dissenting.
I cannot agree with the Court that Mr. Davis is not a transferee of Cowboy’s funds. I do agree that the bank was merely a conduit; the fact that the funds flowed through the bank is immaterial to the central issue, which is the bank’s lack of dominion and control over the funds. The Court similarly strays from the central issue by analyzing the flow of funds through various accounts rather than asking if Mr. Davis, in fact, exercised dominion and control over the funds as required by the test established in Bonded Fin. Serv., Inc. v. European Am. Bank, 838 F.2d 890 (7th Cir.1988), and adopted by us in Malloy v. Citizens Bank of Sapulpa (In re First Sec. Mortgage Co.), 33 F.3d 42 (10th Cir.1994). I believe that the most persuasive case law as well as the concerns of equity and fairness support finding Mr. Davis to be the initial transferee and the Markgrafs to be subsequent transferees.
As the Court begins its analysis with the Seventh Circuit’s decision in Bonded, so do I. Both Bonded and our decision in Malloy held that a bank which acted as a conduit of funds *945was not an initial transferee under § 550 because the bank did not exercise dominion and control over those funds. Malloy, 33 F.3d at 44; Bonded, 838 F.2d at 893-94. Neither case directly answered the question of who, if not the bank, is the initial transferee. However, those decisions do indicate that we must apply the dominion and control test to determine whether Mr. Davis or the Markgrafs is the initial transferee of Cowboy’s funds. See Malloy 33 F.3d at 43-44; Bonded 838 F.2d at 893-94. In IRS v. Nordic Village, Inc. (In re Nordic Village, Inc.), 915 F.2d 1049 (6th Cir.1990), rev’d on other grounds, 503 U.S. 30, 112 S.Ct. 1011, 117 L.Ed.2d 181 (1992), the Sixth Circuit confronted a similar factual situation. In that case, Joseph Lah, an officer and shareholder of the debtor, drew a check on the corporate account made payable to a bank, which in turn issued several cashier’s checks to Lah. The cashier’s check at issue was made payable to the IRS, and bore the notation “RE-MITTER: SWISS HAUS, INC.,” under which name the debtor was doing business at the time. Id. at 1050. The court held that the IRS was liable, but it did so on the grounds that the IRS did not demonstrate that it took the cashier’s check for value, in good faith, and without knowledge of the fraud, without reaching the “initial transferee” issue. Id. at 1055-56 & n. 3. In a line of reasoning that is applicable to the case before us, the court stated in dictum:
If Lah is viewed as acting for Nordic, then the IRS is the “initial transferee.” If Lah is viewed as having taken money illegally from Nordic, he is the “initial transferee” and the delivery of the cashier’s check to the IRS makes the IRS an “immediate transferee” of Lah, the “initial transferee.” If the IRS is considered as an “immediate transferee” of Lah, the IRS can prevail if the IRS shows that it took for value, in good faith, and without knowledge of the voidability of the transfer.
Id. at 1055 (footnote omitted).3
Based on this reasoning, it seems clear that Mr. Davis is the initial transferee of Cowboy’s funds. The Davises purchased the cashier’s cheek using Cowboy funds, and then used it to satisfy Mr. Davis’s personal indebtedness to the Markgrafs. Surely, under these circumstances, Mr. Davis must be viewed as having taken Cowboy’s money illegally, see Nordic, 915 F.2d at 1055, making him the initial transferee of the fraudulent conveyance.
Other ease law supports this conclusion. In Ross v. United States (In re Auto-Pak, Inc.), 73 B.R. 52 (D.D.C.1987), the owner of the debtor company converted a cheek of the debtor intended for the IRS into a cashier’s cheek drawn on the debtor’s account made payable to the IRS. He then directed the cashier’s check to the IRS with a notation to apply it toward taxes owed by another, non-bankrupt, corporation he owned. Sitting in an appellate capacity, the district court held that by converting the debtor’s check into a cashier’s check and writing the name of another corporation on it, the owner “essentially took control of the funds underlying the cashier’s check.... It would defy logic to hold an innocent and mediate party such as the IRS a party to this alleged fraudulent conveyance.” Id. at 54.
Another case, Kendall v. Sorani (In re Richmond Produce Co.), 151 B.R. 1012 (Bankr.N.D.Cal.1993), illustrates how a principal can rise to the level of a transferee by exercising sufficient control over a transaction. In Kendall, the court noted:
The evidence is clear that Clow [the principal] picked up the Cashier’s Check from Mechanics Bank and delivered it to Ban-Cal.... If a messenger service had picked up the Cashier’s Cheek from Mechanics Bank and delivered it to BanCal, the messenger service would not have been deemed a transferee....
However, clearly, Clow was no mere messenger. Rather, Clow exercised complete control over the transaction. It was devised and executed for his benefit. Un*946der these circumstances, the Court must conclude that the Cashier’s Check was transferred to Clow when he picked it up from Mechanics Bank.
Id. at 1021.
In the present case, Mr. Davis and his wife exercised a similar degree of control over the transaction. The Davises chose to instruct the bank to make the cashier’s check payable to the Markgrafs, but, in their position as principals of Cowboy, they could have ordered the bank to issue the cashier’s check to anyone, and for any purpose. Mr. Davis then delivered the cashier’s check to the Markgrafs in satisfaction of his personal debt. Moreover, the fact that the bank sent the cashier’s check to Mr. Davis’s home address illustrates his dominion and control over the funds. Once he had possession of the cashier’s check, Mr. Davis could have returned it to the bank, altered it in some way, or otherwise chosen not to deliver it. See Laird v. Bartz (In re Newman Cos.), 140 B.R. 495, 498 (Bankr.E.D.Wis.1992). Of course, if the cashier’s check had never reached the Markgrafs, the Markgrafs would not be deemed a transferee of the funds. I cannot agree with the Court’s contention that Mr. Davis was “at most, a mere courier.” Ct.Op. at 940. The intermediate steps taken by the Davises evince the dominion and control required to find them to be the initial transferees.
Finally, in Robinson v. Home Sav. of Am. (In re Concord Senior Hous. Found.), 94 B.R. 180 (Bankr.C.D.Cal.1988), the debtor hired Karl Gerwer, doing business as Charter Pacific Management, Inc. (“CPM”), to manage its apartment braiding. Gerwer collected rent for debtor and deposited it into a certificate of deposit at defendant bank in the name of Gerwer and CPM only. Subsequently, Gerwer used the deposited amounts as collateral to secure loans by Gerwer from the bank for his personal use. When Gerwer defaulted, the bank seized the CD. When the trustee tried to recover the seized funds from the bank, the court held that Gerwer, not the bank, was the initial transferee. The court held that “[t]he initial transfer of property from [debtor’s] estate occurred when Gerwer used the CD for his own benefit by pledging it to secure his loans.” Id. at 183. This misappropriation of the estate’s funds by the estate’s agent constituted a transfer of those funds to the agent. Similarly, when the Davises stepped outside of their role as principals of Cowboy and effectuated a fraudulent conveyance of Cowboy’s funds to satisfy a personal obligation of Mr. Davis, Mr. Davis became the initial transferee of those funds.
The majority relies heavily on Richardson v. FDIC (In re Blackburn Mitchell, Inc.), 164 B.R. 117 (Bankr.N.D.Cal.1994), and General Electric Capital Auto Lease, Inc. v. Broach (In re Lucas Dallas, Inc.), 185 B.R. 801 (Bankr.9th Cir.1995), to reject the above cited cases as bad law and hold that Mr. Davis is not the initial transferee, but rather the “entity for whose benefit such transfer was made.” I disagree with this reasoning. It is true, as the Bankruptcy Appellate Panel stated in General Electric, that corporations must always act through individuals. 185 B.R. at 809. However, there is an important difference between a principal’s actions on behalf of a corporation for corporate purposes, albeit misguided, and actions for purely personal purposes. Making such a distinction does not “collapse the two prongs of strict liability [for initial transferees and entities for whose benefit the transfer was made] into a single party,” as the General Electric Court warns. Id. Rather, it merely recognizes that the statute does not hold two separate parties strictly liable in all eases, but only in those where there are separate identifiable parties who are the “initial transferee” and the “entity for whose benefit such transfer was made.” As I believe Mr. Davis is a transferee, he cannot be the entity for whose benefit such transfer was made.
Moreover, the Richardson case, on which the reasoning of General Electric is largely based, both misapplies the legal rule established in Bonded, and is distinguishable on its facts. The Richardson court explained:
The Court believes that the proper focus when analyzing who is a transferee, is in the flow of funds. In order to be an initial transferee, one must be a transferee in the ordinary sense of the word.
*947This Court does not disagree that in order to be a transferee one must obtain dominion and control over funds. But that does not mean that merely because one has dominion and control of funds (as principals ordinarily do) that one is also a transferee. ... In order to be a transferee of the debtor’s funds, one must (1) actually receive the funds, and (2) have full dominion and control over them for one’s own account, as opposed to receiving them in trust or as agent for someone else.
Id. at 126. This is not what Bonded holds. The Bonded court specifically shifts the emphasis from the mechanical movement of money through accounts and establishes the dominion and control test. Bonded, 838 F.2d at 893. It is clear that “transferee” cannot simply be defined under the “ordinary sense of the word;” rather, Bonded holds that “ ‘[transferee’ is not a self-defining term; it must mean something different from ‘possessor’ or ‘holder’ or ‘agent.’ ” Id. at 894. In Richardson, the court created an overly restrictive definition by combining the dominion and control requirement of Bonded with the “ordinary” meaning of “transferee” as someone who actually holds the funds in his account. Of course, not every principal in a business, who ordinarily would have dominion and control over company funds, should be considered a transferee. As discussed above, there is a difference between a principal acting in his official capacity and a principal who abuses his official role and effectively steals money from his company to achieve personal rather than corporate purposes. See Nordic, 915 F.2d at 1055. It is this second type of principal whose wrongdoing causes him to abandon his official capacity and exercise personal dominion and control over corporate funds. The facts of this case clearly support the conclusion that Mr. Davis and his wife exercised personal dominion and control over Cowboy’s funds by improperly using their official status to satisfy personal obligations. These facts demonstrate more clearly than in Richardson that Mr. Davis did exercise the requisite dominion and control to be considered the initial transferees. Arguably, Ms. Mitchell in Richardson could be compared to the hypothetical situation in Bonded in which Ryan in his official capacity tells the bank to “use this check to reduce Ryan’s loan.” In that case, the court explained that the bank, and not Ryan, would be the initial transferee. Bonded, 838 F.2d at 892. Here, however, the situation of Mr. Davis is more akin to the dominion and control exercised by parties who were found to be initial transferees in the cases discussed above. See Kendall, 151 B.R. at 1021; Robinson, 94 B.R. at 183.
If, as the Court contends, Mr. Davis is the “entity for whose benefit such transfer was made,” the Markgrafs would be the initial transferees. However, not every case includes a party who can be labeled as the “entity for whose benefit such transfer was made.” In Bonded, the court makes clear that “the categories ‘transferee’ and ‘entity for whose benefit such transfer was made’ are mutually exclusive.” Bonded, 838 F.2d at 896; see also Danning v. Miller (In re Bullion Reserve of N. Am.), 922 F.2d 544, 548 (9th Cir.1991). Therefore, in a simple example where the debtor fraudulently transfers funds to A, who in turn transfers them to B, there is no party who can be properly described as the “entity for whose benefit such transfer was made.” A is the initial transferee and B is an immediate or mediate transferee.
The single distinguishing characteristic of an “entity for whose benefit such transfer was made” is, therefore, that it is not a transferee under § 550. Rather, in order to avoid redundancy, I interpret the phrase “entity for whose benefit such transfer was made” to describe a party that exists outside of the chain of transfer, but who nonetheless receives a direct benefit from the occurrence of the transfer, such as a guarantor of a loan or other third party beneficiary. Accord Laird, 140 B.R. at 498-99 (“The structure of section 550 distinguishes transferees ... from entities that get a benefit because someone else received the money or property. Since the bank was a subsequent transferee, it could not be the ‘entity for whose benefit’ the initial transfer was made.”). Employing the test of Bonded and its progeny, an “entity for whose benefit such transfer was made” must therefore be a party that does not exercise dominion and control over *948the transferred funds and is outside the chain of transfer. The facts presented clearly support finding Mr. Davis to be a transferee; consequently, under Bonded, he cannot be “the entity for whose benefit such transfer is made.” Mr. Davis was no mere courier; his improper intent and dominion and control over the transaction is crystal clear, compelling my conclusion that he was a transferee of the funds, and more specifically, the initial transferee. As Mr. Davis is the initial transferee, the Markgrafs are subsequent transferees who can utilize the good faith defense of § 550(b)(1).
Because the Court chooses to limit its focus to the form of the transaction, totally ignoring its substance, I respectfully dissent.
. As indicated above, the Supreme Court later reversed the Sixth Circuit’s decision in Nordic, 503 U.S. 30, 112 S.Ct. 1011, 117 L.Ed.2d 181 (1992). However, it did so on the alternative ground that the judgment against the govemment was jurisdictionally barred by the doctrine of sovereign immunity. Id. at 39, 112 S.Ct. at 1017. In disposing of the case on jurisdictional grounds, the Court expressed no opinion at all regarding the merits.