced 75% of the amount of the invoice. The remaining 25% balance served as a “holdback” of the purchase price. If the account debtor did not pay in full, Factor kept the holdback. If payment was made in full, Factor rebated the holdback to the Debtor, after deducting therefrom whatever discounts and other items applied. Factor was, however, also entitled to retain from such remittances any amount needed to make itself whole on other purchased receivables that might be in default. The 25% level of recourse, coupled with the ability to cross-collateralize receivables, is unusually aggressive, compared to what we would have agreed to give “true sale” opinions on. I will discuss the implications later in this post.
* Last, and most critically for the decision, the Factoring Agreement provided that Factor assumed the risk of non-payment on purchased accounts only if non-payment was “due to the occurrence of an account debtor’s financial inability to pay, an `Insolvency Event.'” Notwithstanding this provision, Factor also had the right to put back to the account seller any invoice that was more than 89 days old.
After reciting authorities that provide an overview of the “true sale / disguised financing” issue, Judge France cites relatively modern case law from SDNY for the proposition that, “To constitute a bona fide factoring agreement under New York law, the factor need only assume the risk that the seller’s account debtor will be unable to pay.” In fact, every quotation she supplies on this point includes “only” or “merely,” making the point very clearly that the analysis is a fairly straightforward one. “[A]ll other risks associated with the sale of the accounts receivable remain with the client (e.g., commercial disputes …).”
After general observations that the language of the agreement is not dispositive, and courts look “beyond labels and into the details of the transaction”, Judge France’s analysis begins – oddly, I thought -- by noting that the agreement called for the Debtor to hold payments it received in trust in the exact form received and to forward them immediately to Factor. What troubles me about this observation is not just that its exclusive focus on the language of the agreement seems at odds with the immediately preceding proposition that the language of the agreement should be de-emphasized, but also, in the factual recitals, the Judge had recited at least one instance in which Debtor received payment of a factored receivable directly and initially paid over only the amount advanced on a given receivable, and Factor had to follow up to receive the balance. That seems to undercut the significance attached to the language of the agreement. The opinion obliquely takes up the topic of deviating from the language of the agreement, not as something directly bearing on the ultimate issue, but as a subsidiary question of whether the parties’ conduct had effectively amended the terms of the agreement; pointing to merger clauses and the usual boilerplate, the Judge concludes it hadn’t. I think this – which may have been how the debtor’s lawyer framed it – is a misguided perspective. The right focus is on how the property at issue was handled, as the initial lines of the Judge’s analysis state. It is irrelevant whether the agreement was or wasn't amended by the parties’ conduct; the conduct itself is what matters.
Further, it is unclear from the opinion, which recites some confusion among the litigants about how many receivables were at issue, whether there had been receivables paid to Debtor that, at the petition date, Debtor had failed to pay over to Factor. It may be that the confusion resulted from the account debtors paying the Factor directly but some clarity on the details might provide more insight.
Judge France goes on to state that:
“The ability of a buyer to demand that it receive payment directly from account debtors supports the finding that the transaction is a sale. Here, § 4.4 of the Amended Factoring Agreement gives Durham that right. “Durham may notify any Customer [i.e., account debtor] to make payments directly to Durham for any Account.” Durham Ex. 3 §4.4. After payment of several invoices was delayed, Durham exercised this right and demanded payment directly from Dryden’s account debtors. Had Durham exercised this right at the inception of the agreement it would have been abundantly clear that the transfer of the accounts was a sale. Durham may have preferred not to exercise this right initially to avoid disrupting the business relationship between Dryden and its clients, but in any event, it was entitled to exercise that right at any time under the terms of the Amended Factoring Agreement.”
Here too, I regretfully submit, the Judge over-emphasizes this provision. The power to take over collection is not unique to factoring: every “plain vanilla” security agreement made by a borrower in favor of a lender concerning accounts receivable contains language to this effect. It should therefore have been given no weight here.
The Judge also considers arguments that the pricing formula and the fact that the Debtor was responsible to “service” collection of the factored accounts support characterization of the arrangement as a financing and not a sale. Correctly, I think, the Judge rejects those arguments as well. Servicing by the account seller is a garden-variety feature of all securitization and, absent some abnormal or especially pertinent evidence it affected the main issue of recourse, it should be given no weight, as the Judge concluded; else there would never be a successful securitization. The provision for additional time-based charges, which definitely smack of a financing, concern me more, but, in and of themselves, they don’t tip the balance. They could be rationalized as just a greedy Factor looking for arbitrary excuses to ratchet up the income it’s going to earn; but, more importantly, it wouldn’t take a lot to rewrite the fees so that all 90 days’ worth were charged upfront. Instead, the Factor established incentive compensation for the debtor as servicer that just happened to mirror the timing and amount of the second-stage fees, so I think here, too, the Judge reached the right result. Those fees don’t affect the issue of recourse enough to drive a different result.
So, turning to that, here is what the Judge has to say about the extent of recourse:
“Courts have held that the most important single factor when determining whether a transaction is a true sale is the buyer’s right to recourse against the seller. One of the core attributes of owning a receivable is the risk that it will not be paid. If the buyer “sells” the receivable, but retains the risk of non-payment, it is more likely that the transaction will be recharacterized as a loan. An agreement “without recourse” means that the purchaser/factor agreed to assume the full risk of collecting the money owed to the seller, whereas an agreement “with recourse” means that the seller retains the risk of collection.” Filler v. Hanvit Bank, 339 F. Supp. 2d 553, 556 (S.D.N.Y. 2004), aff’d, 156 F. App’x 413 (2d Cir. 2005). Generally, if there is a full right of recourse against the seller, this weighs in favor of the existence of a loan because there is no transfer of risk. Recourse can take many forms including an obligation to repurchase accounts, a guaranty of the collectibility of accounts, or a reserve which is released when the receivables are paid. See Aicher & Fellerhoff, supra at 186.
“The Amended Factoring Agreement provided that Durham accepted the risk of “non-payment on Purchased Accounts, so long as the cause of non-payment is solely due to the occurrence of an account debtor’s financial inability to pay, an “Insolvency Event.” Durham Ex. 3 §4.10. As to this discrete event, Durham had no recourse against Dryden. The agreement does, however, specify some events which would afford Durham recourse for non-payment. For example, Dryden agreed to “accept back (repurchase) from Durham any Purchased Account subject to a dispute between Customer and Client of any kind whatsoever.” Id. at § 4.11. This included Durham’s right to require Dryden to repurchase disputed accounts, all Purchased Accounts if there was an event of default, and accounts unpaid after ninety days if an insolvency event had not previously occurred. Id. at §6.4.1. While the foregoing provisions limit Durham’s risk and provide some forms of recourse, they are insufficient to support recharacterization of the transaction as a loan.
“Even the existence of a right of full recourse is not dispositive. Thus, for example, “the presence of recourse in a sale agreement without more will not automatically convert a sale into a security interest.” Major’s Furniture Mart, Inc., 602 F.2d at 544. “The question for the court then is whether the nature of the recourse, and the true nature of the transaction, are such that the legal rights and economic consequences of the agreement bear a greater similarity to a financing transaction or to a sale.” Id. Put somewhat differently, if a seller conveys its entire interest in a receivable, the transfer is a true sale, even if the seller has a recourse obligation. See generally Harris & Mooney, supra (proposing that the more critical factor is whether the seller retains a significant interest in the property, not whether the seller has a recourse obligation). Here, Dryden transferred the full economic interest in the Purchased Accounts to Durham. Further, Dryden did not have a full recourse obligation, although it is misleading to characterize the transaction as “nonrecourse” when the agreement included a hold back provision (the “Reserve” in ¶ 4.9) and Durham could require Dryden to repurchase accounts “on demand” as set forth in ¶ 6.4.”
This is the correct framework for analysis and the only issue for debate is the weight to attach to the recourse provisions. I find this a much closer call than the Judge. I do agree with her conclusion regarding the insignificance of the chargebacks for disputes. That is a standard provision and has little to do with the issue of who bears the credit risk of the account debtor, which doesn’t arise unless the account debtor is legally obligated in the first place. But, the other provisions she cites are much harder calls. The ability to put back an account merely for being 90 days outstanding is anomalous and in the absence of a legal dispute over the obligation, difficult to square with the proposition that the factor has taken on the account debtor’s credit risk.
Additionally, 25% recourse is at least double, and in some cases triple, anything I ever saw in a securitization. Now, granted, the companies I was working with were ones for whom securitization was an option, a way to shave some basis points off the cost of financing their working capital, not, as was the case here, a last resort for the Debtor to stay afloat. But, that said, isn’t that evidence of a financing, that the amount of recourse demanded reflected the seller’s creditworthiness, not the account debtors’ creditworthiness? In our practice, whether we were giving an opinion or advising on the strength of a bankruptcy-remote structure, it was a cardinal point that the amount of recourse either had to be explicitly tied to the creditworthiness of the account debtor(s), or, more commonly, where the deal was a securitization program that would operate for several years, had to reasonably resemble the historical loss experience of the debtor on similar accounts. And, as a lesser-included point, the Factor's ability in Dryden to apply a rebate owed the Debtor on one account to a default under another is certainly not helpful to the proposition that the factor had acquired the credit risk of the account debtors, although not in and of itself fatal.
In expressing these doubts, I do not go so far as to say the decision is wrong, for a couple of reasons.
First, in the background here, I note, although I left it out of my summary of the facts, the opinion mentions that the Factor was recommended to the Debtor by the Small Business Administration and that could have had at least an unconscious effect on the Judge’s approach; she may not have wanted to resolve a close issue in a manner that might disrupt small-business financing in general or any SBA practices in workout situations. While not analytically satisfying, the impact on real-world financing practices is and should be a concern for judges at all levels in the judiciary, because bankruptcy is just a small part of a larger body of public policies.
Second, as I have suggested in passing a couple of times, in contrast with opinion-giving, where one can only opine on the terms of agreements as supplemented by assumptions about compliance therewith, the resolution of a litigation over “true sale” should be based on actual facts and conduct at least as much as the bare bones of the agreement. Here, while there was some evidence of deviation from a perfectly pristine transfer of the accounts to the Factor, it wasn’t particularly material; the Factor jumped on top of the issue right away and implemented strict compliance with the procedures designed to conform to a purchase relationship. It is hard on the record recited in the opinion to find conduct consistent with a lender-borrower relationship. Certain provisions of the agreement, such as the size of the holdback and the right to put back accounts more than 90 days old deviate materially from what I consider to be safe "true sale" practice. But, did they ever come into play as an economic matter? To me, that is the critical question for adjudication, not the words on a page. Did any invoice go past 90 days and, if so, did the Factor put that receivable back, or did it continue to hold the credit risk, consistent with a "true sale"? Did the Factor ever dip into recourse to cover a payment default, or just for fees? If the Debtor couldn't show an actual event in which the Factor shifted the loss upon default to the Debtor, it is hard for me to say this wasn't a "true sale" in fact.
Finally, and most importantly, I wasn’t there at the hearing and didn’t see the testimony or hear the arguments of counsel. The Judge’s opinion reduces her analysis to writing but doesn’t capture the full record of the litigation before her. It may well have been that the Debtor just didn’t make the case well enough to win. I believe, by the way, that the Debtor had the ultimate burden of persuasion under 363(p) as it was the one asserting the interest in the factored receivables, for purposes of using the proceeds thereof as cash collateral. Ultimately, from what I see in the opinion, I would have been pretty undecided about whom to rule in favor of here, and the burden of proof allocation might well have been the dispositive factor on this record, had I been the judge.
Overall, I think the Judge did a very commendable job on a highly sophisticated issue, constructing the right framework for analysis weighing of the factors, perhaps a little glibly but certainly defensibly, and arriving at an outcome that, considering the burden of persuasion, is probably the right bottom-line result.