In the
 United States Court of Appeals
               For the Seventh Circuit
                         ____________

No. 02-3127
YVETTE GAFF KLEVEN, Trustee,
MARK A. WARSCO, Trustee,
DAVID R. DUBOIS, Trustee,
                                           Plaintiffs-Appellants,
                                v.

HOUSEHOLD BANK F.S.B.,
                                             Defendant-Appellee.
                         ____________
            Appeal from the United States District Court
     for the Northern District of Indiana, Fort Wayne Division.
              No. 1:02CV26—William C. Lee, Judge.
                         ____________
    ARGUED JANUARY 24, 2003—DECIDED JUNE 30, 2003
                    ____________


  Before RIPPLE, EVANS, and WILLIAMS, Circuit Judges.
  EVANS, Circuit Judge. Today, some people can buy a
luxury automobile, say a Lexus or a Cadillac, and pay it
off with an interest-free loan. The signs are everywhere:
“Zero % financing.” People in the market for luxury cars,
however, are not in the market for Refund Anticipation
Loans (RALs), the subject of the appeal in this case. RALs
attract millions of people who are short on cash—those
who either can’t wait, or choose not to wait, for the Inter-
nal Revenue Service to process their income tax returns
and send them a refund check through the mail. These
people want their money now, and to get it they use a tax
2                                               No. 02-3127

preparer like H & R Block and strike a bargain with a
bank like Household, the defendant here. The bargain
struck is a good one for only one of the two parties. Guess
which one?
   We start with a little background on RALs and some
statistics from Milwaukee, a fairly typical American city.
According to the national research, over half of all U.S. tax
filers use commercial tax preparers, but the marketing
of RALs is aimed mainly at low income families eligible
for the earned income tax credit (EITC). Most of these
people are without personal bank accounts in which to
receive a direct deposit tax refund. A recent report, pre-
pared by the University of Wisconsin-Milwaukee Employ-
ment and Training Institute in cooperation with The
Brookings Institution, reviews the use of RALs by ZIP
Code areas in central city Milwaukee neighborhoods. The
report used the tax year 2000 and data obtained by
Brookings staff from the Internal Revenue Service. Nine
central city Milwaukee ZIP Codes in the heart of Mil-
waukee’s Community Development Block Grant (CDBG)
neighborhoods—the poorest neighborhoods—were analyzed,
as well as other more affluent areas of the city and its
suburbs. The study disclosed that in Milwaukee County,
29,458 tax filers claiming the federal earned income tax
credit used tax preparers and borrowed against their
refunds through RALs prior to receiving them from the
IRS. The lion’s share of these filers lived in Milwaukee’s
distressed CDBG neighborhoods.
  The Brookings study found that national tax prepara-
tion chains typically charge around $100 for preparing
and filing a simple tax return and an additional $75 to
$100 fee for a RAL. At these rates, the study concludes,
lower-income Milwaukee filers who used commercial tax
preparers and RALs lost $5.5 million in tax refunds and
federal EITC payments in 2000.
No. 02-3127                                                3

  In a fairly routine case, a tax return filer like those
identified in the study pays a fee for preparation of the
return—say $100. The return discloses a refund due most
likely from the EITC of, say, $600. After paying the tax
preparer’s fee, the filer has a choice: wait for an IRS check
for $600 or enter into a RAL for a fee, usually around
$75. Because walking out with $525 (or $425 if the tax
preparation charge is rolled into the deal) apparently
seems like a better option than waiting for $600, the
bargain is struck. And unlike the Lexus buyer paying
zero percent annual interest, the filer pays what amounts
to an exorbitant interest rate—often over 100 percent—
when computed on an annual percentage basis. For more
on the study we have referred to see Alan Berube, Anne
Kim, Benjamin Forman and Megan Burns, The Price of
Paying Taxes: How Tax Preparation and Refund Loan
Fees Erode the Benefits of the EITC (The Brookings In-
stitution and the Progressive Policy Institute, May 2002);
Anne Kim and Alan Berube, “Fast Cash for the Tax Man,”
Blueprint Magazine, May 21, 2002; and “Use of Refund
Anticipation Loans by Earned Income Tax Credit Filers
in Central City Milwaukee Neighborhoods,” a study by
Lois M. Quinn, Employment and Training Institute,
School of Continuing Education, University of Wisconsin-
Milwaukee, June 2002.
  As we see it, an attack on RALs based on fairness and
equity would certainly have some appeal. But the appeal
we consider today attacks RALs from a different angle. In
our case, over 100 adversary proceedings were consoli-
dated in the bankruptcy court, and all of them present
the same question: What happens when a tax filer enters
bankruptcy proceedings within 90 days of signing off on
a RAL? The plaintiffs are trustees in chapter 7 cases
seeking to recover the tax refund payments received by
Household Bank. The trustees want to get their hands on
the funds under two theories, arguing that the funds are
4                                              No. 02-3127

impermissible setoffs or avoidable preferences under §§ 553
and 547 of the Bankruptcy Code.
  The transactions in question proceeded in the follow-
ing manner. During the 90-day period before filing for
bankruptcy, the debtor applied for a RAL from Household
by filling out a form entitled “Loan Application, Authoriza-
tion and Certification for a Refund Anticipation Loan.”
The debtor also established a bank account at Household
for the sole purpose of electronically receiving a federal
income tax refund. In addition, the debtor completed a
Form 8453 consenting to have the IRS deposit his tax
refund directly into the account with Household and,
with the bank’s helping hand, he designated a routing
number and bank account number corresponding to that
account on his return. At or about the same time, the
debtor was provided with two additional items—a dis-
closure document and a summary sheet.
  The bankruptcy court rejected the trustees’ pleas, con-
cluding that the transactions resulted in money in the
bank’s pocket that was more appropriately considered a
recoupment and that the bank came upon the funds in
the ordinary course of business. Warsco v. Household
Bank F.S.B., 

(Bankr. N.D. Ind. 2002). The
district court agreed with this determination when it
rejected the trustees’ appeal. We begin with a closer look
at the mechanics of a RAL.
  It is undisputed that the bank retained exclusive con-
trol over the RAL accounts and that no funds other than the
tax refund were deposited into them. And the bank’s
exclusive control, of course, meant that the debtors could
not get their hands on the funds once they reached the
designated account. The undisputed facts also provide
the following information about each individual RAL
transaction: (1) the name and social security number of
the debtor, (2) the loan application date, (3) the loan
No. 02-3127                                               5

amount, (4) the date the loan check was issued, (5) the date
the electronic fund transfer from the IRS was received
into the debtor’s account at the bank, and (6) the date the
bank took control of the deposit. The parties have also
stipulated, in each case, as to the date the bankruptcy
petition was filed and the prior years, if any, when the
debtor participated in the bank’s RAL program. As to
this point, it is noteworthy that only 15 of the more than
100 debtors in our case were first-time parties to RALs
with Household. Lastly, the parties agree that all of the
transfers were made in accordance with the ordinary
business practices of parties engaged in the tax refund
loan industry.
  In responding to the trustees’ claims, both below and
here, the bank contends that each debtor made an
absolute assignment of his tax refund or, alternatively,
that it held a perfected security interest in the refund or
its proceeds. As an affirmative defense to the preference
claims, the bank asserts that the money it received is
protected by the ordinary course of business defense
under § 547(c)(2). It also characterizes each payment as
a recoupment, rather than a setoff, which prevents the
trustees from recovering the funds under § 553(b) or
§ 549(a).
  Upon submitting an application, or perhaps a day
later, the debtor received a check from Household in the
amount of his anticipated federal tax refund, less House-
hold’s fee. Sometime thereafter, the IRS deposited the
debtor’s tax refund into the debtor’s designated account
at Household. Household then immediately transferred
the funds to itself. It is this transfer, by which the
bank actually received the money, that the trustees seek
to avoid.
  Section 547(b) of the Bankruptcy Code allows a trustee
to avoid some payments to creditors made during the 90
6                                             No. 02-3127

days prior to bankruptcy. These payments are regarded
as “preferential,” and thus avoidable, because they allow
the favored creditor to receive more than if he had to
wait in line with other creditors and share with them
what is usually rather slim pickings from the debtor’s
estate. The purpose of this section is two-fold. It is de-
signed to promote equality of distribution among all
creditors while simultaneously it deters creditors from
“racing to the courthouse to dismember the debtor dur-
ing [his] slide into bankruptcy,” Union Bank v. Wolas,

, 161 (1991). See also In re Milwaukee Cheese
Wis., Inc., 

, 847 (7th Cir. 1997).
  But not all payments to creditors during the 90 days
prior to bankruptcy are avoidable preferential transfers.
In crafting § 547, Congress created exceptions protect-
ing certain transfers from a trustee’s reach. See 11 U.S.C.
§ 547(c). Among them is the “ordinary course of business”
exception. Where a transfer is (a) “in payment of a debt
incurred by the debtor in the ordinary course of business
or financial affairs of the debtor and the transferee,” (b)
“made in the ordinary course of business or financial
affairs of the debtor and the transferee,” and (c) “made
according to ordinary business terms,” it is insulated from
avoidance. See 11 U.S.C. § 547(c)(2)(A)-(C). This excep-
tion was designed to “leave undisturbed normal com-
mercial and financial relationships and protect recurring,
customary credit transactions which are incurred and
paid in the ordinary course of business of both the debtor
and the debtor’s transferee.” See In re Armstrong, 

, 729 (Bankr. E.D. Ark. 1999), aff’d, 

(E.D. Ark. 2001). These transactions are not considered
“preferential,” although they would otherwise fit comfort-
ably within the statutory definition. The creditor has the
burden of proving that the debtor’s payment to it qualifies
for this protection. See 11 U.S.C. § 547(g).
No. 02-3127                                             7

  We have examined the ordinary course of business
defense on several occasions. See, e.g., Barber v. Golden
Seed Co., 

(7th Cir. 1997); In re Milwaukee
Cheese Wis., Inc., 

(7th Cir. 1997); In re
Midway Airlines, Inc., 

(7th Cir. 1995); In re
Tolona Pizza Prod. Corp., 

(7th Cir. 1993). And
we must consider whether the transaction, the RAL here,
was “ordinary” not only as between the debtor and the
bank, but also in the context of how the RAL industry
operates as a whole. See Midway 

.
Because it is undisputed that the transactions were “ordi-
nary” within the tax refund loan industry, if they were
also ordinary as between Household and the individual
debtors, they are not avoidable preferences. In making
this determination we consider several factors, including:
   (1) the length of time the parties were engaged in
   the transaction at issue;
   (2) whether the amount or form of tender differed
   from past practices;
   (3) whether the debtor or creditor engaged in any
   unusual collection or payment activity; and
   (4) whether the creditor took advantage of the debtor’s
   deteriorating financial condition.

. This is not an exhaustive list
of factors. Others, in appropriate considerations, may be
a better fit.
  In our case, where the debtor was a prior RAL cus-
tomer of the bank, that history persuades us that the
transactions challenged were ordinary as between House-
hold and the debtor because they occurred in the same
manner and under the same circumstances as in prior
years. Where, however, the challenged transaction was
the first RAL between the parties, a closer question is
presented.
8                                                No. 02-3127

  We have not directly addressed this “first-time” issue but
other courts have, with mixed results, although most
side with the view that a first-time transaction is not
per se ineligible for protection from avoidance under
§ 547(c)(2). Compare In re Finn, 

(6th Cir.
1990); In re Russell Cave Co., 

(Bankr. E.D.
Ky. 2001); In re Air South Airlines, 

(Bankr.
D. S.C. 2000); In re Armstrong, 

(Bankr. E.D.
Ark. 1999), aff’d, 

(E.D. Ark. 2001); In re
Morren Meat & Poultry Co., 

(W.D. Mich. 1988)
(a first-time transaction can qualify as a payment in the
ordinary course of business) with In re Brown Transp.
Truckload, Inc., 

(Bankr. N.D. Ga. 1992)
(a first-time transaction is categorically ineligible for pro-
tection as payment in the ordinary course of business).
   Although a history of dealing between parties is cer-
tainly the strongest factor supporting a determination
that the business between a debtor and an alleged pre-
ference creditor is ordinary, we do not believe it is abso-
lutely necessary in every case. In some instances, and
this is one, the ordinary course of business may be estab-
lished by the terms of the parties’ agreement, until that
agreement is somehow or other modified by actual perfor-
mance. In the absence of modifying behavior, we see
no reason why we should not look to the terms of the
parties’ agreement in order to determine their ordinary
course of business. Indeed, as Judge Grant in the bank-
ruptcy court astutely observed here,
    the court can imagine little (short of the certain knowl-
    edge that its debt will not be paid) that would dis-
    courage a potential creditor from extending credit to
    a new customer in questionable financial circum-
    stances more than the knowledge that it would not
    even be able to raise the ordinary course of business
    defense, if it is subsequently sued to recover an al-
    leged preference.
No. 02-3127                                             9

  In our case, every RAL transaction, including the first-
time transactions between Household and a particular
debtor, was conducted in accordance with the terms of the
parties’ written agreement. This included the timing
and the manner in which Household applied the funds
from each debtor’s account to the loan balance. We con-
clude that all of the transactions were ordinary as be-
tween the parties. Because each transaction was ordinary
as between the parties and when measured against the
tax refund loan industry as a whole according to ordinary
business terms, they are insulated from avoidance under
§ 547(c)(2).
  In closing, we comment on one final matter. The trus-
tees argue that RALs actually involve two separate trans-
actions. In the first, the debtor applies for a tax refund
with the IRS. In the second, the bank makes a loan to
the debtor. Viewed this way, the trustees see a setoff.
The reality of a RAL process, however, is more accurately
viewed as a single integrated transaction. And because
Household held title to the funds when they were re-
ceived at the bank and then transferred to the bank’s own
account, the bank was not engaging in “setting off” a debt
owed by the debtor.
 The judgment of the district court is AFFIRMED.

A true Copy:
      Teste:

                       ________________________________
                       Clerk of the United States Court of
                         Appeals for the Seventh Circuit




                  USCA-02-C-0072—6-30-03