Filed: Feb. 20, 2019
Latest Update: Mar. 03, 2020
Summary: PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _ Nos. 17-3701 & 17-3823 _ In re: PAUL H. TITUS, Alleged Debtor ROBERT SHEARER, Trustee Appellant in No. 17-3823 v. PAUL H. TITUS; BONNIE TITUS, Appellants in No. 17-3701 _ Appeal from the United States District Court for the Western District of Pennsylvania (D.C. Civil Action No. 2-17-cv-00479/548) District Judge: Honorable Joy Flowers Conti _ Argued January 7, 2019 Before: AMBRO, SHWARTZ, and FUENTES, Circuit Judges (Opinion f
Summary: PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _ Nos. 17-3701 & 17-3823 _ In re: PAUL H. TITUS, Alleged Debtor ROBERT SHEARER, Trustee Appellant in No. 17-3823 v. PAUL H. TITUS; BONNIE TITUS, Appellants in No. 17-3701 _ Appeal from the United States District Court for the Western District of Pennsylvania (D.C. Civil Action No. 2-17-cv-00479/548) District Judge: Honorable Joy Flowers Conti _ Argued January 7, 2019 Before: AMBRO, SHWARTZ, and FUENTES, Circuit Judges (Opinion fi..
More
PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
________________
Nos. 17-3701 & 17-3823
________________
In re: PAUL H. TITUS, Alleged Debtor
ROBERT SHEARER, Trustee
Appellant in No. 17-3823
v.
PAUL H. TITUS; BONNIE TITUS,
Appellants in No. 17-3701
________________
Appeal from the United States District Court
for the Western District of Pennsylvania
(D.C. Civil Action No. 2-17-cv-00479/548)
District Judge: Honorable Joy Flowers Conti
________________
Argued January 7, 2019
Before: AMBRO, SHWARTZ,
and FUENTES, Circuit Judges
(Opinion filed February 20, 2019)
Douglas A. Campbell (Argued)
Kathryn L. Harrison
Campbell & Levine
310 Grant Street
1700 Grant Building
Pittsburgh, PA 15219
Counsel for Appellants/Cross-Appellees
Neal H. Levin (Argued)
Freeborn & Peters
311 South Wacker Drive
Suite 3000
Chicago, IL 60606
Counsel for Appellee/Cross-Appellant
________________
OPINION OF THE COURT
________________
AMBRO, Circuit Judge
When his old law firm broke its lease, attorney Paul
Titus was on the hook for millions of dollars in unpaid
commercial rent. The landlord tried to recover the rent by
targeting the wages Mr. Titus was earning at his new firm. But
Mr. Titus’s wages never passed through his hands alone;
instead, they were deposited by his new firm directly into a
bank account owned by both Mr. Titus and his wife as tenants
by the entireties.
2
Eventually, Mr. Titus was forced into bankruptcy and
the landlord’s claim became a claim of the bankruptcy trustee.
Now, after two trials in the Bankruptcy Court and two appeals
to the District Court, we reach three conclusions. First, Mr.
and Mrs. Titus are liable for a fraudulent transfer. When the
wages of an insolvent spouse are deposited into a couple’s
entireties account, both spouses are fraudulent transferees.
Second, as for the precise measure of the Tituses’ liability, the
bankruptcy trustee waived any challenge to the method used
by previous courts to calculate fraudulent-transfer liability.
Going forward, however, we clarify how future courts should
measure liability when faced with an entireties account like the
Tituses’ — an account into which deposits consist of both
(fraudulent) wages and (non-fraudulent) other sources, and
from which cash is spent on both (permissible) household
necessities and (impermissible) other expenditures.1 Until
now, a trustee somehow had to show that wage deposits were
impermissibly spent on non-necessary expenditures, even
though wage and nonwage deposits had become commingled
in the account. Rather than expect a trustee to trace the
untraceable, future courts should generally presume that wage
deposits were spent on non-necessary expenditures in
proportion to the overall share of wages in the account as a
whole. Third, in evaluating the Bankruptcy Court’s
1
Judge Shwartz joins the opinion in all respects except
its discussion of the pro rata approach because, among other
things, the panel has unanimously concluded that the trustee
waived his challenge to the method of calculating the Tituses’
liability, and thus it is unnecessary to discuss the pro rata
approach. Judge Shwartz is also of the view that the method
for calculating the amount of fraudulent-transfer liability
should be left to the discretion of the trial judge based upon the
evidence provided. Cf. In re Gen. Motors Corp. Pick-Up Truck
Fuel Tank Prods. Liab. Litig.,
55 F.3d 768, 821 (3d Cir. 1995).
3
application of the method in play at the time of its decision, we
perceive no clear error. Thus we affirm.
I. Background
A. Facts
In 1999, the Pittsburgh law firm of Titus & McConomy
dissolved. One of the firm’s named partners, Paul Titus, joined
another firm, Schnader Harrison Segal & Lewis LLP. The
Schnader firm began depositing Mr. Titus’s wages into a bank
account he owned jointly with his wife.
Evidently the dissolved Titus firm had walked away
from its commercial lease. To recover rent that had gone
unpaid since the dissolution, the landlord brought a breach-of-
contract suit against the former partners of the Titus firm and
ultimately secured a multimillion dollar judgment against the
partners, including Mr. Titus.
Armed with the breach-of-contract judgment, the
landlord set its sights on the wages that Mr. Titus’s new
employer, the Schnader firm, was depositing into the Tituses’
bank account. It brought a fraudulent-transfer action in
Pennsylvania state court against Mr. and Mrs. Titus. This
triggered an involuntary bankruptcy against Mr. Titus. Thus
the landlord’s fraudulent-transfer claim became a claim of the
bankruptcy trustee in the Bankruptcy Court.2
2
Several former partners of the Titus firm have faced the
same fate since the firm’s dissolution. See In re Wettach,
811
F.3d 99 (3d Cir. 2016); Cardiello v. Arbogast, 533 F. App’x
150 (3d Cir. 2013); Cohen v. Sikirica,
487 B.R. 615 (W.D. Pa.
2013); In re Oberdick,
490 B.R. 687 (Bankr. W.D. Pa. 2013).
4
B. Procedural History in Bankruptcy
After a first trial, the Bankruptcy Court concluded that
the direct deposit of wages into the Tituses’ bank account was
a fraudulent transfer that the trustee could recover from either
Mr. or Mrs. Titus, who jointly owned the account as tenants by
the entireties. As for the measure of liability, Mr. and Mrs.
Titus were liable for the amount of Mr. Titus’s wages that were
“not spent on necessities.” In re Titus (Titus I),
467 B.R. 592,
620 (Bankr. W.D. Pa. 2012).
On appeal, the District Court affirmed that the wage
deposits were a fraudulent transfer. It remanded for a new trial,
however, to give the Tituses a second chance to identify both
the source of certain “unexplained deposits” into the bank
account and the destination of certain “unknown expenditures”
from the account. Titus v. Shearer (Titus II),
498 B.R. 508,
520, 525 (W.D. Pa. 2013).
After a second trial, the Bankruptcy Court made the
following findings as to deposits into, and expenditures from,
the bank account:
DEPOSITS EXPENDITURES
Total Wages Necessities
$1,125,255.58 $1,134,000.67
Entireties Account
($91,272 preexisting)
Explained
Nonwages
$634,998.83 Non-Necessities
$1,000,133.51
Unexplained
Nonwages
$268,167.09
5
See In re Titus (Titus III),
566 B.R. 755, 797 (Bankr. W.D. Pa.
2017). A note on terminology: The Bankruptcy Court divided
the spending from the account into “necessity expenditures”
and “non-necessity expenditures,” which it sometimes called
“Non-Objectionable Expenditures” and “Objectionable
Expenditures,” respectively. See, e.g., id.; see also
id. at 765,
768, 777–78. For simplicity, we use the necessity/non-
necessity nomenclature. To give some content to these terms,
“necessities” included items such as a lawnmower and
batteries,
id. at 792–93, while the Tituses’ “non-necessities”
included, among other things, their grandson’s application fee
to Notre Dame,
id. at 797.
Using the figures set out above, the Bankruptcy Court
went about calculating the Tituses’ liability. But the Court
immediately hit a roadblock: Because money is fungible and
wage and nonwage deposits commingled in the account, it was
impossible to determine whether a dollar of wages was
eventually spent on a permissible “necessity” or an
impermissible “non-necessity.” As a result, the Court had to
calculate liability indirectly.
It did so using what it could measure: nonwage deposits
and non-necessity spending, which are represented below the
dotted line in the chart. The Court’s underlying assumption
was that all explained, nonwage sources of cash in the account
(both explained nonwage deposits and cash already in the
account) were spent on non-necessities before any wage
deposits were impermissibly spent on whatever non-
necessities remained. Thus the Tituses’ total liability was:
6
(Non-Necessities) – (Explained Nonwages) – (Initial Balance)
= $1,000,133.51 – $634,998.83 – $91,272.00
= $273,862.68.
Id. at 799. Over the Tituses’ objection, the Court did not offset
their non-necessity spending even further by the amount of
unexplained nonwage deposits. Had it done so, an additional
$268,167.09 would have been shaved off the remaining non-
necessity spending, leaving a judgment against the Tituses of
only $5,695.59.
The District Court affirmed. Neither side is fully
satisfied with various rulings of the Bankruptcy and District
Courts, and both have appealed.
II. Jurisdiction and Standard of Review
The District Court had jurisdiction to review the final
order of the Bankruptcy Court under 28 U.S.C. § 158(a). We
have jurisdiction to review the final order of the District Court
per 28 U.S.C. § 158(d) and 28 U.S.C. § 1291. Because the
District Court acted as an appellate court, we review its
determinations de novo. In re Bocchino,
794 F.3d 376, 379 (3d
Cir. 2015). We review the legal conclusions of the Bankruptcy
Court de novo and its factual determinations for clear error.
Id.
at 380.
III. Discussion
After two trials in the Bankruptcy Court, two appeals to
the District Court, and four rounds of briefing in our Court,
there are three issues for our review:
(1) Are the Tituses liable for a fraudulent transfer?
7
(2) Did the trustee waive any challenge to the method
used to calculate the Tituses’ liability?
(3) Did the Bankruptcy Court clearly err in applying the
method it used?
We include an additional area of discussion on the
second issue. Even if the trustee waived its challenge to the
calculation method, should future courts measure liability for
commingled accounts differently?
A. Threshold Fraudulent-Transfer Liability
The bankruptcy trustee “may avoid any transfer of an
interest of the debtor in property” if the transfer “is voidable
under applicable law by a creditor.” 11 U.S.C. § 544(b)(1).
The applicable law here, the Pennsylvania Uniform Fraudulent
Transfer Act (PUFTA), permits a creditor to avoid a fraudulent
transfer “to the extent necessary to satisfy the creditor’s claim.”
12 Pa. Cons. Stat. § 5107(a)(1).3
We reach three conclusions on the threshold question of
the Tituses’ fraudulent-transfer liability. First, the wage
deposits into the Tituses’ entireties account were a “transfer”
under the PUFTA. Second, Mrs. Titus is personally subject to
PUFTA liability as an entireties tenant. Third, Mr. Titus is
subject to transferee liability even though he is the debtor-
transferor as well. As a result, the wage deposits constituted a
fraudulent transfer that the bankruptcy trustee could avoid.
3
As of February 22, 2018, the PUFTA was renamed the
Pennsylvania Uniform Voidable Transactions Act. See 12 Pa.
Cons. Stat. § 5101(a).
8
1. The wage deposits constituted a “transfer.”
“[T]he direct deposit of wages into an entireties account
is a ‘transfer’ of an ‘asset’ under the PUFTA.”
Wettach, 811
F.3d at 115. This statement settles the question whether the
wage deposits from the Schnader firm into the Tituses’
entireties account were a transfer.
The reasoning behind this conclusion is as follows. On
a macro level, Mr. Titus’s wages (i) began as his “asset” for
purposes of the PUFTA and (ii) were not his “asset” once they
were in the entireties account. Id at 114–15. That change in
status is deemed a “transfer.”
Id. at 115.
As to the first point, Mr. Titus “exercised control over
where his employer deposited his wages.” See
id. at 114. This
control overrode Pennsylvania’s baseline rule that “wages” are
exempt from creditors “while in the hands of the employer.”
42 Pa. Cons. Stat. § 8127(a). In other words, when the
Schnader firm “initiated the direct deposit,” the wages left
Schnader’s “hands” and became, temporarily, an asset of Mr.
Titus. See
Wettach, 811 F.3d at 114–15.
As to the second point, the wages ceased being an
“asset” of Mr. Titus once they were in the entireties account.
The definition of “asset” under the PUFTA excludes property
held in a tenancy by the entireties “to the extent it is not subject
to process by a creditor holding a claim against only one
tenant.” 12 Pa. Cons. Stat. § 5101(b). That predicate is
satisfied here: The landlord held a claim against only one
entireties tenant (Mr. Titus), and “entireties property is
unavailable to satisfy the claims of the creditor of only one of
the tenants.” Garden State Standardbred Sales Co. v. Seese,
611 A.2d 1239, 1243 (Pa. Super. Ct. 1992). As a result, once
the wage deposits reached the entireties bank account, they
were no longer an “asset” of Mr. Titus under the PUFTA.
9
Putting these points together, Mr. Titus started with an
“asset” and later relinquished it to the entireties account. This
maneuver meets the PUFTA’s definition of “transfer” as an
“indirect . . . disposing of or parting with an asset or an interest
in an asset.” 12 Pa. Cons. Stat. § 5101(b). To sum up, Mr.
Titus, “by depositing [his] own funds into an entireties bank
account, is converting [the] funds into entireties property and,
thereby, transferring [that] property for purposes of the
[PUFTA].” See In re Meinen,
232 B.R. 827, 841 (Bankr. W.D.
Pa. 1999) (setting out the reasoning above).
2. Mrs. Titus is a transferee subject to PUFTA liability.
Even assuming that the wage deposits were a “transfer”
under the PUFTA, Mrs. Titus has a further objection of her
own — that her status as a co-tenant by the entireties cannot
open her up to personal liability for wages deposited by her
insolvent husband into their joint account.
Case law prevents this position’s success. “[W]hen a
spouse conveys individual property to a tenancy by the
entireties in fraud of creditors, the creditor may nevertheless
execute against the property so conveyed.” Garden
State, 611
A.2d at 1243; see also Stinner v. Stinner,
446 A.2d 651, 652
(Pa. Super. Ct. 1982) (same). Numerous courts have applied
this rule to hold an insolvent debtor’s spouse personally liable
for a fraudulent transfer. See, e.g.,
Wettach, 811 F.3d at 114–
15;
Meinen, 232 B.R. at 840–41; Garden
State, 611 A.2d at
1243;
Stinner, 446 A.2d at 652.
In doing so, courts have acknowledged the point that
Mrs. Titus urges here — that this liability rule leads to a harsh
result. See, e.g.,
Stinner, 446 A.2d at 654–55 (McEwen, J.,
concurring) (applying the rule “even though I find it most
difficult to accept the determination that the regular routine
deposit of salary and bonuses by a husband, into a household
10
checking account owned by himself and his wife as tenants by
the entireties, is a conveyance of individual property in fraud
of creditors”). One court even “invite[d] the General
Assembly” to scrap the usual creditor-friendly presumption of
a fraudulent transfer “when the conveyance results in the
creation of entireties property.” Garden
State, 611 A.2d at
1243–44. All courts nonetheless continue to apply the rule.
3. Mr. Titus is subject to PUFTA transferee liability
even though he is the debtor-transferor as well.
Mr. Titus is both transferor and transferee. As an
individual debtor-transferor, he is subject to liability under the
landlord’s claim for breach of the lease agreement. As a
transferee, he has fraudulent-transfer liability as a tenant of the
entireties account. In sum, his different capacities make him
liable in different ways. His argument that he cannot be both
transferor and transferee therefore fails.
Nor is there a risk of double recovery — that is, one
recovery from Mr. Titus as an individual debtor and another
from him as an entireties tenant. Once the trustee secures a
recovery from one source, he will not have recourse against the
other source. Cf. In re Integra Realty Res., Inc.,
354 F.3d 1246,
1266 (10th Cir. 2004) (noting that the trustee may only “restore
the estate to the financial condition it would have enjoyed if
the transfer had not occurred”) (quotations omitted). The
fraudulent-transfer suit was only instituted as part of the overall
collection effort to satisfy the judgment entered on the breach
of the lease agreement.
B. Method for Calculating PUFTA Liability
Having concluded that Mr. and Mrs. Titus can be
individually liable for a fraudulent transfer of Mr. Titus’s
wages, we reach the question of how to measure that liability.
11
We begin with the baseline rule that a transfer is not
“fraudulent” under the PUFTA if the wages deposited into the
entireties account are “used to pay for reasonable and
necessary household expenses.”
Wettach, 811 F.3d at 105
(quotations omitted) (discussing 12 Pa. Cons. Stat.
§ 5104(a)(2)); see also 11 U.S.C. § 550(a) (providing that the
trustee in a § 544 action “may recover” from a transferee either
“the property transferred” or “the value of such property,” but
not explaining how to identify the property or calculate its
value in the first place). This rule accounts for the Bankruptcy
Court’s decision, explained above, to segregate expenditures
from the account into “necessities” and “non-necessities.” But,
having determined the amounts of wage deposits, nonwage
deposits, necessity spending, and non-necessity spending, the
question remains how to combine these inputs to reach a figure
for the Tituses’ liability. This question, in turn, depends on the
allocation of burdens in a fraudulent-transfer action in
Pennsylvania.
Between the two trials in this case, we clarified the
burdens in a PUFTA action. First, our Court presumes that
“funds deposited into an entireties account were not in
exchange for reasonably equivalent value.”
Wettach, 811 F.3d
at 111. In our case, this means we presume that the wages in
question were not spent on necessities. See
id. Second, the
Tituses may rebut that presumption by producing “some
evidence as to uses of funds in the entireties account.”
Id. at
109 (quotations omitted). Imposing this burden of production
on the Tituses “serves an information-forcing purpose” by
requiring them “to come forward with information” about
“how they used funds transferred into [the] entireties account.”
Id. Third, once the Tituses have met their burden of
production, the trustee bears the burden of persuasion “as to all
elements of a constructive fraudulent-transfer claim under the
PUFTA.”
Id. at 107. Among other things, the trustee must
12
prove by a preponderance of the evidence that wage deposits
were not spent on necessities.
But the trustee is faced with what appears to be an
impossible task in a commingled account, circumstances that
did not exist in other cases before us. Because money is
fungible, and funds from multiple sources commingle in the
entireties account, “it may be impossible to determine what
deposit was used for a particular expenditure.” In re Wettach,
489 B.R. 496, 507 (Bankr. W.D. Pa. 2013); see also
Oberdick,
490 B.R. at 710 n.15 (describing the “impossible burden” on
the trustee to trace deposits into the account to spending from
the account).
No wonder, then, that the trustee here could not carry
his burden of persuasion. For example, in the first trial in this
case, the Bankruptcy Court explained that it was “at least as
likely as not” that a given dollar of deposits went toward
necessity spending as toward non-necessity spending. Titus
I,
467 B.R. at 624. As a result, the trustee “ha[d] not
preponderantly proven” that wages were improperly spent on
non-necessities.
Id. The uncertainty of the dollar’s final
destination, moreover, is compounded by a “fairness issue” —
namely, Mr. and Mrs. Titus “created the uncertainty by
commingling the funds,” but the trustee “is expected to
somehow unravel it.” See
Oberdick, 490 B.R. at 710 n.15.
Faced with this commingling problem, every court to
encounter the issue has adopted a baseline assumption: All
explained nonwage deposits were spent on non-necessities
before any wage deposits were spent on non-necessities. E.g.,
Titus v. Shearer (Titus IV), No. AP 10-2338,
2017 WL
5467712, at *5 (W.D. Pa. Nov. 14, 2017) (“The bankruptcy
court found . . . that all non-necessities were purchased with
explained nonwages.”). Another way of stating the same
assumption is that all wage deposits were spent on necessities
13
before being spent on anything else. Either phrasing expresses
the same idea: Deposited funds travel horizontally, rather than
diagonally, across the charts reproduced in this opinion. See
supra p. 5; infra p. 15. The assumption has supported a method
for calculating PUFTA liability, which parties and courts have
dubbed the “Non-Necessities Approach” or the “Other Deposit
Methodology.” It provides the following formula for
calculating fraudulent-transfer liability:
Liability = (Non-Necessity Spending) – (Nonwage Deposits).
In this case, the trustee waived any challenge to the
selection of this method and formula. An issue is waived on
remand if it was “not raised in a party’s prior appeal.”
Skretvedt v. E.I. DuPont De Nemours,
372 F.3d 193, 203 (3d
Cir. 2004). Also, “when a party fails to raise an issue in the
bankruptcy court, the issue is waived.” In re Kaiser Grp. Int’l
Inc.,
399 F.3d 558, 565 (3d Cir. 2005). The trustee here did
not object to the Non-Necessities Approach during the first
trial in the Bankruptcy Court or in the first appeal to the District
Court. Instead, he waited to challenge it for the first time on
remand to the Bankruptcy Court for a second trial. See Titus
III, 566 B.R. at 769. Thus the Bankruptcy Court and District
Court agreed that the choice of method had been waived. See,
e.g.,
id. at 773. We affirm that conclusion. Hence we turn to
its application in the next section.
Before doing so, however, we set out a different way to
calculate liability for future courts facing commingled funds:
the pro rata approach. Under this approach, we presume,
absent other evidence, that spending out of the entireties
account was made up of a mixture of wage and nonwage
deposits in proportion to the overall ratio of wage to nonwage
deposits in the account. As we explain below, this approach
addresses practically the commingling of fungible funds in the
account and is not foreclosed by precedent in our Circuit.
14
1. The pro rata approach accounts practically for the
commingling of fungible funds.
As noted, the first Bankruptcy Court in this case stated
that it was “at least as likely as not” that a dollar of nonwage
deposits funded non-necessity spending, and that the trustee
had therefore failed to prove by a preponderance of the
evidence that wage deposits impermissibly funded non-
necessities. See Titus
I, 467 B.R. at 624. This is incorrect.
Absent unusual circumstances, a dollar deposited into the
account is equally likely to be spent on necessities and on non-
necessities only if the same amount is spent on both necessities
and non-necessities. In all other situations, the deposited dollar
is more likely to be spent on whichever category of spending
(necessities or non-necessities) is larger. By the same token, a
dollar spent from the account is more likely to come from
whichever category of deposits (wage or nonwage) is larger.
The pro rata approach accounts for the fungibility of
wage and nonwage funds that are commingled in the entireties
account. In our case, the liability would be calculated based on
the inflows and outflows found by the second Bankruptcy
Court decision (for simplicity, and to be consistent with our
conclusion in the next section, we have eliminated
“unexplained” nonwage deposits):
DEPOSITS EXPENDITURES
Total Wages Necessities
$1,125,255.58 $1,134,000.67
Entireties Account
($91,272 preexisting)
Nonwage
Deposits Non-Necessities
$634,998.83 $1,000,133.51
15
Titus
III, 566 B.R. at 797. The total amount of money flowing
into the account is the sum of wage deposits, nonwage
deposits, and preexisting cash in the account:
Total Inflows = (Wages) + (Nonwages) + (Preexisting Cash)
= ($1,125,255.58) + ($634,998.83) + ($91,272)
= $1,851,526.41.
Thus the calculation of wage deposits as a percentage of total
inflows is:
(Wage Deposits) / (Total Inflows)
= ($1,125,255.58) / ($1,851,526.41)
= 60.8%.
Hence we can presume that, of the $1,000,133.51 spent on non-
necessities, 60.8% impermissibly came from wage deposits.
The Tituses’ liability would be that wage-derived portion:
(0.608) * ($1,000,133.51) = $607,825.96.
Eyeballing these figures, we note that this measure of
liability makes intuitive sense: Wages account for just under
two-thirds of all deposits into the account, so it stands to reason
that just under two-thirds of all non-necessity spending came
from wage deposits. Appropriately, then, the Tituses’ liability
under the pro rata approach would be just under two-thirds of
all non-necessity spending.
We add one further note on the mechanics of the pro
rata approach. We alluded earlier to “unusual circumstances”
that could overcome the default presumption that spending out
of the entireties account is made up of a mixture of wage and
16
nonwage dollars in proportion to the overall ratio of wage to
nonwage deposits into the account. See supra p. 15. Recall
that the pro rata approach rests on our observation that a
trustee should not be asked to trace the untraceable. It follows
that the presumption would yield where a factfinder could trace
the (ordinarily) untraceable — in other words, where the
factfinder could track a dollar from a given category of deposits
into a given category of spending.
Say, for instance, that a trial court could trace X dollars
of nonwage deposits into an account to X dollars of non-
necessity spending from the account. (An example might be
monies placed into the account from a bequest requiring its
spending on what is not necessary.) Before performing the pro
rata calculation for the rest of the cash inflows and outflows,
the trial court would reduce both its nonwage deposit figure
and its non-necessity spending figure by X. The result of this
reduction in nonwage deposits (while wage deposits remain
constant) would be a greater percentage share of all deposits
into the account coming from wages. This greater percentage
share would then be applied to the reduced amount of total non-
necessity spending.
In sum, the trial court should trace whatever is traceable
before using the pro rata approach to proportionally derive the
untraceable flows. We leave to the trial court’s discretion the
threshold decision whether it is able to trace the ordinarily
untraceable.
2. The pro rata approach is not foreclosed by
precedent in our Circuit.
The only possible precedent of this Court, Wettach,
811
F.3d 99, does not foreclose the pro rata approach. In the
course of setting out the rebuttable presumption and the
burdens of production and persuasion, the Wettach Court noted
17
that the Bankruptcy Court had applied the Non-Necessities
Approach, which we abandon today for certain situations in
which an account is commingled. We explained that Mr. and
Mrs. Wettach, who bore the burden of production,
produced no evidence to demonstrate how they
spent the wages deposited into the entireties
account. The bankruptcy court even offered
them a “dollar-for-dollar reduction against any
liability” for other deposits into the account. . . .
Having failed to carry their burden of production
and absent clear error by the bankruptcy court,
the Wettachs have no claim for relief on appeal.
Wettach, 811 F.3d at 111 (quoting
Wettach, 489 B.R. at 507)
(emphasis added). The “dollar-for-dollar reduction” for
nonwage deposits refers to the formula dictated by the Non-
Necessities Approach (Liability = Non-Necessity Spending –
Nonwage Deposits).
For three reasons, however, this statement in Wettach
does not mandate the Non-Necessities Approach for accounts
that commingle wage and nonwage deposits.
First, the passage is not a ringing endorsement of the
approach. We merely noted that the Bankruptcy Court “even
offered” the Wettachs another potential offset. The statement
is hardly a holding.
Second, even if the statement were taken as a holding,
Wettach is distinguishable. It confronted a mixture of deposits
much simpler than those facing us here: The Wettachs had
produced “no evidence of any ‘other deposits.’” See Titus
III,
566 B.R. at 772 (discussing Wettach). In other words, “the
only deposits that proved to be at issue in Wettach were the
wage deposits of the debtor.”
Id. As a result, the Wettach
18
Court had no need to “directly address” the choice of method
for measuring liability.
Id. When the only source of funding
into an account is wage deposits, there is no mystery about the
source of a dollar spent on non-necessities, and thus no reason
to consider a way of measuring liability like the pro rata
approach. Hence Wettach does not bind us now.
Third, even the Wettach Bankruptcy Court expressed
“some reservations” about the dollar-for-dollar offset dictated
by the Non-Necessities Approach.
Wettach, 489 B.R. at 507
(Agresti, J.). Bankruptcy Judge Agresti applied the approach
only because a prior District Court decision had already
approved it. See
id. at 507–08 (citing Cohen,
487 B.R. 615).
In yet another case arising out of the dissolution of the Titus
firm, moreover, Judge Agresti had sounded the alarm about the
dollar-for-dollar offset. See
Oberdick, 490 B.R. at 710 n.15
(“The Court does have some concerns about the [Non-
Necessities Approach] as to this point.”). Were he “writing on
a blank slate,” he explained he would “give serious
consideration” to employing “a presumption” in which “a pro
rata share of the non-necessary expenditures could be deemed
to have come from funds originating from the other deposits.”
Id. (emphasis added). Now faced with such a blank slate, we
follow Judge Agresti’s prescient thinking.
* * *
In sum, the trustee missed his chance to challenge the
use of the Non-Necessities Approach in the first appeal to the
District Court. Hence his objection to the method for
measuring liability has been waived. Going forward, however,
courts faced with the situation here — in which wages and
nonwages are commingled in a single account and are
subsequently spent on both necessities and non-necessities —
should apply a pro rata approach. They should presume,
absent other evidence, that any spending out of the account was
19
made up of a mixture of wage and nonwage dollars in
proportion to the overall ratio of wage to nonwage deposits in
the account.
C. Application of Non-Necessities Approach
Because the trustee waived his argument that another
method should apply in calculating liability, we turn to the
parties’ dispute over the application of the Non-Necessities
Approach. To repeat, the approach proceeds from the
assumption that all nonwage deposits into the account were
spent on non-necessities before any wage deposits were
impermissibly spent on non-necessities. This, in turn, informs
its formula for fraudulent-transfer liability:
Liability = (Non-Necessity Spending) – (Nonwage Deposits).
The final skirmish in the case centers on the last term in
the formula: nonwage deposits. In the Bankruptcy Court, the
parties stipulated that nonwage deposits could be divided into
$634,998.83 of explained nonwage deposits and $268,167.09
of unexplained nonwage deposits. See Titus
III, 566 B.R. at
798 (“The Parties agree that [the $268,167.09] were not wage
deposits, but they also agree that other than being non-wage
the source or sources of such deposits are unexplained.”). In
applying the Non-Necessities Approach, the Bankruptcy Court
counted only explained nonwage deposits toward the
“nonwage deposits” term in the formula. Had the Court also
counted the $268,167.09 of stipulated unexplained nonwage
deposits, the Tituses’ ultimate liability would have been
reduced from $273,862.68 to $5,695.59. Thus the question
presented in this section is whether the parties’ joint stipulation
that the unexplained deposits were not wages was sufficient to
meet the Tituses’ burden of production as to nonwage deposits.
20
To begin, we review the Bankruptcy Court’s decision
on this point for clear error. See
Wettach, 811 F.3d at 111. The
Court did not commit clear error in the face of the “bright line
rule” established by the District Court that “unexplained
deposits may not be used to set off liability.” Titus
III, 566
B.R. at 768 (“[T]he District Court . . . does not provide for any
discretion in the matter — it provides there shall be no
reduction for unexplained deposits.”). In remanding the case
for a second trial, the District Court made clear that the Tituses
would need to “produce evidence about the source of the
unknown funds” in order to offset their liability. Titus
II, 498
B.R. at 521.
Aside from expecting the Tituses to follow this
straightforward directive, there are at least three compelling
reasons to apply a bright-line rule in situations like this. First,
allowing an offset for unexplained deposits would
“incentivize” debtors “not to come forward with any
information that they had regarding the source of those funds,”
Cohen, 487 B.R. at 625, even though the debtor is “certainly in
a better position than the [t]rustee to determine the source of
the unexplained deposits,” Titus IV,
2017 WL 5467712, at *6.
Second, allowing an offset for unexplained deposits would
allow debtors “to avoid judgment . . . merely by having funds
deposited into the account that could not be traced.”
Cohen,
487 B.R. at 625. Third, even a “nonwage” unexplained deposit
could be a fraudulent transfer. “For example, if Paul Titus
individually owned a rare painting, sold it for $268,167.09, and
deposited the proceeds into the entireties checking account,
those funds could not be used to offset objectionable
expenses.” Titus IV,
2017 WL 5467712, at *6; see also Titus
III, 566 B.R. at 768 n.8.
In sum, we affirm the decisions of the Bankruptcy Court
and District Court not to allow unexplained nonwage deposits
to offset the Tituses’ liability.
21
IV. Conclusion
To recap, we reach three conclusions on the path to
affirming the judgment of the District Court. First, fraudulent-
transfer liability attaches to both Mr. and Mrs. Titus for the
deposit of Mr. Titus’s wages from his law firm directly into the
Tituses’ entireties bank account. The wage deposits into the
account constituted a “transfer” under the PUFTA. Mrs. Titus
is personally subject to fraudulent-transfer liability as a joint
owner of the account. And Mr. Titus is subject to transferee
liability even though he is the debtor-transferor as well. As a
result, the wage deposits were a fraudulent transfer that the
bankruptcy trustee could avo
id.
Second, the trustee waived any objection to the
Bankruptcy Court’s chosen method to calculate the Tituses’
liability. The Court followed the so-called Non-Necessities
Approach, which holds that fraudulent-transfer liability is non-
necessity spending less nonwage deposits, and the trustee did
not challenge the approach in his first appeal to the District
Court. Going forward for commingled accounts, however, the
Non-Necessities Approach rests on an unreasonable
expectation that a trustee can show by a preponderance of the
evidence that a dollar of wages was impermissibly spent on a
non-necessity. When deposits from different sources are
commingled in an account, the Non-Necessities Approach
almost always forces a trustee to explain the unexplainable.
Absent other evidence, future courts instead should presume
that any spending out of an entireties account is made up of a
mixture of wage and nonwage dollars in proportion to the
overall ratio of wage to nonwage deposits in the account. This
pro rata approach accounts practically for the commingling of
fungible funds and is not foreclosed by precedent in our
Circuit.
22
Third, the Bankruptcy Court did not clearly err in its
application of the Non-Necessities Approach. The District
Court had set out a simple rule that the Tituses had to explain
the source of their deposits into the account. Despite the
parties’ stipulation that certain unexplained deposits were not
wages, the Bankruptcy Court did not clearly err in refusing to
offset the Tituses’ liability by the amount of those unknown
deposits.
Thus we affirm.
23