On the first day of its bankruptcy, Kmart sought permission to pay immediately, and in full, the pre-petition claims of all "critical vendors." (Technically there are 38 debtors: Kmart Corporation plus 37 of its affiliates and subsidiaries. We call them all Kmart.) The theory behind the request is that some suppliers may be unwilling to do business with a customer that is behind in payment, and, if it cannot obtain the merchandise that its own customers have come to expect, a firm such as Kmart may be unable to carry on, injuring all of its creditors. Full payment to critical vendors thus could in principle make even the disfavored creditors better off: they may not be paid in full, but they will receive a greater portion of their claims than they would if the critical vendors cut off supplies and the business shut down. Putting the proposition in this way implies, however, that the debtor must prove, and not just allege, two things: that, but for immediate full payment, vendors would cease dealing; and that the business will gain enough from continued transactions with the favored vendors to provide some residual benefit to the remaining, disfavored creditors, or at least leave them no worse off.

Bankruptcy Judge Sonderby entered a critical-vendors order just as Kmart proposed it, without notifying any disfavored creditors, without receiving any pertinent evidence (the record contains only some sketchy representations by counsel plus unhelpful testimony by Kmart's CEO, who could not speak for the vendors), and without making any finding of fact that the disfavored creditors would gain or come out even. The bankruptcy court's order declared that the relief Kmart requested — open-ended permission to pay any debt to any vendor it deemed "critical" in the exercise of unilateral discretion, provided that the vendor agreed to furnish goods on "customary trade terms" for the next two years — was "in the best interests of the Debtors, their estates and their creditors". The order did not explain why, nor did it contain any legal analysis, though it did cite 11 U.S.C. § 105(a). (The bankruptcy court issued two companion orders covering international vendors and liquor vendors. Analysis of all three orders is the same, so we do not mention these two further.)

Kmart used its authority to pay in full the pre-petition debts to 2,330 suppliers, which collectively received about $300 million. This came from the $2 billion in new credit (debtor-in-possession or DIP financing) that the bankruptcy judge authorized, granting the lenders super-priority in post-petition assets and revenues. See In re Qualitech Steel Corp., 276 F.3d 245 (7th Cir.2001). Another 2,000 or so vendors were not deemed "critical" and were not paid. They and 43,000 additional unsecured creditors eventually received about 10¢ on the dollar, mostly in stock of the reorganized Kmart. Capital Factors, Inc., appealed the critical-vendors order immediately after its entry on January 25, 2002. A little more than 14 months later, after all of the critical vendors had been paid and as Kmart's plan of reorganization was on the verge of approval, District Judge Grady reversed the order authorizing payment. 291 B.R. 818 (N.D.Ill.2003). He concluded that neither § 105(a) nor a "doctrine of necessity" supports the orders.

Appellants insist that, by the time Judge Grady acted, it was too late. Money had changed hands and, we are told, cannot be refunded. But why not? Reversing preferential transfers is an ordinary feature of bankruptcy practice, often continuing under a confirmed plan of reorganization. See Mellon Bank, N.A. v. Dick Corp., 351 F.3d 290 (7th Cir.2003). If the orders in question are invalid, then the critical vendors have received preferences that Kmart is entitled to recoup for the benefit of all creditors. Confirmation of a plan does not stop the administration of the estate, except to the extent that the plan itself so provides. Compare In re Hovis, 356 F.3d 820 (7th Cir.2004), with In re UNR Industries, Inc., 20 F.3d 766 (7th Cir.1994). Several provisions of the Code do forbid revision of transactions completed under judicial auspices. For example, the DIP financing order, issued contemporaneously with the critical-vendors order, is sheltered by 11 U.S.C. § 364(e): "The reversal or modification on appeal of an authorization under this section to obtain credit or incur debt, or of a grant under this section of a priority or a lien, does not affect the validity of any debt so incurred, or any priority or lien so granted, to an entity that extended such credit in good faith, whether or not such entity knew of the pendency of the appeal, unless such authorization and the incurring of such debt, or the granting of such priority or lien, were stayed pending appeal." Nothing comparable anywhere in the Code covers payments made to pre-existing, unsecured creditors, whether or not the debtor calls them "critical." Judges do not invent missing language.

Now it is true that we have recognized the existence of a longstanding doctrine, reflected in UNR Industries, that detrimental reliance comparable to the extension of new credit against a promise of security, or the purchase of assets in a foreclosure sale, may make it appropriate for judges to exercise such equitable discretion as they possess in order to protect those reliance interests. See also In re Envirodyne Industries, Inc., 29 F.3d 301, 304 (7th Cir.1994). Thus once action has been taken to distribute assets under a confirmed plan of reorganization, it would take some extraordinary event to turn back the clock. These appeals, however, do not question any distribution under Kmart's plan; to the contrary, the plan (which was confirmed after the district court's decision) provides that adversary proceedings will be filed to recover the preferences that the critical vendors have received. No one filed an appeal, which means that it is appellants in this court that now wage a collateral attack on the plan of reorganization.

Appellants say that we should recognize their reliance interests: after the order, they continued selling goods and services to Kmart (doing this was a condition of payment for pre-petition debts). Continued business relations may or may not be a form of reliance (that depends on whether the vendors otherwise would have stopped selling), but they are not detrimental reliance. The vendors have been paid in full for post-petition goods and services. If Kmart had become administratively insolvent, and unable to compensate the vendors for post-petition transactions, then it might make sense to permit vendors to retain payments under the critical-vendors order, at least to the extent of the post-petition deficiency. Because Kmart emerged as an operating business, however, no such question arises. The vendors have not established that any reliance interest — let alone any language in the Code — blocks future attempts to recover preferential transfers on account of prepetition debts.

Handleman Company, which received $49 million as a critical vendor, makes a different procedural objection: that the district court's order does not affect it because Capital Factors' notice of appeal did not name Handleman as an appellee. Handleman was not a "party" in the district court and, consistent with the due process clause of the fifth amendment, cannot be bound by the district judge's decision — or so it says. We permitted Handleman to intervene in this court. Thus it is a party today and will be bound by our decision, so it is hard to see why it matters whether the district judge's resolution would have had independent effect.

Notices of appeal in bankruptcy must name "all parties to the judgment, order, or decree appealed from". Fed. R. Bankr.P. 8001(a)(2). Handleman was not a "party" to the critical-vendors order; Kmart was the sole party at the time. Kmart filed an ex parte application that did not specify any particular creditor. It had notified only 65 creditors of its impending request, and none of these was among the 2,000 vendors to be left high and dry. The bankruptcy judge's order likewise did not identify any creditor that acquired rights, for no creditor acquired rights. All the order did was authorize Kmart to pay any vendor that Kmart in its discretion deemed "critical." The party that Capital Factors had to name thus was Kmart itself, and this it did. If the lack of personal notice about the proceedings before the district judge deprived Handleman of due process, then Kmart's application to the bankruptcy judge deprived about 47,000 unsecured creditors of due process! That would render the critical-vendors order void, and Handleman would be worse off — for then it would have to repay the money even if the order's entry otherwise would have been lawful. But there is no constitutional obligation to make every creditor a party to every contested matter in the bankruptcy. As a rule, a trustee or debtor in possession represents the interests of many stakeholders. Kmart vigorously represented the interests of Handleman and the other vendors Kmart deemed "critical".

Other creditors must look out for their own interests and intervene if need be — as Handleman could have done had it devoted to these proceedings the care that a $49 million stake warrants. Handleman will be a party, and receive all the notice that the Constitution requires, if Kmart initiates a preference-recovery action against it. As a party in this court, Handleman will not be allowed to contest matters resolved here; even the 2,327 critical vendors that are not parties in this court must accept the precedential effect of our decision. No rule of law requires personal notice to all entities that might be affected by the precedential (as opposed to the preclusive) force of an appellate decision. Today's opinion affects thousands of "critical vendors" and other unsecured creditors; decisions by the Supreme Court may affect millions of persons. Only those persons who will be formally bound by a decision are entitled to individual notice, and then only when practical (the lesson of many a class action, see Mirfasihi v. Fleet Mortgage Corp., 356 F.3d 781, 785-86 (7th Cir.2004).) So there was no flaw in the notice of appeal or the district judge's view that Kmart and Capital Factors were the only parties to the proceedings.

Thus we arrive at the merits. Section 105(a) allows a bankruptcy court to "issue any order, process, or judgment that is necessary or appropriate to carry out the provisions of" the Code. This does not create discretion to set aside the Code's rules about priority and distribution; the power conferred by § 105(a) is one to implement rather than override. See Norwest Bank Worthington v. Ahlers, 485 U.S. 197, 206, 108 S. Ct. 963, 99 L. Ed. 2d 169 (1988); In re Fesco Plastics Corp., 996 F.2d 152, 154 (7th Cir.1993). Cf. United States v. Noland, 517 U.S. 535, 542, 116 S. Ct. 1524, 134 L. Ed. 2d 748 (1996). Every circuit that has considered the question has held that this statute does not allow a bankruptcy judge to authorize full payment of any unsecured debt, unless all unsecured creditors in the class are paid in full. See In re Oxford Management Inc., 4 F.3d 1329 (5th Cir.1993); Official Committee of Equity Security Holders v. Mabey, 832 F.2d 299 (4th Cir.1987); In re B & W Enterprises, Inc., 713 F.2d 534 (9th Cir.1983). We agree with this view of § 105. "The fact that a [bankruptcy] proceeding is equitable does not give the judge a free-floating discretion to redistribute rights in accordance with his personal views of justice and fairness, however enlightened those views may be." In re Chicago, Milwaukee, St. Paul & Pacific R.R., 791 F.2d 524, 528 (7th Cir.1986).

A "doctrine of necessity" is just a fancy name for a power to depart from the Code. Although courts in the days before bankruptcy law was codified wielded power to reorder priorities and pay particular creditors in the name of "necessity" — see Miltenberger v. Logansport Ry., 106 U.S. 286, 1 S. Ct. 140, 27 L. Ed. 117 (1882); Fosdick v. Schall, 99 U.S. 235, 25 L. Ed. 339 (1878) — today it is the Code rather than the norms of nineteenth century railroad reorganizations that must prevail. Miltenberger and Fosdick predate the first general effort at codification, the Bankruptcy Act of 1898. Today the Bankruptcy Code of 1978 supplies the rules. Congress did not in terms scuttle old common-law doctrines, because it did not need to; the Act curtailed, and then the Code replaced, the entire apparatus. Answers to contemporary issues must be found within the Code (or legislative halls). Older doctrines may survive as glosses on ambiguous language enacted in 1978 or later, but not as freestanding entitlements to trump the text. See, e.g., Lamie v. United States Trustee, ___ U.S. ___, 124 S. Ct. 1023, 1029-31, 157 L. Ed. 2d 1024 (2004), United States v. Ron Pair Enterprises, Inc., 489 U.S. 235, 242-46, 109 S. Ct. 1026, 103 L. Ed. 2d 290 (1989); Bethea v. Robert J. Adams & Associates, 352 F.3d 1125, 1128-29 (7th Cir.2003). See also Noland (courts lack authority to subordinate creditors that judges, as opposed to legislators, believe should be lower in the hierarchy).

So does the Code contain any grant of authority for debtors to prefer some vendors over others? Many sections require equal treatment or specify the details of priority when assets are insufficient to satisfy all claims. E.g., 11 U.S.C. §§ 507, 1122(a), 1123(a)(4). Appellants rely on 11 U.S.C. §§ 363(b), 364(b), and 503 as sources of authority for unequal treatment. Section 364(b) reads: "The court, after notice and a hearing, may authorize the trustee to obtain unsecured credit or to incur unsecured debt other than under subsection (a) of this section, allowable under section 503(b)(1) of this title as an administrative expense." This authorizes the debtor to obtain credit (as Kmart did) but has nothing to say about how the money will be disbursed or about priorities among creditors. To the extent that In re Payless Cashways, Inc., 268 B.R. 543 (Bankr.W.D.Mo.2001), and similar decisions, hold otherwise, they are unpersuasive. Section 503, which deals with administrative expenses, likewise is irrelevant. Pre-filing debts are not administrative expenses; they are the antithesis of administrative expenses. Filing a petition for bankruptcy effectively creates two firms: the debts of the pre-filing entity may be written down so that the post-filing entity may reorganize and continue in business if it has a positive cash flow. See Boston & Maine Corp. v. Chicago Pacific Corp., 785 F.2d 562 (7th Cir.1986). Treating pre-filing debts as "administrative" claims against the post-filing entity would impair the ability of bankruptcy law to prevent old debts from sinking a viable firm.

That leaves § 363(b)(1): "The trustee [or debtor in possession], after notice and a hearing, may use, sell, or lease, other than in the ordinary course of business, property of the estate." This is more promising, for satisfaction of a pre-petition debt in order to keep "critical" supplies flowing is a use of property other than in the ordinary course of administering an estate in bankruptcy. Capital Factors insists that § 363(b)(1) should be limited to the commencement of capital projects, such as building a new plant, rather than payment of old debts — as paying vendors would be "in the ordinary course" but for the intervening bankruptcy petition. To read § 363(b)(1) broadly, Capital Factors observes, would be to allow a judge to rearrange priorities among creditors (which is what a critical-vendors order effectively does), even though the Supreme Court has cautioned against such a step. See United States v. Reorganized CF & I Fabricators of Utah, Inc., 518 U.S. 213, 116 S. Ct. 2106, 135 L. Ed. 2d 506 (1996); Noland, supra. Yet what these decisions principally say is that priorities do not change unless a statute supports that step; and if § 363(b)(1) is such a statute, then there is no insuperable problem. If the language is too open-ended, that is a problem for the legislature. Nonetheless, it is prudent to read, and use, § 363(b)(1) to do the least damage possible to priorities established by contract and by other parts of the Bankruptcy Code. We need not decide whether § 363(b)(1) could support payment of some pre-petition debts, because this order was unsound no matter how one reads § 363(b)(1).

The foundation of a critical-vendors order is the belief that vendors not paid for prior deliveries will refuse to make new ones. Without merchandise to sell, a retailer such as Kmart will fold. If paying the critical vendors would enable a successful reorganization and make even the disfavored creditors better off, then all creditors favor payment whether or not they are designated as "critical." This suggests a use of § 363(b)(1) similar to the theory underlying a plan crammed down the throats of an impaired class of creditors: if the impaired class does at least as well as it would have under a Chapter 7 liquidation, then it has no legitimate objection and cannot block the reorganization. See generally Bank of America v. 203 N. LaSalle St. Partners, 526 U.S. 434, 119 S. Ct. 1411, 143 L. Ed. 2d 607 (1999). For the premise to hold true, however, it is necessary to show not only that the disfavored creditors will be as well off with reorganization as with liquidation — a demonstration never attempted in this proceeding — but also that the supposedly critical vendors would have ceased deliveries if old debts were left unpaid while the litigation continued. If vendors will deliver against a promise of current payment, then a reorganization can be achieved, and all unsecured creditors will obtain its benefit, without preferring any of the unsecured creditors.

Some supposedly critical vendors will continue to do business with the debtor because they must. They may, for example, have long term contracts, and the automatic stay prevents these vendors from walking away as long as the debtor pays for new deliveries. See 11 U.S.C. § 362. Fleming Companies, which received the largest critical-vendors payment because it sold Kmart between $70 million and $100 million of groceries and related goods weekly, was one of these. No matter how much Fleming would have liked to dump Kmart, it had no right to do so. It was unnecessary to compensate Fleming for continuing to make deliveries that it was legally required to make. Nor was Fleming likely to walk away even if it had a legal right to do so. Each new delivery produced a profit; as long as Kmart continued to pay for new product, why would any vendor drop the account? That would be a self-inflicted wound. To abjure new profits because of old debts would be to commit the sunk-cost fallacy; well-managed businesses are unlikely to do this. Firms that disdain current profits because of old losses are unlikely to stay in business. They might as well burn money or drop it into the ocean. Again Fleming illustrates the point. When Kmart stopped buying its products after the contract expired, Fleming collapsed (Kmart had accounted for more than 50% of its business) and filed its own bankruptcy petition. Fleming was hardly likely to have quit selling of its own volition, only to expire the sooner.

Doubtless many suppliers fear the prospect of throwing good money after bad. It therefore may be vital to assure them that a debtor will pay for new deliveries on a current basis. Providing that assurance need not, however, entail payment for pre-petition transactions. Kmart could have paid cash or its equivalent. (Kmart's CEO told the bankruptcy judge that COD arrangements were not part of Kmart's business plan, as if a litigant's druthers could override the rights of third parties.) Cash on the barrelhead was not the most convenient way, however. Kmart secured a $2 billion line of credit when it entered bankruptcy. Some of that credit could have been used to assure vendors that payment would be forthcoming for all post-petition transactions. The easiest way to do that would have been to put some of the $2 billion behind a standby letter of credit on which the bankruptcy judge could authorize unpaid vendors to draw. That would not have changed the terms on which Kmart and any of its vendors did business; it just would have demonstrated the certainty of payment. If lenders are unwilling to issue such a letter of credit (or if they insist on a letter's short duration), that would be a compelling market signal that reorganization is a poor prospect and that the debtor should be liquidated post haste.

Yet the bankruptcy court did not explore the possibility of using a letter of credit to assure vendors of payment. The court did not find that any firm would have ceased doing business with Kmart if not paid for pre-petition deliveries, and the scant record would not have supported such a finding had one been made. The court did not find that discrimination among unsecured creditors was the only way to facilitate a reorganization. It did not find that the disfavored creditors were at least as well off as they would have been had the critical-vendors order not been entered. For all the millions at stake, this proceeding looks much like the Chapter 13 reorganization that produced In re Crawford, 324 F.3d 539 (7th Cir.2003). Crawford had wanted to classify his creditors in a way that would enable him to pay off those debts that would not be discharged, while stiffing the creditors whose debts were dischargeable. We replied that even though classification (and thus unequal treatment) is possible for Chapter 13 proceedings, see 11 U.S.C. § 1322(b), the step would be proper only when the record shows that the classification would produce some benefit for the disfavored creditors. Just so here. Even if § 362(b)(1) allows critical-vendors orders in principle, preferential payments to a class of creditors are proper only if the record shows the prospect of benefit to the other creditors. This record does not, so the critical-vendors order cannot stand.

AFFIRMED.