Pearson v. Component Technology Corp.

247 F.3d 471, 2001 WL 369784

• Court: Court of Appeals for the Third Circuit
• Decided: April 13, 2001
• Docket: 00-3056
• Status: Unknown

Opinion

OPINION OF THE COURT

BECKER, Chief Judge.

The Worker Adjustment and Retraining Notification Act of 1988 (the WARN Act), 29 U.S.C. § 2101 et seq., mandates that employers provide workers with 60 days’ notice (subject to certain exceptions not at issue in this appeal) prior to a plant closing or mass layoff, and allows various remedies for workers when closures are not preceded by the requisite notification. Because a plant closure often presages a corporation’s demise, leaving workers with no source of satisfaction from their employer, plaintiffs have frequently sought damages from affiliated corporations. In a parallel series of cases, plaintiffs with claims arising from non-WARN Act sources of law against debt-laden or bankrupt corporations have occasionally attempted to sue the corporations’ major secured lenders, on the theory that the lenders have exercised such control over the corporations that veil-piercing is appropriate. This case implicates both lines of precedent.

The question before us is whether the former employees of Component Technology (CompTech), a now-defunct company, have set forth sufficient evidence to create a genuine issue of material fact as to whether, under standards we must fashion for this Circuit, CompTech’s major secured lender, General Electric Capital Corporation (GECC), should incur WARN Act liability for CompTech’s unnoticed plant closure. This question, in turn, requires us to consider not only the prerequisites for pareni/subsidiary liability in the WARN Act context (as will be shown, that jurisprudence is apposite here), but also whether the prerequisites change in the context of lender/borrower relationships.

The Department of Labor (DOL) has issued a regulation setting forth relevant factors for courts to use when considering whether to impose WARN Act liability on a parent corporation. See 20 C.F.R. § 639.3(a)(2). These factors closely resemble, but do not precisely mirror, the “single employer” or “integrated enterprise” test, frequently utilized for similar purposes in labor and employment law. The Department has also issued a statement explaining its intention that jurisprudence under the WARN Act not deviate from “existing law” with regard to liability for affiliated corporations. See 54 Fed. Reg. 16,045 (Apr. 20, 1989). The tension between “existing law” and the regulatory factors has led to considerable confusion among courts as to the appropriate standards to apply for WARN Act veil-piercing. Compare United Paperworkers Int’l Union v. Alden Corrugated Container Corp., 901 F.Supp. 426, 436-39 (D.Mass.1995), with Wholesale & Retail Food Distrib. Local 63 v. Santa Fe Terminal Servs., Inc., 826 F.Supp. 326, 334-35 (C.D.Cal.1993). To further compound the problem, when it comes to lenders rather than parents in other areas of law, courts have been extremely reluctant to hold lenders liable for their borrowers’ actions; usually, some version of the “alter ego” or “instrumentality” test for liability is used, often with an especial vigor. These tests are far less hospitable to plaintiffs than labor law’s “integrated enterprise” test, and, apparently, than the Department of Labor factors. Thus, the law is presently unsettled as to the proper test for liability under the WARN Act, and as to the significance for WARN Act purposes of an affiliated corporation’s status as “lender” or “parent.”

In this case, GECC loaned large sums of money to CompTech, and CompTech fell into default. Exercising its rights under the loan agreements, GECC voted Comp-Tech’s stock and installed a new slate of directors and a new Chief Executive Officer, to whom title of the stock was transferred. For the next few years, GECC and CompTech maintained a close relationship as CompTech struggled to survive as a going concern; when CompTech finally was unable to turn a profit, GECC declined to provide further cash infusions. CompTech, unable to secure new financing, collapsed and shut down its operations without giving WARN Act notice. A class of former CompTech employees have now brought suit against GECC under the WARN Act.

We conclude that the appropriate test to employ under the WARN Act for affiliated corporate liabihty is the multi-factored test promulgated by the DOL. We believe that the DOL’s instruction that courts apply “existing law” to questions involving inter-corporate liabihty was not intended to undermine the force of its own regulation on the subject, but was instead intended to instruct courts that existing precedent applying other tests (such as the “integrated enterprise” test) may be useful and appropriate to resolve analogous questions arising under the WARN Act. We also observe that the regulation indicates that the hsted factors are not an exhaustive list, which we interpret as a reminder that the test is one of balancing, and that, as with any balancing test, a number of circumstances not specifically enumerated may be relevant.

We must also determine whether GECC’s initial status as a secured lender affects the test we choose to employ for WARN Act liabihty, and we must further decide whether, under the appropriate test, the District Court erred in concluding that plaintiffs had not put forth sufficient evidence to create a genuine issue of material fact as to GECC’s liability. We ultimately hold that because the lines separating “parents” from “lenders” are not often bright ones, the simpler approach is to apply the same test for liability regardless of the formal label the corporations have attached to their association. Our conclusion is strengthened by our recognition that deter mining liabihty by reference to whether a lender has behaved in a “typical” manner, as did the District Court in this case, carries with it the risk of unintentionally altering what is “typical,” as lenders structure their relationships with borrowers to respond to the practices that we ourselves have proclaimed “typical.” Therefore, we take a more functional approach to determining whether or not to “pierce the veil” under WARN by focusing on the nature and degree of control possessed by one corporation over another; in so doing, however, particular weight must be accorded, where applicable, to a lack of ownership interest between corporations.

Applying the DOL factors to the circumstances presented in this case, we hold that even if GECC did, in the course of its relationship with CompTeeh, technically become a “parent” corporation, its actions never reached the point where even a more conventional parent would become liable. Therefore, the District Court’s grant of summary judgment to GECC was proper, and we will affirm the judgment.

I. Facts and Procedural History

Component Technology, a Delaware corporation with headquarters in Erie, Pennsylvania, was a custom injection molder whose business was to manufacture plastic objects in accordance with corporate customers’ specifications, principally in the business machine and medical products markets. R&R Plastics (R&R) was, at that time, a wholly-owned subsidiary of CompTeeh.

In June 1989, CompTeeh was poised to enter into a profitable new arrangement with Kodak, whereby CompTeeh would manufacture plastic components for a revolutionary new photocopier. In anticipation of the capital expansion that the venture would require, CompTeeh sought and obtained a $25,000,000 loan from GECC. The loan was formally structured as a loan to the Chicago Plastics Products Corporation (Chicago Plastics), a holding company formed for the purpose of acquiring Comp-Tech. As security for its loan, GECC received pledge agreements for all of the stock in Chicago Plastics, CompTeeh, and R&R, including the right to vote the stock in the event of a default.

Shortly after Chicago Plastics purchased CompTech, the Kodak project was canceled due to “technical obsolescence.” CompTech’s business immediately faltered, resulting in default on the GECC loan in 1991. GECC exercised its rights under the pledge agreements and voted its stock to install new boards of directors of the three corporations. The new directors, in turn, chose new corporate officers. On July 10, 1991 GECC hired a consultant with experience in the plastics industry, Thomas Gaffney, to serve as Chief Executive Officer of Chicago Plastics, CompTech, and R&R.

At some point in late 1991 or early 1992, GECC Vice-President Jeanette Chen began to manage the CompTech account. In March, she drafted an internal credit memorandum outlining a proposed strategy to restructure CompTech’s loans. The memo explained that for GECC to “[m]ax-imize debt recovery” it would need to “hold[] investment for sufficient time period to implement merger/acquisition strategy, rebuild customer base, and return Company to profitability.” Pursuant to the recommendations contained in the memo, GECC wrote off $20,000,000 of CompTech’s debt and restructured the remaining debt as term loans of $8,500,000, a revolving line of credit of $8,500,000, and nonvoting preferred stock of $4,000,000.

GECC then renewed its agreement with Gaffney and arranged for a private foreclosure sale of CompTech’s stock from Chicago Plastics. The stock was sold to Component Technology Acquisition Company (Acquisition), a wholly-owned subsidiary of Component Technology Holdings Corporation (Holdings). Acquisition was then merged into CompTech. Sixty percent of the stock in Holdings was transferred to Gaffney; forty per cent was transferred to Richard Brooks, the newly-appointed president of CompTech. GECC retained a pledge on all of the CompTech common stock, and neither Gaffney nor Brooks paid any compensation for their stock. As part of the consulting agreement, GECC indemnified Gaffney for all liabilities (other than misconduct) arising out of his duties as CEO of the companies and as shareholder of Holdings.

As a final step in the restructuring outlined in the Chen memo, CompTech acquired a plastics company known as Accu-form and merged it with R&R. Thus, after these maneuvers, all of CompTech’s stock was owned by a single holding company, Holdings, which, in turn, was owned by Gaffney and Brooks. Gaffney, the CEO of CompTech, was under contract to GECC to run the company, and had been indemnified by GECC for liabilities arising both out of his activities as CEO and his activities as a shareholder.

The loan agreement, together with the $4,000,000 worth of nonvoting preferred stock held by GECC, provided GECC with a considerable amount of control over CompTech’s finances. The loan agreement and GECC’s stock ownership entitled GECC to receive certain scheduled payments and, in the event of two consecutive defaults or four defaults overall, to vote in new directors and assume control of CompTech until the deficiency was paid. No dividends could be paid on common stock until GECC had received its preferred payments. Without prior authorization from GECC, CompTech was forbidden to engage in stock reorganization, lending or borrowing, mergers or acquisitions, large-scale capital expenditures, or selling of assets encumbered by liens securing GECC loans. Further, GECC retained the right to approve any employee salaries in excess of $100,000 per year — a restriction apparently intended to enable GECC to monitor the hiring of key personnel. Finally, GECC received warrants on 75% of CompTech’s stock, 24% of which were to be gradually returned to Comp-Tech as the loans were repaid, although there is no indication in the record of when, if ever, the remaining 51% were to be relinquished. As it turned out, Comp-Tech was only able to pay a single dividend to GECC, in 1993, by using monies drawn from the revolving line of credit.

As CompTech continued to operate with the new loans in place, GECC exercised continuing oversight of its finances pursuant to the loan agreement, occasionally agreeing to waive penalties and extend further loans to the cash-strapped company. CompTech, in accordance with the agreement, sought approval of a number of its decisions, including executive compensation and benefits, and the sale of equipment. CompTech also provided GECC with updates concerning its financial condition. GECC’s approval was required when CompTech sought to create a Mexican subsidiary to service one of its customers, and in connection with the deal, GECC waived its security interest in relevant equipment, receiving in return a pledge on all of the stock owned by CompTech in the new venture.

In early 1994, CompTech sold R&R, and the proceeds of the sale were used to prepay GECC in accordance with the terms of the loan agreement. Later that year, Gaffney wrote to GECC outlining CompTech’s current status and its proposed projects, seeking approval for those proposals relating to the Accuform acquisition. The letter concluded by saying “We need to know what G.E. wants us to do. We are proceeding with items 2, 8 and 4, but we need to get confirmation from G.E. on each one of these items. Obviously, without G.E.’s help we cannot proceed to complete our plans. I am prepared to do whatever G.E. wants relative to Comp-Tech.”

The next month, GECC determined that as CEO of CompTech, Gaffney was too focused on “grandiose schemes” for the company and spent too little time on the day-to-day operations. According to the deposition of GECC Vice-President Ed Christie, it was for this reason GECC elected not to renew Gaffney’s contract and asked him to step down as CEO. Gaffney had recently fired Brooks as president and replaced him with Charles Villa; GECC chose to hire Villa as the new president and CEO. In order to transfer power from Gaffney to Villa, both Gaffney and Brooks sold their stock back to Holdings for a nominal price, and Holdings reissued the stock to Villa. Villa then pledged the shares to GECC, and granted GECC the unconditional power to transfer and assign the stock.

At some point after the plans for these transfers had been laid, but before the shift in ownership had actually taken place, Villa sent a letter to Chen detailing the current status of CompTech’s attempts to resuscitate its business and requesting GECC’s further support to finance its growth and upgrade its facilities. The letter concluded by stating that “G.E. needs to make decisions now on all the issues outlined in this memo. How you respond will then dictate what the management of COMPTECH will need to do to accomplish the task at hand. We can either move forward in an aggressive, systematic approach to take advantage of industry dynamics and market windows available now or we can remain stagnant and loose [sic] any competitive edge that may exist resulting in a slow death of a one-time INDUSTRY LEADER. The choice is yours to make.”

When the stock transfers finally were completed in late June 1994, a number of new rights were granted to GECC. At any time within the first six months, GECC could force Villa to sell his shares to GECC for the same nominal amount that he had paid (the Villa Call). Further, at any time — either before or after the first six months — GECC could force Villa to sell a portion of the stock for value, either to GECC or to a GECC designee (the Bring-Along Call). The purpose of the Villa Call was to enable GECC to “assess Villa’s abilities on a trial basis,” and, should he prove to be unsatisfactory, easily transfer the shares to a new president. The stated purpose of the Bring-Along Call was to allow GECC to sell CompTech if it so chose. Villa, as the stockholder, had the right to vote the shares so long as Comp-Tech stayed current with loan payments; however, it appears that CompTech was in default either at the time of the transfer or shortly thereafter, so that in fact, GECC at all times possessed the voting rights. GECC acknowledged in its internal memo-randa that, as a result of these options, it was potentially responsible for the company’s liabilities under the Employee Retirement Income Security Act of 1974 (ERISA), 29 U.S.C. § 1001 et seq. However, because by this time CompTech had frozen its benefits plans and could not create new ones without GECC approval, there was no risk that GECC would involuntarily become liable for payments under ERISA.

Simultaneous with the stock transfer, GECC requested that CompTech hire Stuart Benton as an industry consultant to “help assess the Company’s ongoing cash needs and determine an appropriate account strategy for GECC going forward.” Benton was also to function as a “backup” president should Villa fail to meet GECC expectations. Although Benton was formally under contract to CompTech, several CompTech employees testified in deposition that they understood him to be a GECC representative.

Benton observed the CompTech operation for the next few months, until the plant closed. Deposition testimony from CompTech’s upper-level managers demonstrates that although he occasionally made suggestions to the managers, Benton largely worked with Villa and did not direct or interfere with the duties of the other management personnel. None of the upper-level managers, including Villa, reported any attempts by GECC to control CompTech’s operations, through Benton or through anyone else. Other than the financial controls and stock interests described above, GECC had no formal authority to control CompTech’s business decisions, including its strategic planning, its relationships with clients, its marketing, its product selection, its product design, or its quality control.

By August 1994, CompTech was contemplating a plant closure, going so far as to contact its attorneys to inquire about its WARN Act obligations. In September 1994, CompTech informed GECC that it would require further cash outlays to continue operations through the end of the year. CompTech also sought financing from Citibank Venture Capital Limited, but Citibank would only provide financing if GECC continued to do so as well. GECC refused to provide further funds, either with or without Citibank’s participation, and, on September 30, Chen drafted an internal memorandum sketching out a proposal for an “orderly liquidation” of the company. Negotiations for Comp-Tech’s liquidation began in October 1994, and on October 14, 1994, employees were formally notified of the plant closing, effective immediately.

The former employees of CompTech filed this action against GECC in the District Court for the Western District of Pennsylvania on November 2, 1994. On December 17, 1999, after discovery, the District Court granted GECC’s motion for summary judgment on the ground that the plaintiffs had failed to demonstrate that GECC, as a lender, had “exhibited] such a high degree of control over the debtor corporation” so as to become an “employer” within the meaning of the WARN Act. Pearson v. Component Tech. Corp., 80 F.Supp.2d 510, 518 (W.D.Pa.1999). The court also rejected the plaintiffs’ attempts to argue for veil-piercing liability based on GECC’s status as a parent corporation of CompTech, reasoning that GECC was not a parent and that its actions were “consistent with its role as CompTech’s secured creditor.” Id. at 523.

On appeal, the plaintiffs argue that: (1) GECC was an employer within the meaning of the WARN Act as a matter of law by virtue of its ownership of CompTech stock options; and (2) GECC was a parent corporation of CompTech under circumstances justifying veil-piercing under WARN Act standards. The District Court had jurisdiction pursuant to 28 U.S.C. § 1331, and we have jurisdiction pursuant to 28 U.S.C. § 1291. We exercise plenary review over a District Court’s grant of summary judgment. See Coolspring Stone Supply, Inc. v. American States Life Ins. Co., 10 F.3d 144, 146 (3d Cir.1993). We set forth the familiar summary judgment standard in the margin. 1

II. Liability for Affiliated Corporations Under the WARN Act

A. Introduction

In the wake of numerous plant closings and mergers in the 1970s and 1980s, Congress passed the WARN Act. See Hotel Employees & Rest. Employees Int’l Union Local 54 v. Elsinore Shore Assocs., 173 F.3d 175, 182 (3d Cir.1999). The Act was intended to protect workers by requiring that companies with advance knowledge of an imminent closing provide notice to employees, so as to allow “workers and their families some transition time to adjust to the prospective loss of employment, to seek and obtain alternative jobs and, if necessary, to enter skill training or retraining that will allow these workers to successfully compete in the job market.” 20 C.F.R. § 639.1(a). Thus, the Act states that:

An employer shall not order a plant closing or mass layoff until the end of a 60 day period after the employer serves written notice of such an order ... to each representative of the affected employees as of the time of the notice or, if there is no such representative at that time, to each affected employee....

29 U.S.C. § 2102(a)(1).

The Act defines an employer as “any business enterprise” that employs 100 or more employees. Id. § 2101(a). Employers violating the Act are liable for backpay and back benefits. See id. § 2104(a). Thus, the question presented by this litigation is whether, under these facts, GECC was the plaintiffs’ employer. In order to make such a showing, the plaintiffs must establish GECC to be a single “business enterprise” with CompTech such that it is responsible for CompTeeh’s WARN Act obligations.

The question when affiliated corporations will be considered a single employer for WARN Act purposes tends to arise in two contexts: (1) when plaintiffs seek to impose liability for violations on affiliates of insolvent corporations, see, e.g., Local 397, Int’l Union of Electronic, Elec. Salaried Mach. & Furniture Workers v. Midwest Fasteners, Inc., 779 F.Supp. 788 (D.N.J.1992); and (2) when plaintiffs seek to establish that two or more affiliated corporations should be viewed as a single enterprise in order to meet the 100 employee WARN Act threshold, see, e.g., Watts v. Marco Holdings, L.P., No. 3:95CV88-B-A, 1997 WL 578783 (N.D.Miss. Aug. 8, 1997). The WARN Act itself does not address such situations, but the Department of Labor regulations issued under the Act provide that:

Under existing legal rules, independent contractors and subsidiaries which are wholly or partially owned by a parent company are treated as separate employers or as a part of the parent or contracting company depending upon the degree of their independence from the parent. Some of the factors to be considered in making this determination are (i) common ownership, (ii) common directors and/or officers, (iii) de facto exercise of control, (iv) unity of personnel policies emanating from a common source, and (v) the dependency of operations.

20 C.F.R. § 639.3(a)(2). The five factors will hereinafter be referred to as the “DOL factors.”

The Department of Labor’s “supplementary information” regarding its WARN Act regulations explains that:

The intent of the regulatory provision relating to independent contractors and subsidiaries is not to create a special definition of these terms for WARN purposes; the definition is intended only to summarize existing law that has developed under State Corporations laws and such statutes as the NLRA, the Fair Labor Standards Act (FLSA) and the Employee Retirement Income Security Act (ERISA). The Department does not believe that there is any reason to attempt to create new law in this area especially for WARN purposes when relevant concepts of State and federal law adequately cover the issue.

54 Fed. Reg. 16,045 (Apr. 20,1989).

The intersection of the regulatory factors and the supplementary information has created considerable confusion among courts searching for a single test to determine the status of affiliated corporations. See Cynthia Nance, Affiliated Corporation Liability Under the WARN Act, 52 Rutgers L. Rev. 495, 535-36 (2000) [hereinafter Nance, WARN Act] (describing contradictory holdings). The problem arises because the jurisprudence contains several tests for determining when two corporations compose a single entity depending on whether the cause of action accrues under state or federal law, as well as on the particular type of claim at issue. Further, the DOL factors do not precisely correspond to any of the established tests for such determinations. Courts examining affiliated corporations under the WARN Act have often applied two or more tests, purporting to “average” the results, usually without any systematic method for doing so. See, e.g., United Paperworkers Int’l Union v. Alden Corrugated Container Corp., 901 F.Supp. 426, 436-39 (D.Mass.1995) (conducting, inter alia, a state alter ego test, but ultimately jettisoning the results on the ground that federal liability standards should not turn on state protections for corporations).

A further complication comes from the fact that we have before us in this case not the traditional parent/subsidiary relation ship but a relationship that began as an arrangement between a secured lender and a borrower — a situation unaddressed by either the Act or the regulations. Thus, we must determine whether the “standard” WARN Act test — whatever test that might be — is even applicable under these circumstances. To that end, we will first briefly sketch some of the methods available for determining intercorpo-rate WARN Act liability, ultimately concluding that rather than simply choosing one of the “established” tests and importing it to the WARN Act context, the appropriate test is the one specifically delineated in the DOL regulation. We further conclude that the supplementary information provided by the Department of Labor was not intended to encourage courts to choose a different test, but was merely intended to clarify that courts may draw on concepts in existing precedent when interpreting and applying the DOL factors. We will then turn to the question whether the DOL factors should apply to situations involving lenders rather than parents, ultimately concluding that the factors should be the same for both.

B. Liability Between Parents and Subsidiaries

1. Traditional Veil-Piercing Theories

The corporate form was created to allow shareholders to invest without incurring personal liability for the acts of the corporation. These principles are equally applicable when the shareholder is, in fact, another corporation, and hence, mere ownership of a subsidiary does not justify the imposition of liability on the parent. See United States v. Bestfoods, 524 U.S. 51, 69, 118 S.Ct. 1876, 141 L.Ed.2d 43 (1998); American Bell Inc. v. Federation of Tel. Workers of Pa., 736 F.2d 879, 887 (3d Cir.1984). Nor will liability be imposed on the parent corporation merely because directors of the parent corporation also serve as directors of the subsidiary. See Bestfoods, 524 U.S. at 69, 118 S.Ct. 1876. However, under both state and federal common law, abuse of the corporate for m will allow courts to employ the “tool of equity” known as veil-piercing, i.e., disregard of the corporate entity to impose liability on the corporation’s shareholders. Publicker Indus., Inc. v. Roman Ceramics Corp., 603 F.2d 1065, 1069 (3d Cir.1979). Courts have held veil-piercing to be appropriate “when the court must prevent fraud, illegality, or injustice, or when recognition of the corporate entity would defeat public policy or shield someone from liability for a crime,” Zubik v. Zubik, 384 F.2d 267, 272 (3d Cir.1967), or when “the parent so dominated the subsidiary that it had no separ-rate existence,” New Jersey Dep’t of Envtl. Prot. v. Ventron Corp., 94 N.J. 473, 468 A.2d 150, 164 (1983).

The Third Circuit alter ego test is fairly typical of the genre. 2 It requires that the court look to the following factors: gross undercapitalization, failure to observe corporate formalities, nonpayment of dividends, insolvency of debtor corporation, siphoning of funds from the debtor corporation by the dominant stockholder, nonfunctioning of officers and directors, absence of corporate records, and whether the corporation is merely a facade for the operations of the dominant stockholder. See American Bell, 736 F.2d at 886. Other (similar) formulations are set forth in the margin. 3

The test, whether or not a particular version requires an element of fraudulent intent, see supra note 2, is demonstrably an inquiry into whether the debtor corporation is little more than a legal fiction. Such a burden is notoriously difficult for plaintiffs to meet. For instance, courts have refused to pierce the veil even when subsidiary corporations use the trade name of the parent, accept administrative support from the parent, and have a significant economic relationship with the parent. See, e.g., Jackson v. General Elec. Co., 514 P.2d 1170 (Alaska 1973). Thus, in order to succeed on an alter ego theory of liability, plaintiffs must essentially demonstrate that in all aspects of the business, the two corporations actually functioned as a single entity and should be treated as such. See RRX Indus., Inc. v. Lab-Con, Inc., 772 F.2d 543, 545 (9th Cir.1985) (veil-piercing is appropriate when “the personalities of the corporation and individual are no longer separate”); Akzona Inc. v. E.I. Du Pont De Nemours & Co., 607 F.Supp. 227, 237 (D.Del.1984) (a subsidiary is an alter ego or instrumentality of the parent when “the separate corporate identities ... are a fiction and ... the subsidiary is, in fact, being operated as a department of the parent”).

2. “Integrated Enterprise” Test

Veil-piercing doctrine has been criticized for employing the same formulations of the test across the different contexts in which plaintiffs seek to impose liability. See Phillip I. Blumberg, The Law of Corporate Groups: Substantive Law § 6.01, at 107-OS (1987); cf. William H. Lawrence, Lender Control Liability: An Analytical Model Illustrated with Applications to the Relational Theory of Secured Financing, 62 S. Cal. L. Rev. 1387, 1388 (1989) [hereinafter Lawrence, Lender Control Liability] (criticizing the use of similar “indicia of control” for lender liability cases regardless of context). It is often argued that because public policy varies from contract to tort to property, for example, veil-piercing standards should vary as well. See, e.g., Robert B. Thompson, Piercing the Corporate Veil: An Empirical Study, 76 Cornell L. Rev. 1036 (1991). These concerns have been partially addressed through the “integrated enterprise” test for the presence of a single employer, a sort of labor-specific veil-piercing test, first developed by the National Labor Relations Board.

Because the Board was concerned only with labor law and policy, it developed a test for corporate “sameness” that, likewise, concerned itself only with those aspects of corporations having a di rect relevance to labor relations. So, for example, the integrated enterprise test is not concerned with such traditional alter ego hallmarks as “nonpayment of dividends,” because such aspects of a corporation’s finances are not as directly related to management’s labor policy as are other aspects of corporate functioning. See Nance, WARN Act, supra, at 533. Rather, the test looks to four labor-related characteristics of affiliated corporations: interrelation of operations; common management; centralized control of labor relations; and common ownership or financial control. See, e.g., Radio & Television Broad. Techs. Local Union 1264 v. Broadcast Serv. of Mobile, 380 U.S. 255, 256, 85 S.Ct. 876, 13 L.Ed.2d 789 (1965) (per curiam). No single factor is dispositive; rather, single employer status under this test “ultimately depends on all the circumstances of the case.” NLRB v. Browning-Ferris Indus. of Pa., Inc., 691 F.2d 1117, 1122 (3d Cir.1982).

As originally designed, the integrated enterprise test was used by the National Labor Relations Board to determine whether two firms were sufficiently related to meet its jurisdictional minimum amount of business volume. See Stephen F. Befort, Labor Law and the Double-Breasted Employer: A Critique of the Single Employer and Alter Ego Doctrines and a Proposed Reformulation, 1987 Wis. L. Rev. 67, 75. Later, the Board came to use the same test to determine whether nominally separate firms constituted “neutral” entities in the context of secondary boycotts, and to determine whether an employer had impermissibly “double-breasted” operations so as to avoid the obligations of a collective bargaining agreement. See id. at 75-76. 4

Since its initial formulation, the test has been applied by courts in other employment contexts, including the Labor Management Relations Act, see International Bhd. of Teamsters Local 952 v. American Delivery Serv. Co., Inc., 50 F.3d 770 (9th Cir.1995); Title VII and the Age Discrimination in Employment Act, see Frank v. U.S. West, Inc., 3 F.3d 1357 (10th Cir.1993); the Americans with Disabilities Act, see EEOC v. Chemtech Int’l Corp., 890 F.Supp. 623 (S.D.Tex.1995); and the Fair Labor Standards Act, see Takacs v. Hahn Auto. Corp., No. C-3-95-404, 1999 WL 33117265 (S.D.Ohio Jan. 4, 1999). But see Papa v. Katy Indus., Inc., 166 F.3d 937, 940-43 (7th Cir.1999) (rejecting the integrated enterprise test in the context of antidiscrimination law). Department of Labor regulations have also adopted the integrated enterprise test for the Family Medical Leave Act. See 29 C.F.R. § 825.104(c)(2).

The integrated enterprise test, with its focus only on labor relations and its emphasis on economic realities as opposed to corporate formalities, see Phillip I. Blum-berg, The Law of Corporate Groups: Problems of Parent and Subsidiary Corporations Under Statutory Law of General Application § 13.03, at 398 (1989), is demonstrably easier on plaintiffs than traditional veil piercing. Ultimately, “the policy underlying the single employer doctrine is the fairness of imposing liability for labor infractions where two nominally independent entities do not act under an arm’s length relationship.” Murray v. Miner, 74 F.3d 402, 405 (2d Cir.1996).

3. Direct Liability

Although not often employed to hold parent corporations liable for the acts of subsidiaries in the absence of other hallmarks of overall integration of the two operations, it has long been acknowledged that parents may be “directly” liable for their subsidiaries’ actions when the “alleged wrong can seemingly be traced to the parent through the conduit of its own personnel and management,” and the parent has interfered with the subsidiary’s operations in a way that surpasses the control exercised by a parent as an incident of ownership. United States v. Bestfoods, 524 U.S. 51, 64, 118 S.Ct. 1876, 141 L.Ed.2d 43 (1998) (quoting William O. Douglas & Carrol M. Shanks, Insulation from Liability Through Subsidiary Corporations, 39 Yale L.J. 193, 207 (1929)). In such situations, the parent has not acted on its own (in which case there would be no need even to consider the subsidiary’s actions), nor has it acted in its capacity as owner of the subsidiary; rather, it has forced the subsidiary to take the complained-of action, in disregard of the subsidiary’s distinct legal personality. See Esmark, Inc. v. NLRB, 887 F.2d 739, 756-57 (7th Cir.1989). Thus, in the labor context, “direct” liability may attach if the parent has overridden the subsidiary’s ordinary decision-making process and ordered it to institute an unfair labor practice, or to create discriminatory hiring policies. See id. at 757. In this way, direct liability functions essentially as a kind of “transaction-specific” alter ego theory. Id. at 756.

Although direct liability is rarely used independently to hold parents liable for them subsidiary’s actions, it has often been used in conjunction with the “integrated enterprise” test for liability, particularly to satisfy the “control of labor” prong. For instance, the Ninth Circuit in UA Local 313 of the United Ass’n of Journeymen & Apprentices of the Plumbing & Pipefitting Industry of the United States & Canada v. Nor-Cal Plumbing, Inc., 48 F.3d 1465 (9th Cir.1994), held that the “control of labor” prong of the integrated enterprise test may be established either by a showing of day-to-day control of labor, or by a showing that the parent was specifically responsible for the labor practice at issue in the litigation. See id. at 1471. Other courts have explained that all four factors of the integrated enterprise test are to be employed solely with an eye to discerning which entity — the parent or the subsidiary — was the final decisionmaker for the challenged practice. See, e.g., Hukill v. Auto Care, Inc., 192 F.3d 437, 444 (4th Cir.1999); Lusk v. Foxmeyer Health Corp., 129 F.3d 773, 777 (5th Cir.1997). Thus, the “directness” of a parent’s involvement in the employment decision under dispute may be conceived as a sliding scale; if the parent has sufficiently overwhelmed its subsidiary in taking the challenged action, such a showing is sufficient to create liability; if the parent was involved to a lesser degree, there must be some demonstration of the presence of the other aspects of the integrated enterprise test.

4. Choosing a Test for WARN Liability

Given these variations in the methods by which courts determine when corporations shall be liable for the acts of their affiliates, it is not surprising that there has been a good deal of inconsistency among the courts attempting to apply “existing law” in the context of the WARN Act. 5 In the first reported case on the subject, Local 397, International Union of Electronic, Electrical Salaried Machine & Furniture Workers v. Midwest Fasteners, Inc., 779 F.Supp. 788 (D.N.J.1992), the court employed three different tests — Third Circuit federal veil-piercing, the integrated enterprise test, and the DOL factors — to determine whether parent and grandparent corporations could be held liable for the debts of a subsidiary. The court concluded that the corporations were separate under an “alter ego” analysis, but identical under integrated enterprise analysis and the DOL factors. In reconciling these different outcomes, the court ultimately explained that the WARN Act was enacted to protect workers, and that the “wrongdoer” should not escape liability merely because “corporate formalities” were observed — a principle that the court noted had been established in federal labor statutes generally. Id. at 800. Thus, the court held that the outcomes of the integrated enterprise and DOL factors would control the analysis, rendering its entire discussion of alter ego not only superfluous, but also inapposite.

Other courts have followed the multi-part Midwest Fasteners approach, although the application of its principles varies widely. For instance, in Watts v. Marco Holdings, L.P., No. 3:95CV88-B-A, 1997 WL 578783 (N.D.Miss. Aug. 8, 1997), the court chose to apply state, rather than federal, veil-piercing analysis, all the while acknowledging that, as Midwest Fasteners had stated, for the purpose of determining whether two nominally separate companies constituted a single employer, state law veil-piercing was probably inappropriate. See id. at *2; see also United Paperworkers Int'l Union v. Alden Corrugated Container Corp., 901 F.Supp. 426, 436-39 (D.Mass.1995) (applying state corporate law, integrated enterprise, and the DOL factors and concluding that because WARN is a federal labor statute, the outcomes of the federal tests, rather than the state alter ego test, should control).

On the opposite end of the spectrum, the court in Wholesale & Retail Food Distribution Local 63 v. Santa Fe Terminal Services, 826 F.Supp. 326 (C.D.Cal.1993), also applying state alter ego principles, rejected the plaintiffs’ assertions that state veil-piercing was less important under WARN than the DOL factors, and chose not to employ the integrated enterprise test at all. See id. at 334-35. In United Mine Workers of America, District 2 v. Florence Mining Co., 855 F.Supp. 1466 (W.D.Pa.1994), the court, although purporting to follow Midwest Fasteners, actually appeared to apply only the DOL factors in concluding that two corporations did not constitute a single employer for WARN Act purposes. See id. at 1480. Finally, in International Brotherhood of Teamsters Local 952 v. American Delivery Service, 50 F.3d 770 (9th Cir.1995), the Ninth Circuit expressed doubts about the need to apply several different tests for liability, yet still chose to apply both the integrated enterprise test and the DOL factors, albeit concurrently due to the tests’ similarity. See id. at 776. 6

The current trend toward applying more than one test for affiliated corporate liability is manifestly unworkable. Not only does this approach generate considerable uncertainty for parties affected by the WARN Act (the briefs presented to us are exemplars; they spend an inordinate amount of time simply running through different possible tests for liability), but it also obfuscates the purposes of the inquiry itself, i.e., whether the affiliated corporation should be legally responsible for issuing WARN notice. Further, although the importation of state law standards into federal law is permissible when state law is deemed to effectuate federal policy, see Textile Workers Union of Am. v. Lincoln Mills of Ala., 353 U.S. 448, 457, 77 S.Ct. 912, 1 L.Ed.2d 972 (1957), state veil-piercing standards hardly seem likely to do so when such standards may generate inconsistency in an area of law that has always been characterized by insistence on uniformity. Cf. Antol v. Esposto, 100 F.3d 1111, 1115 (3d Cir.1996) (discussing the need for uniformity in the interpretation of collective bargaining agreements). The use of state law standards also has the potential to permit “[t]he policy underlying a federal statute” to be “defeated by ... an assertion of state power.” Anderson v. Abbott, 321 U.S. 349, 365, 64 S.Ct. 531, 88 L.Ed. 793 (1944). Finally, the multitest approach is both “unduly complicated,” American Delivery Serv., 50 F.3d at 776, and ultimately yields no definitive answer to the question of liability: When liability is uncertain enough to result in different outcomes for each of the different tests, there is no method of reconciling the results, much in the same way that a man with one watch always knows what time it is, but a man with two watches is never sure. Cf. Papa v. Katy Indus., Inc., 166 F.3d 937, 940 (7th Cir.1999) (criticizing the integrated enterprise test on the ground that there is no way to reconcile the results of the prongs).

Given these variations in the methods by which courts determine when corporations shall be liable for the acts of their affiliates, we decline to interpret the Department of Labor’s statement that it does not intend to create “new” law for WARN Act liability as a direction to courts to employ multiple tests within a single case. Rather, we conclude that the most prudent course is to employ the factors listed in the Department of Labor regulations themselves. This approach not only has the virtue of simplicity (if anything in this area of law can be described as “simple”), but also allows for the creation of a uniform standard of liability for the enforcement of a federal statute. Cf. United States v. Pisani, 646 F.2d 83, 87-88 (3d Cir.1981) (holding that federal veil-piercing standards are appropriate in Medicare disputes due to the need for a uniform federal approach). Finally, and most importantly, the DOL factors are the best method for determining WARN Act liability because they were created with WARN Act policies in mind and, unlike traditional veil-piercing and some of the other theories, focus particularly on circumstances relevant to labor relations.

The DOL factors are quite similar to the integrated enterprise test, which is understandable because the integrated enterprise test was also specifically intended to deal with labor relations. However, in addition to those factors that are analogous to the integrated enterprise factors, the Department of Labor’s version has included a fifth, catch-all factor — that of “de facto exercise of control” — that has the potential to tip the balance in an otherwise close case. This factor is arguably problematic, because read in isolation, it might well encourage the imposition of liability merely as a result of the control ordinarily exercised by a parent corporation over a subsidiary by virtue of its ownership. Such a result would cause a type of liability that is not only at odds with the purpose of limited liability in general, but also would be inconsistent with the “existing legal rules” regarding parental liability that the Department of Labor would have courts apply. See Bestfoods, 524 U.S. at 61-62, 118 S.Ct. 1876 (describing as “horn-book” law that a parent’s exercise of control through ownership of stock is not grounds for holding the parent hable for the subsidiary’s actions).

In reconciling this apparent tension, we observe that the DOL factors are, by their wording, more focused than their integrated enterprise test counterparts. For instance, rather than looking to “centralized control of labor relations” — the factor that, in the integrated enterprise context, could be satisfied either upon a showing that the parent and subsidiary functioned as a single entity, or, alternatively, upon a showing that the parent directed the subsidiary to institute the policy at issue — the DOL formulation is “unity of personnel policies,” a rendering that appears to be more targeted toward discerning whether the nominally separate corporations actually functioned as a single entity with respect to such policies on a regular, day-to-day basis. Similarly, the “common management” prong of the integrated enterprise test, which allowed courts to focus not only on employees holding formal officer positions or directorships but also on employees occupying supervisory positions, see, e.g., Hukill v. Auto Care, Inc., 192 F.3d 437, 443 (4th Cir.1999); Penntech Papers, Inc. v. NLRB, 706 F.2d 18, 25-26 (1st Cir.1983), has been changed to “common officers and/or directors,” a facially more specific requirement.

In light of these changes, and in light of the instruction that the test should draw upon existing legal rules, we read the “de facto exercise of control” factor as an endorsement of the sort of hybrid direct liability analysis heretofore employed in the context of the integrated enterprise test — allowing consideration not only of whether the two corporations shared the same labor policies, as the DOL’s “unity” factor would suggest, but also of whether the parent company directly exercised control over the particular policy at issue. We further conclude that the regulation’s specific instruction that the “factors” are a nonexhaustive list is meant only as a reminder that the inquiry is a balancing test, and that, as with most balancing tests, a number of circumstances may be relevant. Just as the integrated enterprise test is often described as ultimately an inquiry into whether the two companies operated at arm’s length, see NLRB v. Browning-Ferns Indus. of Pa., Inc., 691 F.2d 1117, 1122 (3d Cir.1982), we believe that the Department of Labor’s instructions are intended to allow the consideration of evidence that might otherwise fall outside of the listed factors in order to conduct such an inquiry.

5. Conclusion

In the long history of the corporate form and limited liability, both the common law and various pieces of legislation have developed numerous methods for determining when affiliated corporations should be treated as unified or as distinct entities. These methods are often quite different from each other, and vary across contexts. In light of this history, if we were to interpret the DOL’s instruction that courts apply “existing law” to determine WARN Act liability for affiliated corporations as a literal direction to employ the various tests that have been developed over the years, we would find ourselves ensnared in a web of complicated — and conflicting — lines of jurisprudence. We cannot believe that the Department intended such a result.

Rather, in our view, the Department intended for courts to test for affiliated corporate liability under WARN along the dimensions specifically enumerated in its regulation. These dimensions, in turn, were adapted from other tests developed for intercorporate liability, most notably labor law’s “integrated enterprise” test. In light of the similar considerations inherent in the DOL factors and in other such “veil-piercing” tests, we believe that the DOL’s instruction that courts apply “existing law” is intended only to encourage courts to make use of established precedent in interpreting and applying its factors. Further, via its statement that the factors are meant as a nonexhaustive list, the DOL has made room for the exercise of the flexibility that this area of law requires. Accordingly, in determining whether two or more corporations constitute a single “employer,” the factfinder may consider not only the aspects of corporate organization specifically listed in the regulation, but also may consider the other indicia of corporate “sameness” that have characterized this area of the law, such as nonfunctioning of officers and directors, gross undercapitalization, and other circumstances that demonstrate a lack of an arm’s-length relationship between the companies.

We also interpret the DOL’s inclusion of the “de facto exercise of control” factor to be an endorsement of the hybrid direct liability analysis heretofore employed in the context of the integrated enterprise test. Thus, the “de facto exercise of control” prong allows the factfinder to consider whether the parent has specifically directed the allegedly illegal employment practice that forms the basis for the litigation.

C. Lender Liability Under the WARN Act

1. Discussion

The preceding discussion focused on the standards to be employed for parent/subsidiary liability (or between sister corporations). But at the time their venture began, GECC was a major secured lender of CompTech, and not a parent. Neither the WARN Act itself, nor the regulations, explicitly discuss the statute’s applicability to lenders, but we agree with both the Eighth and the Ninth Circuits that, under some circumstances, a lender can become so entangled with its borrower’s affairs so as to engender WARN Act liability. See Adams v. Erwin Weller Co., 87 F.3d 269, 271 (8th Cir.1996); Chauffeurs, Sales Drivers, Warehousemen & Helpers Union Local 572 v. Weslock Corp., 66 F.3d 241, 244 (9th Cir.1996). Thus, the question becomes what circumstances must exist before such liability can attach.

Courts have grappled with the question of lender liability in a wide variety of situations, such that the catch-phrase “lender liability” has now taken on a broad meaning to refer to any kind of liability that can grow out of the lender/borrower relationship. See, e.g., Lawrence, Lender Control Liability, supra (describing various theories under which lenders can be held liable either to their borrowers or to third parties). For our purposes, the most relevant lines of precedent are those where third parties seek to impose liability on major lenders on the theory that the lenders have so controlled the borrowing corporation that the corporation was functionally being run by the lenders, or solely for the lenders’ benefit, to the detriment of other creditors. See, e.g., Krivo Indus. Supply Co. v. National Distillers & Chem. Corp., 483 F.2d 1098 (6th Cir.1973). Often, these claims arise in the context of a bankruptcy proceeding, whereby the creditors or the trustee seek equitable subordination of the major lender’s claims. See, e.g., In re W.T. Grant Co., 699 F.2d 599 (2d Cir.1983). In other situations, creditors simply sue the major lender on the theory that the lender’s control over the borrower rendered the lender the real party in interest for the incurred debt. See, e.g., Combustion Sys. Servs., Inc. v. Schuylkill Energy Res., Inc., Civ.A. 92-4228, 1993 WL 514496 (E.D.Pa. Dec.1, 1993).

In these situations, the test usually applied is some version of traditional veil-piercing, be it alter ego, instrumentality, or some other formulation. See, e.g., In re Clark Pipe & Supply Co., Inc., 893 F.2d 693, 699 (5th Cir.1990) (explaining that, in the absence of fraud, a lender must have used its debtor as an instrumentality to justify equitable subordination); Great West Cas. Co. v. Travelers Indem. Co., 925 F.Supp. 1455, 1462-63 (D.S.D.1996) (utilizing state veil-piercing standards to determine whether a lender would be liable to a third party for the debtor’s debts). This is precisely the test that was employed by the District Court when it concluded that because GECC was a lender and not a parent, the integrated enterprise test was inapplicable. See Pearson v. Component Tech. Corp., 80 F.Supp.2d 510, 520 (W.D.Pa.1999). We believe, however, that traditional lender/borrower veil-piercing jurisprudence is inappropriate in the WARN Act context for many of the same reasons that we rejected such jurisprudence in the context of parent/subsidiary liability.

To begin with, the precedent on this point does not draw as sharp a distinction between “lenders” and “parents” as the District Court perceived. Although Krivo and its progeny employed strict veil piercing standards to suits against lenders, they did so in situations where even parents would have been examined under such standards. The only differences between parents and lenders came in the test’s application, both via the court’s awareness of the changed context, see Krivo, 483 F.2d at 1110 (observing that the lender’s “control” was limited to its financial interest as a major creditor), and the court’s statement that the lack of stock ownership is “a factor to be considered in assessing the relationship between two companies,” see id. at 1109; see also Riquelme Valdes v. Leisure Res. Group, Inc., 810 F.2d 1345, 1353 (5th Cir.1987) (explaining that complete ownership is a “symptom but not the sine qua non of alter ego status”). It follows that when the appropriate test for parental liability is something other than the strict alter ego test, there should be a parallel change in the test for liability for lenders.

Further, traditional veil-piercing jurisprudence tends to sweep quite broadly, allowing liability to attach only when there is complete unity of identity in all aspects of corporate functioning. See, e.g., Krivo, 483 F.2d at 1105 (liability attaches to lenders only when there has been “total control” over the debtor). Although such an inquiry may be appropriate for many types of claims, the mere existence of the integrated enterprise test demonstrates that in the labor context, a more targeted inquiry is appropriate. We acknowledge that the DOL factors are explicitly made applicable in the WARN Act regulation only to “subsidiaries” and not to borrowers, but do not read that reference as precluding the application of the factors to lenders; rather, we believe that by directing courts to examine these particular factors, the Department of Labor was highlighting those aspects of corporate functioning that are most closely tied to the particular problems the WARN Act was intended to address. 7

Additionally, the problem with creating such a sharp distinction in liability rules under the WARN Act for lenders and for parents is that it will not always be clear when a party should be characterized as a “lender,” when a party should be characterized as a parent or owner, and when a party occupies both roles. In the case before us, GECC began its relationship with CompTech as a lender, but subsequently foreclosed on the stock and, rather than merely holding the stock for only a few days before the plant closure (as was the case in Weslock), transferred it (for no consideration) to CompTech’s Chief Executive Officer (then under contract to GECC), yet retained a considerable amount of control over the stock for the next several years as part of its plan to “hold” CompTech until the company could be restored to profitability. Given the ease with which Thomas Gaffney parted with the stock upon being asked by GECC to relinquish his position with CompTech, it is certainly not clear that GECC should not be viewed as having owned Comp-Tech’s stock from the date of foreclosure until the date that the company was finally liquidated.

Lenders may also occasionally be difficult to distinguish from parents because, although generally the difference between a “parent” and a “lender” is the existence of an equity, rather than a debt, interest in the company, lenders often structure their interests in hybrid ways. In this case, in addition to traditional loans, GECC chose to structure part of the debt by having CompTech modify its articles of incorporation so as to create a class of mandatorily redeemable preferred stock tailored to GECC’s interest. Such redeemable preferred stock is currently listed as neither equity nor liability according to U.S. generally accepted accounting principles. See International Accounting Standards, SEC Release Nos. 33-7801 & 34-42430, 65 Fed. Reg. 8,896, 8,911 (Feb. 23, 2000). However, it is sometimes classified as equity in SEC opinions, see, e.g., The Southern Company, SEC Release Nos. 35-27323 & 70-8277, 2000 SEC LEXIS 2860 (Dec. 27, 2000), although international accounting standards list such stock as a liability, see International Accounting Standards, supra, and federal regulations forbid such stock from being listed as stockholders’ equity, see 17 C.F.R. § 210.5-02. See generally Anthony P. Polito, Useful Fictions: Debt and Equity Classification in Corporate Tax Law, 30 Ariz. St. L.J. 761 (1998) (explaining the fluidity of the concepts of “debt” and “equity”).

Several federal statutes have defined the term “parent” in such a way as to include GECC’s interest. For instance, the Internal Revenue Code at 26 U.S.C. § 1563 defines as “parents” in a controlled group those companies that own eighty percent of the of the stock of other corporations in the group, and 26 U.S.C. § 1202 requires only fifty percent ownership. Ownership, in turn, is defined throughout the Code to include stock options. See, e.g., 26 U.S.C. §§ 318, 544, 554, 1563. Under these definitions, GECC was a parent of CompTech after the transfer to Charles Villa, because GECC retained options on all of Comp-Tech’s stock. The Code of Federal Regulations also contains definitions of “parent” and “subsidiary” that would include GECC’s relationship to CompTech as a result of its power to vote the stock in the wake of CompTech’s default. See 17 C.F.R. § 210.1-02. And Pennsylvania law defines the term “subsidiaries” for registered corporations as including those corporations for which another corporation has obtained options on fifty percent of the voting stock, see 15 Pa. Cons.Stat. § 2542, a definition that would also apply to GECC due to its calls obtained during the stock transfer to Villa. That the Supreme Court of Delaware described the right to vote a majority of the board in the event of default as a “creditor’s remedy” in In re Bicoastal Corp., 600 A.2d 343, 350 (Del.1991), merely selves to highlight the hybrid nature of such rights. Obviously, however, the regulatory purposes of these statutes (and the WARN Act itself) vary considerably.

Finally, we note that the commonly understood difference between a “parent” and a “lender” — i.e., the existence of an equity interest — is largely accounted for in the DOL factors themselves, via the “common ownership” prong. Although this factor is typically referred to as the “least important” of the factors, International Bhd. of Teamsters Local 952 v. American Delivery Serv., 50 F.3d 770, 775 (9th Cir.1995), these statements mean only that, by itself, ownership — and even ownership coupled with common management — is not a sufficient basis for liability, see Lusk v. Foxmeyer Health Corp., 129 F.3d 773, 778 (5th Cir.1997). Although “financial control” will suffice to satisfy the “common ownership” prong of the integrated enterprise test, see, e.g., Frank v. U.S. West, Inc., 3 F.3d 1357, 1362 (10th Cir.1993), and it is likely that the DOL factors should be interpreted similarly, there is nothing to prevent courts from requiring a higher showing of control in the absence of true ownership, just as they have done in traditional veil-piercing lender cases. See, e.g., Riquelme Valdes, 810 F.2d at 1354.

2. Conclusion

All things considered, including the absence of a satisfactory alternative, we are satisfied that the DOL factors are an appropriate method of determining lender liability as well as parental liability, and therefore hold that, regardless of whether GECC took on the status of “parent” in addition to its status of “lender” when it foreclosed on the stock, its involvement with CompTeeh will be tested by reference to those factors. We emphasize, however, that just as Krivo and similar cases took special note of the unique relationship between a lender and a borrower, so should courts do the same when utilizing the DOL factors. Thus, courts should place special weight on a lender’s lack of stock ownership, and the mere fact that a lender has loaned money to the borrower — thus making the borrower, in some sense, financially beholden to the lender— will not establish liability, or even “dependency of operations” as that phrase is used in the DOL test, just as a parent’s ownership of stock will not suffice to create liability for the parent.

Our application of the DOL factors will not, however, be dominated by an assessment of whether the defendant’s behavior was “typical” of a secured lender, as other courts (including the District Court in this case) have done. See Pearson, 80 F.Supp.2d at 525; see also Adams, 87 F.3d at 272 (refusing to hold a lender liable for WARN Act violations because the lender’s control was not “unusual for a lender loaning over eighteen million dollars”); Weslock, 66 F.3d at 245 (refusing to hold lender liable because the control exercised was “consistent with the type of control a secured creditor legitimately may exercise over a defaulting debtor” (quotations omitted)). “Typical” lender behavior is a mutable concept, and it will respond to the liability rules we put into place. See Robert E. Scott, A Relational Theory of Secured Financing, 86 Colum. L. Rev. 901, 934 (1986) (explaining that, if not for the liability rules currently in place, creditors would exercise more control over debtors than is now customary). Further, even total control of a delinquent borrower’s business might well be justified as an effort to protect collateral. See id. Thus, although courts should attend to the customary relationship between lender and borrower (just as they have attended to customary relationships between parents and subsidiaries in determining liability), they should also make a functional assessment of the amount of control involved.

D. Summary

Affiliated corporate liability under the WARN Act is ultimately an inquiry into whether the two nominally separate entities operated at arm’s length. Cf. NLRB v. Browning-Ferris Indus. of Pa., Inc., 691 F.2d 1117, 1122 (3d Cir.1982). To that end, the Department of Labor has specifically mandated consideration of: (1) common ownership, (2) common directors and/or officers, (3) de facto exercise of control, (4) unity of personnel policies emanating from a common source, and (5) the dependency of operations. We acknowledge that although these factors do not correspond precisely to established tests for liability, reliance.on analogous precedent may often be useful in the interpretation and application of those factors.

We believe that the DOL’s caveat that the factors are a nonexhaustive list is intended to allow the factfinder to consider other evidence, if any, of a functional integration between the two nominally separate entities — with, as always, an eye to the sorts of circumstances that courts have considered relevant to “veil-piercing” inquiries in the past. So, although ordinarily such hallmarks of integration as “nonfunctioning of officers and directors” and “nonpayment of dividends” are not of great importance in the labor context, certainly the factfinder would be permitted to take such arrangements into account when determining WARN Act liability, as well as any other arrangements that bear on the question whether the two companies failed to maintain an arm’s-length relationship.

Further, the “de facto exercise of control” factor allows the factfinder to consider, as has been done in the “integrated enterprise” context, whether a parent corporation was the final decisionmaker for the challenged practice. If the evidence of the parent’s control with respect to the practice is particularly egregious — for instance, if the parent corporation has “disregard[ed] the separate legal personality of its subsidiary” in directing the subsidiary to act, Esmark, Inc. v. NLRB, 887 F.2d 739, 757 (7th Cir.1989) — such evidence alone might be strong enough to warrant liability.

Finally, we conclude that these factors, with their labor-specific focus, are more appropriate than traditional veil-piercing jurisprudence for gauging WARN Act liability with respect to lenders. As has always been the case, when a plaintiff seeks to hold one corporation liable for the debts of another, particular attention must be paid to any lack of an ownership interest between the two corporations, and such a lack must be weighed heavily against a finding of liability for the affiliated corporation. Further, just as a parent will not be held liable solely because of its ownership of the subsidiary, so too a lender will not be hable solely because of the financial dependence that necessitated the loan in the first place.

However, notwithstanding the importance of a lack of ownership interest between two corporations, we believe that in the application of the DOL factors, the ultimate inquiry should not depend entirely on an assessment of whether the lender has behaved in a “typical” fashion. “Typical” lender behavior is a mutable concept, and it will respond to the liability rules we put into place. Thus, although the customary relationship between lender and borrower is a relevant consideration (just as customary relationships between parents and subsidiaries have always been relevant to the determination of liability), there must also be a functional assessment of the amount of control exercised by the lender.

III. Applying the Test

Now that we have set forth the standard for liability, we must determine whether plaintiffs have put forth enough evidence that GECC and CompTech constituted a single employer under the WARN Act to survive GECC’s motion for summary judgment. Therefore, we will analyze the plaintiffs’ evidence by reference to the DOL factors. We note at the outset that our adoption of the DOL factors as the proper test for WARN Act liability compels a rejection of the plaintiffs’ first contention, i.e., that solely by virtue of its ownership of CompTech stock options, GECC was a WARN Act employer.

In deciding that the DOL factors are appropriate for WARN Act veil-piercing, we also hold that the application of those factors is a “factual” question rather than a “legal” one. Although our own jurisprudence has been somewhat opaque as to whether the veil-piercing/alter ego tests present factual questions or legal ones, compare Craig v. Lake Asbestos of Quebec, Ltd., 843 F.2d 145, 148-49 (3d Cir.1988), with Carpenters Health & Welfare Fund v. Kenneth R. Ambrose, Inc., 727 F.2d 279, 283 (3d Cir.1983), it is well established that the “integrated enterprise” test — which is more akin to the DOL factors — presents a factual question, see, e.g., Limbach Co. v. Sheet Metal Workers Int’l Ass’n, 949 F.2d 1241 (3d Cir.1991). In light of our treatment of the integrated enterprise test, as well as the fact that most jurisdictions treat veil-piercing inquiries as factual, see Crane v. Green & Freedman Baking Co., 134 F.3d 17, 22 (1st Cir.1998), we conclude that the WARN Act test for intercorporate liability presents a question of fact. Of course, as is always the case, our decision to characterize it as such does not preclude an inquiry as to whether plaintiffs have put forth enough evidence to create a genuine issue of material fact so as to survive summary judgment.

A. Common Ownership

• After GECC purchased CompTech at a private foreclosure sale, it immediately transferred the shares to Gaffney and Brooks. When Gaffney stepped down as CompTech’s Chief Executive Officer, Gaff-ney and Brooks transferred their shares back to Holdings, and the shares were reissued to Villa. Technically, then, GECC was not the owner of CompTech at the time of the unnoticed plant closure, but we must acknowledge that the evidence, viewed in the light most favorable to the plaintiffs, allows for the possibility that these transactions were not bona fide transfers of ownership. See, e.g., NLRB v. Big Bear Supermarkets No. 3, 640 F.2d 924, 930 (9th Cir.1980) (upholding a finding of single employer status in part due to the fact that the transfer of ownership to an insider for no consideration did not bear the hallmarks of a bona fide transaction).

Gaffney was under contract to GECC at the time he received the shares; he paid no consideration for them, pledged them to GECC as security for CompTech’s loans, and, in fact, was indemnified by GECC for any liability arising out of his status as shareholder. He relinquished the shares for only nominal consideration at GECC’s request. When Villa received the shares, he paid only nominal consideration, pledged the shares and stock powers to GECC, and took them subject to GECC’s extensive options. Finally, GECC’s own restructuring memo (authored by GECC Vice-President Jeanette Chen) recommended both that CompTech be “held” until the “investment” could be recouped, and that “ownership” be transferred to the management team, suggesting that GECC understood itself to be the true owner of the shares despite the nominal vesting of title in Gaffney, Brooks, and Villa.

We agree with the Ninth Circuit that a prerequisite for lender liability is that whatever responsibility the lender may have assumed for the borrower’s business, such responsibility must have been for the “ordinary operation” of the business. Thus, as in bankruptcy law, the lender may not be hable under WARN for “winding up” or foreclosure activities not taken as part of an effort to operate the business in the “normal commercial sense,” Weslock, 66 F.3d at 245, even if, as a result of the foreclosure, the lender “owns” the shares for some period of time. Cf. In re United Healthcare System, Inc., 200 F.3d 170, 179 (3d Cir.1999) (holding that a fiduciary’s “winding up” activities in the course of a company’s liquidation pursuant to a bankruptcy fifing cannot give rise to WARN Act liability). However, GECC’s own documents demonstrate that it was not involved with CompTech merely long enough to wind up the affairs of the company; rather, in its own words, it intended to “hold” CompTech until its investment could be recouped and for CompTech to continue to conduct ordinary operations. For these reasons, we believe that, for summary judgment purposes, drawing all inferences in the plaintiffs’ favor, the common ownership prong is satisfied. GECC cannot escape the inference that it “owned” CompTech simply by focusing attention on the formal “ownership” of Gaffney, Villa, and Brooks, or by characterizing its activities as those of a lender engaged in liquidation.

B. Common Directors and/or Officers

This factor, which may be analogized to the integrated enterprise test’s “common management” factor, ordinarily looks to whether the two nominally separate corporations: (1) actually have the same people occupying officer or director positions with both companies; (2) repeatedly transfer management-level personnel between the companies; or (3) have officers and directors of one company occupying some sort of formal management position with respect to the second company. See, e.g., Frank v. U.S. West, Inc., 3 F.3d 1357, 1364 (10th Cir.1993) (finding no common management where the affiliated corporations shared no officers and only a single officer of the parent served in a managerial capacity with respect to subsidiaries); Local No. 627, Int’l Union of Operating Eng’rs v. NLRB, 518 F.2d 1040, 1047 (D.C.Cir.1975) (finding the common management prong satisfied by several interchanges of higher-level managers and officers between the corporations), overruled on other grounds by Southern Prairie Constr. Co. v. Local 627, Int’l Union of Operating Eng’rs, 425 U.S. 800, 96 S.Ct. 1842, 48 L.Ed.2d 382 (1976) (per curiam). In this case, there is no allegation that CompTech’s top officers and directors— Gaffney, Brooks, and Villa — ever occupied any sort of director or officer position with GECC. Rather, the plaintiffs’ theory is that these three men functioned as “agents” of GECC, and, because of that relationship, GECC’s officers and directors were also officers and directors of Comp-Tech.

Although we will discuss the application of agency principles to WARN Act liability in more detail below, for now we observe that the plaintiffs’ interpretation of this particular prong is misdirected. The WARN Act test — like the related integrated enterprise test — is intended to discover whether the two nominally separate entities actually functioned as a single business, particularly with regard to labor policy. See NLRB v. Browning-Ferris Indus. of Pa., Inc., 691 F.2d 1117, 1122 (3d Cir.1982); Armbruster v. Quinn, 711 F.2d 1332, 1338 (6th Cir.1983) (a finding of single-employer status is justified when the corporations are “highly integrated with respect to ownership and operations” (citations omitted)). To that end, the “common officers and/or directors” prong of the test should look only to whether some of the same individuals comprise (or, at some point, did comprise) the formal management team of each company. See, e.g., Armbruster, 711 F.2d at 1339 (in the integrated enterprise context, describing as “sketchy” evidence of common management even where a single person had served as president of both companies, and where the subsidiary had formally released decisionmaking power to the parent).

The plaintiffs’ theory, by contrast, is more appropriately employed to satisfy other prongs of the test, such as the “de facto exercise of control” factor, the factor relating to personnel policy, or the factor relating to dependency of operations. See, e.g., Frank, 3 F.3d at 1362-63 (evaluating an allegation that the officers of the subsidiary were controlled by the officers of the parent in the context of the interrelation of operations prong of the integrated enterprise test). This is particularly true in light of the fact that the DOL factors, rather than using the phrase “common management” as is utilized by the integrated enterprise test, instead specifically require the presence of common directors or officers, a formulation that is facially more narrow. Plaintiffs have not demonstrated the existence of any persons who, either simultaneously or in consecutive periods, held a directorate or an officer position both with CompTech and with GECC, and the “agency” theory with respect to Gaff-ney and Villa — which we discuss in more detail below — would not satisfy this requirement even if we were to accept its truth. Thus, we conclude that the plaintiffs have not created an issue of fact with respect to the presence of common directors or officers, and this factor must be weighed in favor of the defendant.

C. Unity of Personnel Policies Emanating from a Common Source

We begin by interpreting the language of this prong to require the factfinder to focus the inquiry less on the hierarchical relationship between the companies (as such relationships may be considered in other aspects of the test) than on whether the companies actually functioned as a single entity with regard to its relationships with employees.

That GECC and CompTech did not share labor policies, or even coordinate their labor policies, is essentially undisputed. 8 Affidavits from CompTech and GECC employees demonstrate that Comp-Tech had its own personnel managers who never received any direction from GECC regarding their duties, and no evidence has been offered to contradict these statements. But even were we to accept the plaintiffs’ theory that Villa, as GECC’s agent, at some level instituted all of Comp-Tech’s policies, including labor policies, on GECC’s behalf, such evidence is not sufficient to satisfy this prong of the test in the absence of indications that GECC had a particular interest in how CompTech’s labor policies were designed, or issued specific directives to Villa on the subject. The evidence that plaintiffs have offered to demonstrate that GECC had an interest in controlling the personnel policies at Comp-Tech is scant, and largely amounts to establishing that CompTech, pursuant to the terms of the loan agreement (and on a very few occasions) sought GECC approval for decisions to institute bonus programs and to pay salaries in excess of $100,000. Such limited monitoring of compensation expenditures as part of a general loan agreement requiring oversight of Comp-Tech’s spending in a number of areas is not sufficient to demonstrate a “unity of personnel policies emanating from a common source.”

There are some inconsistencies in the evidence submitted by GECC. GECC submitted the affidavit of Jeanette Chen, who swore that GECC had never attempted to control the hiring or firing of CompTech employees, and that the loan restriction preventing CompTech from paying salaries in excess of $100,000 was intended to ensure that CompTech maintained a frugal budget. But both of these statements were contradicted by other pieces of evidence in this litigation. Gaffney’s letter to GECC specifically states that he resigned his post as CEO of CompTech at GECC’s request, and other deposition evidence indicates that he was asked to resign due to GECC’s dissatisfaction with his strategies for CompTech. Moreover, internal GECC memoranda characterize the salary restriction as part of an effort to ensure that GECC could monitor the hiring of “key” personnel. Finally, the existence of the “Villa Call” and the references to Benton as a “back up president” seem to show that GECC, if nothing else, conceived itself as having control over the hiring and firing of CompTech’s president.

Despite these inconsistencies, we do not believe that plaintiffs have created a genuine issue of material fact as to the existence of a “unity” of policy. Even if we were to disregard Chen’s affidavit entirely and focus solely on the other evidence, the fact that GECC may have controlled the hiring and firing of the company’s president and chief executive officer, and monitored the hiring of a few other high-level managers (there is no evidence that GECC ever suggested that a particular person other than Benton be hired, or prevented a candidate from being hired), simply is not enough to find a “unity” of personnel “policy.”

The plaintiffs alternatively urge us to weigh this factor in their favor on the ground that the analogous “centralized control of labor” prong from the integrated enterprise test permits an inquiry into whether the parent corporation, though perhaps not involved in day-to-day employment policies, mandated the employment practice at issue. Plaintiffs allege that GECC made the decision to close the plant, and because the plant closing is the relevant practice giving rise to the litigation, submit that that decision is enough to tip the balance. However, as explained above, we believe that, for WARN Act purposes, allegations of this kind of “direct” control are more appropriately considered as part of the “de facto exercise of control” factor, discussed infra. We conclude that the plaintiffs have not created a genuine issue of fact as to the existence of a unified personnel policy, and we will weigh this factor in the defendant’s favor.

D. Dependency of Operations

We consider the “dependency of operations” factor to be virtually identical to the integrated enterprise test’s “interrelation of operations” factor, and will consider it in that light. When examining the “interrelation of operations” factor, courts generally consider the existence of arrangements such as the sharing of administrative or purchasing services, see, e.g., Penntech Papers, Inc. v. NLRB, 706 F.2d 18, 25 (1st Cir.1983), interchanges of employees or equipment, see Hukill v. Auto Care, Inc., 192 F.3d 437, 443 (4th Cir.1999), and commingled finances, see UA Local 343 of the United Ass’n of Journeymen & Apprentices of the Plumbing & Pipefitting Indus. of the U.S. & Can. v. Nor-Cal Plumbing, Inc., 48 F.3d 1465, 1472 (9th Cir.1994). No evidence of this sort was presented by the plaintiffs. Instead, plaintiffs reiterate their theory that Villa, and Gaffney before him, were stooges, or at least agents, of GECC, and that therefore all of their actions may be attributed to GECC. Because certainly Villa and Gaffney were routinely involved with the day-to-day operation of the business, the plaintiffs maintain GECC was in the position of general manager of the company.

We decide whether to apply agency principles to establish liability under a federal statute in accordance with the degree to which such principles effectuate the policies of the statute. See AT&T v. Winback & Conserve Program, Inc., 42 F.3d 1421, 1429-33 (3d Cir.1994). This is not dissimilar from the rules of common law, where there are different types of agency relationships, and the degree to which the actions of the agent are attributed to the principal vary for each. See id. at 1434-35. Thus, if we are to import any agency principles to the test for WARN Act intercorporate liability, we must do so selectively, with an eye to effectuating WARN Act purposes.

Plaintiffs argue that the “agency” relationships alleged to exist between GECC and Gaffney, and later between GECC and Villa, establish that GECC had control over the day-to-day operations of CompTech. Control over day-to-day operations has been held to be indicative of interrelation of operations. See, e.g., Lusk v. Foxmeyer Health Corp., 129 F.3d 773, 778 (5th Cir.1997); Cook v. Arrowsmith Shelburne, Inc., 69 F.3d 1235, 1241 (2d Cir.1995). However, the mere fact that the subsidiary’s ehain-of-eommand ultimately results in the top officers of the subsidiary reporting to the parent corporation does not estabhsh the kind of day-today control necessary to estabhsh an interrelation of operations. See Frank v. U.S. West, Inc., 3 F.3d 1357, 1362-63 (10th Cir.1993); Martin v. Safeguard Scientifics, Inc., 17 F.Supp.2d 357, 364 (E.D.Pa.1998) (“[I]t is never sufficient to estabhsh only that a chain of command eventually ends at the parent’s headquarters.”).

Moreover, dependency of operations cannot be established by the parent corporation’s exercise of its ordinary powers of ownership, i.e., to vote in directors and set general policies. See Lusk, 129 F.3d at 778. Therefore, if “agency principles” are to be used to weigh the “dependency of operations” factor in plaintiffs’ favor, the pohcies behind WARN Act liability — and limited liability for corporations generally — require that the plaintiffs estabhsh that the scope of the agency relationship was such that GECC had the right to direct and control the manner in which Gaffney and Villa undertook their duties. In other words, for such an all-encompassing factor such as “dependency of operations” — a factor which, by its nature, looks to the daily functioning of the two companies — the plaintiffs must estabhsh the existence of what was known at common law as a “master-servant” agency relationship. See AT&T, 42 F.3d at 1435.

Plaintiffs’ evidence on this point consists of: (1) letters and similar contacts from both Gaffney and Villa to GECC requesting permission to take certain kinds of actions that the loan documents would otherwise forbid; (2) the terms of the consulting agreement between Gaffney and GECC and the circumstances surrounding his employment; (3) the circumstances surrounding the firing of Gaffney and his replacement with Villa; and (4) the fact that consultant Stuart Benton was brought in at GECC’s request, and apparently worked closely with Villa.

We do not believe that this evidence is sufficient to create an issue of fact with regard to an agency relationship that would support a finding of dependency of operations. At most, the evidence establishes that Gaffney was something of an independent contractor, hired, according to the terms of his contract, to “exercise general executive authority over all business operations of the companies subject at all times to the control of the board of directors.” In other words, Gaffney was to run the company as he saw fit, and not in accordance with GECC directives. Under such circumstances, Gaffney’s actions are not attributable to GECC merely by virtue of the existence of his consulting agreement.

Further, because Villa was president and CEO of CompTech at the time of the closing, as well as at the time that plaintiffs argue that WARN Act notice should have issued, it is his relationship with GECC that is more relevant — and Villa’s relationship was even more tenuous than Gaffney’s. Villa was not under contract to GECC, and, in fact, Chen attested (and the plaintiffs have not disputed) that she was unaware that Gaffney had even replaced Brooks with Villa until the chief financial officer mentioned it to her in the course of conversation. Though GECC did apparently view itself as having the power to remove Villa (hence the Villa Call), as we have stated, such general powers of ownership are not a sufficient basis for a finding of liability. Further, the right to hire or fire does not, in itself, distinguish a “master-servant” agency relationship from any other sort of agency relationship. In fact, the most plaintiffs are able to muster with regard to an agency relationship between Villa and GECC is the letter Villa wrote asking GECC to “make decisions now” concerning the requests he had outlined. These requests, however, were for additional financing and (the letter is not entirely clear) perhaps permission for a large-scale merger. These are not the kind of mine-run matters that would support a finding that Villa was behaving as GECC’s agent in his management of CompTech.

The situation in Citibank, N.A v. Data Lease Financial Corp., 828 F.2d 686 (11th Cir.1987), presents a useful contrast. In that case, a corporation defaulted on its loans and the major creditor exercised its rights to elect a new slate of directors. The new directors allegedly mismanaged the company, and the borrower brought suit against the lender on the theory that the directors had acted as the “agents” of the lender. See id. at 691. The court concluded that the borrower had created a genuine issue of material fact as to the existence of an agency relationship, basing its decision in part on the deposition testimony of one of the directors, in which he stated that he both worked for the lender, and that he worked closely with the lender on all major matters of policy. See id. at 692. Such evidence stands in stark contrast to the situation before us, in which the man alleged to be GECC’s agent submitted an affidavit, as well as deposition testimony, denying that GECC ever controlled his actions, and the plaintiffs have produced no specific instances of GECC control to rebut his testimony.

As for Villa’s and Gaffney’s requests for GECC approval of actions involving large-scale expenditures, restructuring, or the disposition of equipment in which GECC retained a security interest, these cannot form the basis of a demonstration of “dependency of operations,” as there is no evidence that GECC ever did anything more than approve or disapprove such requests, and no evidence that it was involved in the details or manner of implementation of any CompTech business plans. GECC was exercising one of the defining rights of a secured lender: the right to prevent the transfer of its collateral. That GECC may also, at this point, have been a “parent” of CompTech does not prevent it from exercising such pedestrian creditor’s rights without incurring liability, any more than parental status prevents a company from controlling its subsidiary’s boards of directors without incurring liability. Cf. Japan Petroleum Co. (Nigeria) Ltd. v. Ashland Oil, Inc., 456 F.Supp. 831, 846 (D.Del.1978) (refusing to pierce the corporate veil in part because the “control” alleged by the plaintiffs was attributable to the fact that the parent corporation was a major creditor of its subsidiary). We do not intend to create a jurisprudence that discourages loans in general or rescues of troubled business enterprises in particular. Further, the fact that in an attempt to keep itself operational, CompTech independently sought additional financing from an outside lender to replace GECC cuts against a conclusion that CompTech was closely integrated with GECC itself.

If GECC’s “loans” to CompTech had not been made at arm’s length (either because of a failure to require interest payments, a lack of formal documentation, etc), such evidence would cut in favor of finding a dependency of operations. See, e.g., Vance v. NLRB, 71 F.3d 486, 493 (4th Cir.1995) (per curiam) (finding an interrelation of operations in part because of numerous interest-free, or unrepaid, “loans” between corporations). But there is nothing to suggest that GECC’s loans to CompTech were anything other than bona fide arm’s length transactions. As for the plaintiffs’ final allegation, that Benton served as something of a roaming GECC “enforcer,” once again, plaintiffs have failed to submit any support for their claims. Villa, Benton, and other top-level CompTech managers all attested that Benton did nothing more than observe and occasionally make suggestions, only some of which were followed. No evidence has been offered to suggest that GECC was using Benton functionally to manage CompTech’s day-to-day activities.

Finally, we address the question whether there can be said to have been “dependency” of operations due to the fact that CompTech was financially dependent on GECC’s loans, and ultimately was unable to stay afloat without them. As we remarked earlier, the loans were made in the ordinary course of-business, and there is no evidence that they included especially favorable terms or bore other indicia of a lack of legitimacy. Under such circumstances, loans — even from a parent to a subsidiary — cannot be sufficient to satisfy this prong, particularly in this context where there is no serious dispute that GECC, rather than attempting to establish a continuing relationship whereby Comp-Tech would be permanently dependent on GECC for financing, was instead by this point conducting a “rescue” operation in an attempt to “return Company to profitability,” We surely do not want to discourage companies from attempting to keep their subsidiary operations afloat with temporary loans by holding that the mere fact that loans were even necessary establishes a “dependency of operations” giving rise to liability. Cf. United Elec., Radio & Mach. Workers of Am. v. 163 Pleasant St., 960 F.2d 1080, 1094 (1st Cir.1992) (refusing to find “inadequate capitalization” in alter ego inquiry where parent repeatedly infused subsidiary with cash in an ultimately unsuccessful attempt to resurrect a failing company).

The same logic holds true for GECC’s enforcement of the prepayment provisions of its loans upon CompTech’s sale of R&R Plastics. Regardless of whether GECC was CompTech’s parent as a result of its options or its stock ownership, it also was indisputably a lender as a result of the revolving lines of credit and the repeated cash advances. As such, an attempt to seek repayment of its loans cannot be realistically characterized as a “siphoning” of funds or as part of a deliberate attempt to keep CompTech undercapitalized.

Therefore, we conclude that the plaintiffs have failed to create a genuine issue of material fact with respect to Comp-Tech’s “dependency” on GECC. As we have explained, there is no evidence that GECC’s loans were intended as a method of keeping CompTech deliberately under capitalized, and if dependency of operations is to be proved solely by the existence of an alleged “agency” relationship between the lender/parent corporation and the officers of the borrower/subsidiary, plaintiffs must submit evidence of far more oversight and control on the part of the principal than has been done in this case.

E. De Facto Exercise of Control

As we have explained, the “de facto exercise of control” factor is not intended to support liability based on a parent’s exercise of control pursuant to the ordinary incidents of stock ownership. Nor may this factor be used to create liability for a lender’s general oversight of its collateral. The factor is appropriately utilized, however, if the parent or lender was the decisionmaker responsible for the employment practice giving rise to the litigation. Further, because the balancing of the factors is not a mechanical exercise, if the de facto exercise of control was particularly striking — for instance, were it effectuated by “disregarding] the separate legal personality of its subsidiary,” Esmark, Inc. v. NLRB, 887 F.2d 739, 757 (7th Cir.1989)—then liability might be warranted even in the absence of the other factors.

In support of its allegations of “de facto” control by GECC, the plaintiffs continue to point to GECC’s general monitoring of CompTech; the fact that GECC’s approval was required under the loan agreements for a number of projects involving Comp-Tech’s finances, budgeting, and GECC’s collateral; the relationship among GECC, Villa, and Gaffney; and the fact that CompTech ultimately closed as a result of GECC’s decision to call the loans. See Part I, supra. Although, for the reasons discussed above, we do not find GECC’s monitoring of CompTech’s finances or its relationship with Villa and Gaffney particularly probative, we are given pause by the extent of GECC’s involvement in the decision to close the plant.

CompTech was kept operational for three years solely as a result of GECC’s own decision to hold on to CompTech and ensure the company’s return to profitability. During this time, CompTech was almost always behind in its payments to GECC, and was only able to survive by GECC’s extension of due dates and additional financing. Therefore, for three years, GECC was aware that its funding was the only thing keeping a troubled company afloat. It continued to invest, but when it finally concluded that Comp-Tech could not be saved, it immediately made the decision, according to its internal documents, not to “refus[e] to loan additional working capital,” Adams v. Erwin Weller Co., 87 F.3d 269, 273 (8th Cir.1996), but instead to “liquidate the company”— thus forcing CompTech to close its doors two weeks later. The decision is thus arguably less like a subsidiary’s independent choice to terminate its business in the face of severe cash constraints than like the decision of a WARN Act employer to close a single site of its operations. 9

On the other hand, the evidence also demonstrates that even before GECC decided to withdraw its support, CompTech was contemplating closure and seeking legal advice as to its labor obligations. CompTech’s repeated requests for financing, coupled with its warnings to GECC that it would be forced to shut down without further cash infusions, demonstrate that CompTech was acting as an independent entity seeking further capital rather than as a branch of GECC operating under GECC’s direction. In the absence of other indicia of a unified status between GECC and CompTech, we cannot conclude that the evidence supports an inference that GECC’s decision to accept CompTech assets as collateral on the loans manifested a “de facto exercise of control” equivalent to an employer’s decision to close a work site.

To be sure, the distinction between a decision to call a loan and a decision to shut down a company is a fine one, and the fact that GECC in some sense “owned” CompTech at the time of closure renders the distinction finer still. But the business of judging is, in considerable measure, one of line-drawing. While the demarcation here is not clear, indeed is quite close, we think it falls on GECC’s side of the line. We acknowledge that large lenders— whether or not they are also parents— cannot help but be aware of the consequences of their decisions to foreclose and dispose of collateral; in other areas of the law, a party is “presumed to intend all the natural and probable consequences flowing from his deliberate acts.” United States v. Applewhaite, 195 F.3d 679, 690 (3d Cir.1999) (quotations omitted). However, we must be scrupulous in our efforts to distinguish between situations in which a parent/lender has ultimately assumed responsibility for the continuing viability of a company (thus incurring liability for WARN Act violations) and situations in which the borrower has retained the ultimate responsibility for keeping the company active.

Here, CompTech’s independent research into its legal obligations, its negotiations with GECC, and its attempts to secure additional financing all reflect CompTech’s own vitality, and demonstrate that GECC’s decision to cut off its funding was not a “de facto exercise of control” over CompTech’s decision to close its doors. After consideration of all of the evidence, we conclude that the plaintiffs have not created a genuine issue of material fact as to whether GECC exercised “de facto” control over CompTech generally, or by virtue of its calling of the loan, and thus this factor weighs in favor of the defendant. 10

F. Summary

The DOL factors, like the integrated enterprise test, require that two corporations be “highly integrated with respect to ownership and operations” before they will be considered a single employer for WARN Act purposes. Armbruster v. Quinn, 711 F.2d 1332, 1338 (6th Cir.1983) (quotations omitted). The evidence proffered by the plaintiffs simply does not establish the high degree of integration required by the analysis set forth in this opinion. Though GECC may have monitored much of the activity at CompTech, there is no question that at all times CompTech remained an entirely separate business entity that did not rely on GECC to supply it with personnel, equipment, facilities, clients, administrative services, or any of the other various resources typically “shared” between companies that are ultimately found liable for each others’ debts. And, as we have discussed, there is a dearth of evidence to support the plaintiffs’ theory that GECC effected a pervasive control over all of CompTech’s functioning by virtue of its relationship with CompTech’s CEOs.

Although the plaintiffs have created a genuine issue of fact with respect to GECC’s ownership of CompTech, they have failed to meet their burden with respect to the other facets of the DOL test. The existence of an “agency” relationship between GECC and CompTeeh’s CEOs would not establish the existence of “common directors and/or officers,” and, without evidence establishing not only that an agency relationship existed, but also the scope of that relationship, the plaintiffs cannot rely on the “agency” theory to establish “dependency of operations.” Additionally, under the circumstances presented here, GECC’s decision to call its loan does not give rise to an inference of “de facto exercise of control.” Finally, the plaintiffs have not put forth any additional facts not included in the DOL factors that bear on the question of GECC’s liability. Therefore, the District Court’s grant of summary judgment to GECC will be affirmed. 11

1 . Summary judgment is proper if there is no genuine issue of material fact and if, viewing the facts in the light most favorable to the non-moving party, the moving party is entitled to judgment as a matter of law. See Fed.R.Civ.P. 56(c); Celotex Corp. v. Catrett, 477 U.S. 317, 322, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). At the summary judgment stage, the judge's function is not to weigh the evidence and determine the truth of the matter, but to determine whether there is a genuine issue for trial. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986).

2 . Although the tests employed to determine when circumstances justifying "veil-piercing” exist are variously referred to as the "alter ego,” "instrumentality,” or "identity” doctrines, the formulations are generally similar, and courts rarely distinguish them. See Phillip I. Blumberg, The Law of Corporate Groups: Substantive Law § 6.01, at 111 (1987). The most important differences across jurisdictions seem to reside largely in two aspects of these different formulations: first, whether an element of "fraudulent intent,” inequitable conduct, or injustice is explicitly required, see id. § 6.02, at 115, and second, a general sense that federal courts are more likely to pierce the veil in order to effectuate federal policy, lest state corporate laws be permitted to frustrate federal objectives, see Anderson v. Abbott, 321 U.S. 349, 365, 64 S.Ct. 531, 88 L.Ed. 793 (1944); United Elec., Radio & Mach. Workers of Am. v. 163 Pleasant St. Corp., 960 F.2d 1080, 1092 (1st Cir.1992).

3 . For comparison, the Massachusetts version requires consideration of common ownership, pervasive control, intermingling of activity and assets, undercapitalization, lack of corporate formalities, absence of records, nonpayment of dividends, insolvency at the time of the relevant transaction, siphoning of corporate assets by shareholders, nonfunctioning officers and directors, use of the corporation for the transactions of dominant shareholders, and use of the corporation for fraud. See Evans v. Multicon Constr. Corp., 30 Mass.App. Ct. 728, 574 N.E.2d 395, 398 (1991). The Illinois version considers the failure to maintain records and formalities, commingling of funds, undercapitalization, and one corporation treating the assets of the other as its own. See Van Dorn Co. v. Future Chem. & Oil Corp., 753 F.2d 565, 570 (7th Cir.1985).

4 . In a "double-breasted” operation, a company divides its business into union and nonunion shops. See Limbach Co. v. Sheet Metal Workers Int’l Ass'n, 949 F.2d 1241, 1245 (3d Cir.1991).

5 . In addition to the tests for liability listed above, there are numerous others that are employed less frequently, or only in specific contexts. Sometimes courts will impose liability on a parent for a subsidiary’s acts based on a theory of "agency,” i.e., that the subsidiary organization, although (unlike an alter ego) possessed of a distinct legal personality from the parent, has acted as the parent's agent in a series of transactions and therefore has the power to bind the parent. See, e.g., A.T. Massey Coal Co., Inc. v. International Union, United Mine Workers of Am., 799 F.2d 142, 147 (4th Cir.1986); Publicker Indus., Inc. v. Roman Ceramics Corp., 603 F.2d 1065, 1070 (3d Cir.1979); A. Gay Jenson Farms Co. v. Cargill, Inc., 309 N.W.2d 285 (Minn.1981).

Moreover, specific statutes have their own tests. For instance, the Multiemployer Pension Plan Amendments Act of 1980, 29 U.S.C. § 1381 et seq., treats as a single employer all businesses "under common control,” borrowing its definition of "common control” from the Internal Revenue Code. That definition, in turn, looks purely to stock ownership to define a "controlled group,” and ownership in this context includes stock options. See IUE AFL-CIO Pension Fund v. Barker & Williamson, Inc., 788 F.2d 118, 123 (3d Cir.1986). Cases construing the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. § 9601 et seq., have also set forth standards for determining when parent corporations will be responsible for the waste management responsibilities of subsidiaries. See United States v. Bestfoods, 524 U.S. 51, 118 S.Ct. 1876, 141 L.Ed.2d 43 (1998). And cases under the Securities Act of 1933, 15 U.S.C. § 77a et seq., and the Securities Exchange Act of 1934, 15 U.S.C. § 78a et seq., have fashioned distinct standards of liability for entities that "control” businesses found to be in violation of the securities laws. See Lawrence, Lender Control Liability, supra, at 1393.

6 . Just recently, in Hollowell v. Orleans Regional Hospital LLC, 217 F.3d 379 (5th Cir.2000), the Fifth Circuit upheld a jury finding that several corporations constituted a single employer for WARN Act purposes where the jury had been instructed to follow only the DOL factors. However, it is unclear whether the court ultimately held that the DOL factors, and only those factors, were the appropriate method of analysis, for it refused to consider the defendants’ arguments that the wrong legal standard had been applied. See id. at 389.

7 . In its own decision granting summary judgment to GECC, the District Court relied on cases involving standards for equitable subordination in the bankruptcy law context, such as In re W.T. Grant, 699 F.2d 599 (2d Cir.1983), and standards regarding allegations that lenders have breached fiduciary obligations to borrowers, such as in Temp-Way Corp. v. Continental Bank, 139 B.R. 299 (E.D.Pa.1992). See Pearson, 80 F.Supp.2d at 521. We do not believe that these precedents are particularly instructive. Both sorts of claims set a high bar for plaintiffs because the causes of action rely upon an allegation of wrongdoing by the lender. WARN Act liability, by contrast, requires no showing of fraud or even a culpable mental state, although a court may, in its discretion, reduce penalties if the employer proves good faith. See 29 U.S.C. § 2104(a)(l)(B)(4). Therefore, the standards justifying equitable subordination cannot be freely transferred. Cf. Lawrence, Lender Control Liability, supra, at 1388 (criticizing courts' tendencies to apply similar definitions of "control" for lender liability, no matter what the cause of action).

8 . We believe that a "coordinated” labor policy would be as relevant as a “unified” policy; for instance, in double-breasted operations, there is obviously no “unity” of policy — and yet this is precisely the type of situation that the test is meant to capture.

9 . Such a conclusion is strengthened by the fact that, although in many disputes arising under labor legislation, the acts giving rise to liability are likely to be ones that are controlled at the "local” level (for instance, acts of discriminatory hiring), the very nature of the acts that give rise to WARN Act liability (a plant closure) are likely to result from decisions made from the highest levels within the corporate structure, rendering an examination of a parent corporation's role in the decision more important.

10 . Plaintiffs argue that the doctrine of issue preclusion bars GECC from denying that it exercised control over CompTech ever since the 1991 foreclosure. At that time, GECC moved in the District Court for the Northern District of Illinois to enjoin the former Comp-Tech directors from filing for bankruptcy. In granting the motion, the court concluded that under the loan agreements, GECC had the right to elect a new board of directors upon a default, that GECC could only hope to recoup its investment in CompTech if the business continued as a going concern, and that if the former directors were to file for bankruptcy, GECC would be irreparably harmed. See General Elec. Capital Corp. v. Chicago Plastics Prods. Ltd. P’ship, No. 91C4291 (N.D.Ill. July 10, 1991). As the District Court in this action properly concluded, the Illinois court did not purport to hold that GECC exercised “control” over CompTech at all, much less the nature and degree of control relevant to an assessment of WARN Act liability. See Pearson v. Component Tech. Corp., 80 F.Supp.2d 510, 526 (W.D.Pa.1999). Thus, issue preclusion is inapplicable. See Temple Univ. v. White, 941 F.2d 201, 212 (3d Cir.1991).

11 . In their final attempt to evade summary judgment, plaintiffs argue that the District Court erred in excluding (without first conducting an in limine hearing) the expert affidavit of Thomas Myers. Myers, a qualified authority on banking practices and bank fraud, attested that in his opinion, GECC "substantively operated" CompTech after the foreclosure, and that the controls it exercised over CompTech "were not primarily limited to the financial controls normally exercised in a typical debtor/creditor relationship.” The District Court concluded on the basis of the transcript of an eight-hour deposition that Myers's lack of familiarity with CompTech corporate structure, coupled with his lack of familiarity with aspects of the loan agreement between GECC and CompTech, rendered his opinion lacking in a sufficient factual basis to be relevant to the plaintiffs' case. See Pearson, 80 F.Supp.2d at 529. We review a district court's decision to conduct a hearing regarding the reliability of expert evidence, as well as its decision to admit or exclude such evidence, for abuse of discretion. See Oddi v. Ford Motor Co., 234 F.3d 136, 154 (3d Cir.2000).

As we have discussed above, WARN Act liability does not turn on whether a party is a "parent” or a "lender,” or whether a "lender” behaved in typical fashion. Indeed, we have already acknowledged that GECC may well have been a parent of CompTech. Thus, because we have declined to create separate standards of liability for lenders and parents, regardless of whether the District Court was correct in concluding that Myers’s affidavit was lacking in a factual basis, his opinion was ultimately not germane and no hearing was necessary. Cf. Padillas v. Stork-Gamco, Inc., 186 F.3d 412, 418 (3d Cir.1999) (holding that in limine hearings are encouraged when courts are concerned with the factual, rather than legal, dimensions of the evidence). We therefore affirm the District Court's decision to exclude the affidavit.