IN THE UNITED STATES COURT OF FEDERAL CLAIMS
)
CAPSTONE ASSOCIATED SERVICES, )
LTD., )
)
Plaintiff, ) No. 21-913
)
v. ) Filed: September 25, 2026
)
THE UNITED STATES, )
)
Defendant. )
___________________________________ )
OPINION AND ORDER
Plaintiff Capstone Associated Services, Ltd. (“Capstone”), by and through Capstone
Holdings Corp., Tax Matters Partner, contests the Internal Revenue Service’s (“IRS”) Notice of
Final Partnership Administrative Adjustment (“FPAA”) dated November 13, 2020, which
disallowed $239,773 in insurance premium expense deductions claimed by Capstone for the 2016
tax year and imposed accuracy related penalties amounting to 20 percent of the alleged
underpayment. Before the Court is Capstone’s Motion for Summary Judgment. Capstone argues
that the undisputed facts establish that the payments constitute insurance premiums, entitling
Capstone to judgment as a matter of law. The Government contends that genuine issues of material
fact, as well as questions of evidentiary weight and credibility, preclude summary judgment. For
the following reasons, the Court DENIES Capstone’s Motion for Summary Judgment.
I. BACKGROUND
A. Captive Insurance Transactions Generally
Businesses generally may deduct insurance premiums from their taxable income as
ordinary and necessary business expenses. See I.R.C. § 162(a); Treas. Reg. § 1.162-1(a).
Businesses may not, however, deduct money set aside in reserves to pay for a business’s expected losses—i.e., self-insurance. See, e.g., Clougherty Packing Co. v. Comm’r of Internal Revenue,
811 F.2d 1297, 1300 (9th Cir. 1987). Where the insurer and the insured are affiliated entities,
often referred to as a “captive insurance” arrangement, id. at 1298 n.1, the concern is that the
taxpayer may be effectively engaging in “self-insurance” as opposed to paying insurance
premiums, id. at 1300. On the other side of the transaction, the Internal Revenue Code generally
taxes insurance companies on their income like other corporations. See I.R.C. §§ 11, 831(a).
However, insurance companies (including captive insurers) below a certain size can alternatively
elect to exclude a certain amount of premiums received from their taxable income. Id. § 831(b).
In 2016, the relevant tax year in this case, the maximum premium limit was $1.2 million. 1 Id.
Under § 831(b), these small insurance companies instead pay taxes only on their investment
earnings. United States v. Dep’t of Ins., 66 F.4th 114, 118 & n.4 (3d Cir. 2023).
While “captive” insurance generally refers to transactions where the insurer and insured
are affiliated entities, Avrahami v. Comm’r of Internal Revenue, 149 T.C. 144, 176 (2017),
“microcaptive” insurance specifically refers to arrangements where small insurance companies
provide coverage to an affiliated entity, or group of related companies, while making an I.R.C.
§ 831(b) election, see id. at 179 (“Combining these two concepts—captive insurance companies
and the section 831(b) tax advantages for small insurance companies—yields the
‘microcaptive.’”). 2 Given the affiliation between the insured and captive insurer and the tax
benefits that accrue to both entities, microcaptive insurance arrangements can pose a heightened
1
For taxable years beginning after December 31, 2016, Congress raised the premium limit to $2.2 million and added diversification requirements. See Consolidated Appropriations Act, 2016, Pub. L. No. 114-113, § 333, 129 Stat. 2242, 3106–08 (2015). These changes are irrelevant to the 2016 tax year at issue in this case.
2
Although the instant case involves a microcaptive arrangement, this opinion refers generally to captive and microcaptive insurance and insurers interchangeably.
2
risk of tax abuse. See id. at 179–80 (describing ways in which such arrangements may operate to
generate “tax deductions while arguably moving money from one pocket to another”); CIC Servs.,
LLC v. Internal Revenue Serv., 593 U.S. 209, 213 (2021) (describing microcaptive insurance
arrangements’ “potential for tax evasion”). Despite the IRS’s concern for potential abuse, where
the microcaptive transaction constitutes real insurance, rather than a loss reserve, the taxpayer is
entitled to the deduction Congress elected to provide under the Code. See CIC Servs., 593 U.S. at
213 (noting that “[t]he result is that the money does not get taxed at all,” and “for better or worse,
[this] is a congressional choice”).
B. Capstone’s Purported Premiums and FinServ and PoolRe’s Purported Insurance
Coverage
The key issue in this case is whether Capstone properly deducted from its taxable income
insurance premium payments it made to FinServ Casualty Corporation (“FinServ”) and PoolRe
Insurance Corporation (“PoolRe”). The answer to that question depends on whether those
payments constituted “insurance” premiums that entitle Capstone to the tax benefits it claimed.
See Pl.’s Mem. of L. in Supp. of Mot. for Summ. J. at 8, ECF No. 181; Def.’s Resp. in Opp’n to
Pl.’s Mot. for Summ. J. at 5–6, ECF No. 188.
Capstone, at all relevant times, was owned 49.5 percent by Stewart Feldman, 49.5 percent
by Mr. Feldman’s wife, Dr. Marla B. Matz, and 1 percent by Capstone Holdings Corp. See Decl.
of Stewart A. Feldman ¶ 15, ECF No. 180-3. Capstone Holdings Corp., at all relevant times, was
owned 50 percent by Mr. Feldman and 50 percent by Dr. Matz. Id. ¶ 18. In 2016, FinServ, which
was domiciled in and regulated by Anguilla, British West Indies, was owned 61 percent by Dr.
Matz and 39 percent by a trust of which Dr. Matz was the trustee. Id. ¶¶ 54, 58; Pl.’s Exs. at 276–
3
77, ECF No. 182-3 (FinServ’s register of shareholders); 3 ECF No. 182-4 at 118 (FinServ’s
Anguilla insurance license). Thus, Capstone and FinServ were corporate siblings—i.e., they
shared a beneficial owner. 4
In 2016, PoolRe was also domiciled in and regulated by Anguilla and was owned by a third
party, Stephen Friedman. ECF No. 180-3 ¶ 82 (Feldman Declaration describing lack of common
ownership between Capstone and PoolRe); id.¶ 79 (describing PoolRe redomiciling to Anguilla in
2009); ECF No. 182-5 at 16 (PoolRe’s Anguilla insurance license). But other evidence indicates
a potential relationship, though not formal ownership, between PoolRe and FinServ as well as
PoolRe and Capstone. For example, Mr. Freidman pledged as collateral on a loan from FinServ
to Mr. Friedman “[a]ll shares of common stock of PoolRe.” ECF No. 190-5 at 16. And PoolRe’s
Conflict of Interest Policy identified PoolRe as “a wholly owned subsidiary of Capstone.” ECF
No. 190-1 at 3. Moreover, one of PoolRe’s directors, Robert L. Snyder II, testified at deposition
that he provided overlapping and simultaneous consulting services to Capstone during his time as
a PoolRe director. Dep. Tr. of Robert Snyder at 11:3–12:25, ECF No. 189-20. Capstone’s
Summary of Third-Party Reinsurance Programs similarly described PoolRe as a “Capstone3
Each party filed an appendix of exhibits in support of their respective briefs. Specifically, Capstone filed a Statement of Undisputed Material Facts and Appendix of Exhibits. See ECF Nos. 180-2, 182–84. After the Court granted Plaintiff’s Motion for Leave to File Exhibits Out of Time and to File Replacement Statement of Facts in Support of its Motion for Summary Judgment, see ECF No. 185, Capstone filed a replacement version of its Statement of Undisputed Material Facts. See ECF No. 187-1. The Government filed a Response in Opposition to Plaintiff’s Statement of Undisputed Facts and Exhibit Index. See ECF Nos. 189–99. After the Court granted the Government’s Motion to File Corrected Exhibits, see ECF No. 200, the Government filed corrected versions of Exhibits 52, 160, and 166–67. See ECF Nos. 201–03. The Court refers to each party’s exhibits by their respective ECF numbers and associated pagination rather than the bates-labeled page numbers.
4
In 2017, FinServ was merged into A.M.Y. Property & Casualty Insurance Corporation. ECF No. 180-3 ¶ 127.
4
administered Risk Pool” that “[l]imit[s] [its] pool participants to Capstone-administered
companies.” ECF No. 190-3 at 2.
Capstone contends that, in 2016, it paid $147,595 in direct premiums to FinServ for 10
insurance policies. ECF No. 180-3 ¶ 96 (Feldman Declaration describing the purported premiums
and policies); see, e.g., ECF No. 184-3 at 29 (commercial crime policy). For nine of these policies,
FinServ was liable for losses below the attachment point (i.e., “primary coverage”), which was 35
percent of the total direct written premiums and pooled premiums received by FinServ for the year.
ECF No. 180-3 ¶¶ 104, 107 (Feldman Declaration describing FinServ’s liability for primary
coverage below the attachment point, which was 35 percent of FinServ’s premiums); see also ECF
No. 184-3 at 62 (Joint Underwriting Stop Loss Endorsement for the nine policies describing the
attachment threshold as 35 percent of the Subject Premiums).
Losses exceeding the primary coverage (i.e., “excess coverage” or “stop-loss coverage”)
were split between FinServ, which covered 20 percent, and PoolRe, which covered 80 percent.
ECF No. 180-3 ¶¶ 105, 107–08, 115 (Feldman Declaration describing the stop-loss coverage
PoolRe provided to Capstone and FinServ); see also ECF No. 184-3 at 62 (Joint Underwriting
Stop Loss Endorsement for the nine policies describing PoolRe’s coverage for 80 percent of losses
exceeding the attachment threshold subject to a ceiling of 160 percent of the Subject Premiums).
PoolRe then ceded its 80 percent share to a quota-share risk pool of 79 participating insurance
companies, including FinServ. ECF No. 180-3 ¶¶ 115–17 (Feldman Declaration describing the
quota-share risk pool); ECF No. 183-1 at 13–14 (2016 Quota Share Reinsurance Policy’s schedule
of reinsurers and respective quota shares). Capstone also claims that it paid $92,178 in insurance
premiums directly to PoolRe for stop-loss coverage on these nine policies. ECF No. 180-3 ¶ 139
5
(Feldman Declaration describing Capstone’s payment of purported premiums for PoolRe’s excess
coverage).
FinServ maintained three lines of insurance in 2016, receiving: (1) direct written premiums
from policies issued to 10 of its affiliated insureds (including Capstone); (2) payments from
PoolRe for risk that PoolRe ceded to FinServ through FinServ’s participation in PoolRe’s quotashare risk pool; and (3) reinsurance premiums for its participation in vehicle-service contracts
ceded from Wesco Insurance Company (“Wesco”) through Aria (SAC), Ltd. (“Aria”) and PoolRe.
Id. ¶ 94 (Feldman Declaration describing FinServ’s 2016 insurance policies and premiums); ECF
No. 182-4 at 198 (FinServ’s 2016 income statement indicating FinServ’s revenues consisted of
direct written premiums, pooled reinsurance, other reinsurance programs, and investment income).
In 2016, PoolRe’s business consisted of issuing stop-loss insurance policies, administering
a quota-share insurance program, and other insurance activities such as the Wesco reinsurance
contracts. ECF No. 180-3 ¶¶ 129–31 (Feldman Declaration describing PoolRe’s 2016 premiums);
ECF No. 184-10 at 502 (PoolRe’s 2016 income statement indicating PoolRe’s revenues consisted
of direct written premiums, reinsurance, insurance services income, and investment income).
Through the stop-loss policies, PoolRe insured other insurers for losses exceeding their primary
coverage. ECF No. 180-3 ¶ 130 (Feldman Declaration describing PoolRe’s stop-loss coverage);
ECF No. 184-3 at 62 (Joint Underwriting Stop Loss Endorsement between FinServ and PoolRe
describing PoolRe’s stop-loss coverage). PoolRe then ceded this liability to the other insurers
participating in the quota-share risk pool. ECF No. 180-3 ¶¶ 110, 115–17, 130 (Feldman
Declaration describing PoolRe’s quota-share risk pool); ECF No. 184-3 at 64 (Joint Underwriting
Stop Loss Endorsement explaining that PoolRe, as the Stop Loss Insurer, “redistributed” its 80
percent of liability for the excess coverage “among the various participating captive reinsurers”).
6
Each insurer assumed a quota-share percentage of the ceded total risk of the combined pool. ECF
No. 183-1 at 13–14 (2016 Quota Share Reinsurance Policy’s schedule of reinsurers and respective
quota shares). One of the insurers in the quota-share risk pool was FinServ. See id. at 13 (2016
Quota Share Reinsurance Policy’s schedule listing FinServ’s quota share as 1.3484 percent).
The Feldman Law Firm LLP (“Feldman Law Firm”) acted as the “Servicing Agent” of
Capstone and other entities with “overlapping ownership” (the “Operating Entities”) under a
Shared Services Agreement (“SSA”) dated November 2, 1998. ECF No. 184-9 at 44. Mr. Feldman
is the managing partner of Feldman Law Firm, which is directly or indirectly owned by Mr.
Feldman and Dr. Matz. 5 See id. at 52 (SSA signed by Mr. Feldman as “Managing Partner” of
Feldman Law Firm); ECF No. 180-3 ¶¶ 1, 99 (Feldman Declaration indicating that, since the early
1990s, he has been a partner of Feldman Law Firm, which is a part of Feldman / Matz Interests,
LLP, and owned directly or indirectly by Feldman and Matz). The SSA explained that the
Operating Entities “wish[ed] to minimize their respective staffing and to better the quality of [their]
services by contracting with Servicing Agent for the performance of administrative duties for the
operation of each Operating Entity.” ECF No. 184-9 at 44 see also Decl. of Paul K. Franks ¶¶ 15–
21, ECF No. 180-1 (Feldman Law Firm’s controller explaining the operation of the SSA).
Pursuant to the SSA, Feldman Law Firm was authorized to issue payments on behalf of Capstone
in connection with the administrative services provided, like the premium payments to FinServ
and PoolRe at issue here. See ECF No. 184-9 at 45–46 (SSA indicating that Feldman Law Firm
“may act as the disbursing agent on behalf of the respective Operating Entity” in connection with
5
As of the date of the SSA, Feldman Law Firm was known as Stewart A. Feldman & Associates, L.L.P. See ECF No. 184-9 at 53 (First Addendum to the SSA dated November 3, 2004, indicating that Stewart A. Feldman & Associates, L.L.P. was at that time known as The Feldman Law Firm LLP). For ease of reference, the Court refers to Feldman Law Firm by its current name.
7
the shared services, including “[a]rranging to have insurance coverage”); ECF No. 180-1 ¶¶ 35–
37 (Franks Declaration explaining that Feldman Law Firm paid the $239,733 in premiums to
FinServ and PoolRe on Capstone’s behalf).
During the relevant time, Capstone served closely held businesses and their owners in the
establishment and administration of captive insurance companies. See Dep. Tr. of Jeffrey Carlson
at 16:8–14, ECF No. 182-1 (Capstone employee Jeffrey Carlson testifying at deposition that “the
primary services provided by Capstone . . . were to establish and administer captive insurance
companies”); ECF No. 180-3 ¶ 13 (Feldman Declaration explaining that “Capstone[’s] focus
included providing services to closely held businesses and their owners in the formation,
administration, and general implementation of captive insurance arrangements”). FinServ and
PoolRe are both Capstone-managed insurers. See ECF No. 180-3 ¶ 65 (Feldman Declaration
indicating that Capstone was an “administrative co-manager” of FinServ); id. ¶ 77 (indicating that
“Capstone served as PoolRe’s statutory insurance manager for all relevant times”); ECF No. 189-23 at 2 (Capstone Services Agreement “formaliz[ing]” PoolRe, Capstone, and Feldman Law
Firm’s “prior oral agreements concerning turnkey administrative services arrangements”). All of
FinServ’s direct insureds were corporate siblings of Capstone. ECF No. 180-3 ¶ 99 (Feldman
Declaration indicating that, “[i]n addition to Capstone, FinServ also insured” nine other named
insureds, “which were all owned directly or indirectly by Dr. Matz, [Mr. Feldman,] or both”); see,
e.g., ECF No. 184-3 at 29 (listing named insureds on FinServ’s commercial crime insurance
policy). The other insurers participating in PoolRe’s quota-share risk pool were all Capstonemanaged. See ECF No. 190-3 at 2 (Capstone’s Summary of Third-Party Reinsurance Programs
explaining that PoolRe “[l]imit[s] [its] pool participants to Capstone-administered companies”);
ECF No. 180-3 ¶ 133 (Feldman Declaration explaining that “[a]s the manager, Capstone
8
employed” underwriters, insurance brokers, and actuaries “to support the pool, the Wesco
program, and the 79 participating reinsurers”).
C. Procedural History
Capstone filed its Complaint on February 11, 2021. ECF No. 1. On October 14, 2021,
after the Government filed its Answer but before discovery had begun, Capstone filed a Motion
for Summary Judgment. ECF No. 22. On November 10, 2021, the Government filed a Motion
for Relief Under Rule 56(d) of the Rules of the United States Court of Federal Claims (“RCFC”)
as to Plaintiff’s Motion for Summary Judgment. ECF No. 24. The Court granted the
Government’s motion and denied Capstone’s prior summary judgment motion without prejudice
on February 1, 2022, emphasizing that this suit is a de novo proceeding involving a fact-intensive
inquiry. Capstone Associated Servs. v. United States, No. 21-913, 2022 WL 601379, at *2–3 (Fed.
Cl. Feb. 1, 2022). The parties then proceeded to discovery.
On December 15, 2025, after the completion of discovery, Capstone filed its pending
Motion for Summary Judgment. ECF No. 180. Capstone argues that it has established as a matter
of law that the captive insurance transactions at issue constituted valid insurance under the
appropriate legal test because: (1) FinServ and PoolRe were separate corporations rather than
shams, ECF No. 181 at 38–41; (2) the transactions between FinServ, Capstone, and PoolRe
transferred Capstone’s risk, id. at 41–42; and (3) FinServ and PoolRe distributed risk, id. at 42–
44. The Government filed its Response on February 3, 2026, arguing that: (1) FinServ lacked the
unrelated risks required for risk-shifting, ECF No. 188 at 20–24; (2) none of FinServ’s alleged risk
distribution strategies achieved their goals, id. at 24–29; (3) PoolRe did not face actual and
insurable risk, id. at 29–33; (4) FinServ is not insurance in the commonly accepted sense, id. at
34–39; and (5) several issues of evidentiary weight, including issues of materiality and credibility,
preclude summary judgment, id. at 39–42. Capstone filed its Reply in Support of its Motion for
9
Summary Judgment on March 2, 2026, arguing that the Government’s Response failed to establish
a genuine dispute of material fact under the applicable legal test. See ECF No. 205 at 17–25. The
Court heard oral argument on August 11, 2026. The motion is ripe for disposition.
II. LEGAL STANDARDS
A. Motions for Summary Judgment
Summary judgment is warranted when “the movant shows that there is no genuine dispute
as to any material fact and the movant is entitled to judgment as a matter of law.” RCFC 56(a). A
dispute is genuine when a finder of fact may reasonably resolve a factual issue in favor of either
party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 250 (1986). And a fact is material when it
“might affect the outcome of the suit” under the substantive law at issue. Id. at 248. “When
deciding a summary judgment motion, all of the nonmovant’s evidence is to be credited, and all
justifiable inferences drawn in the nonmovant’s favor.” Ruiz v. A.B. Chance Co., 234 F.3d 654,
671 (Fed. Cir. 2000) (citing Anderson, 477 U.S. at 255); see also Dairyland Power Co-op. v.
United States, 16 F.3d 1197, 1202 (Fed. Cir. 1994) (similar). The Court need not limit its review
to materials cited by the moving party. See RCFC 56(c)(3). Instead, the Court “has discretion to
consider other materials in the record when making its determination.” 10A Charles Alan Wright
& Arthur R. Miller, Federal Practice and Procedure § 2721 (4th ed.).
“The moving party carries the burden of establishing its entitlement to summary
judgment.” Quapaw Tribe of Okla. v. United States, 120 Fed. Cl. 612, 615 (2015) (citing Celotex
Corp. v. Catrett, 477 U.S. 317, 322–23 (1986)). Thus, “[t]he party who has the burden of proof
on a dispositive issue cannot attain summary judgment unless the evidence that he provides on that
issue is conclusive.” Torres Vargas v. Santiago Cummings, 149 F.3d 29, 35 (1st Cir. 1998) (citing
Calderone v. United States, 799 F.2d 254, 258 (6th Cir. 1986)); see also Albee Tomato, Inc. v. A.B.
Shalom Produce Corp., 155 F.3d 612, 618 (2d Cir. 1998) (“Where the movant has the burden [of
10
proof], its own submissions in support of the motion must entitle it to judgment as a matter of
law.”); Guzman v. Allstate Assurance Co., 18 F.4th 157, 160 (5th Cir. 2021) (explaining that a
summary judgment movant who also bears the burden of proof must “establish beyond
peradventure all of the essential elements of the claim or defense” (emphasis in original) (quoting
Fontenot v. Upjohn Co., 780 F.2d 1190, 1194 (5th Cir. 1986))). Only if the movant succeeds in
this initial showing must the nonmovant identify specific facts showing a genuine issue for trial.
See Guzman, 18 F.4th at 160; Anderson, 477 U.S. at 250.
Because “[c]redibility determinations, the weighing of the evidence, and the drawing of
legitimate inferences from the facts are [factfinder] functions,” Anderson, 477 U.S. at 255,
summary judgment is inappropriate “even where there is no dispute as to the evidentiary facts but
only as to the conclusions to be drawn therefrom,” Gordon v. Kidd, 971 F.2d 1087, 1093 (4th Cir.
1992) (quoting Charbonnages de Fr. v. Smith, 597 F.2d 406, 414 (4th Cir. 1979)).
B. Tax Partnership Cases
Capstone challenges the IRS’s adjustments in the November 13, 2020 FPAA, a question
subject to de novo review by this Court. Alpha I, L.P. ex rel. Sands v. United States, 682 F.3d
1009, 1025 (Fed. Cir. 2012). Accordingly, the Court gives no deference to the factual findings
made by the IRS. Litman v. United States, 78 Fed. Cl. 90, 107 (2007) (citing George E. Warren
Corp. v. United States, 135 Ct. Cl. 305, 314 (1956)). Generally, “the burden is upon the taxpayer
to establish the amount of a deduction claimed.” Helvering v. Taylor, 293 U.S. 507, 514 (1935);
accord Abrahamsen v. United States, 228 F.3d 1360, 1364 (Fed. Cir. 2000) (“In a tax refund case,
the taxpayer bears the burden of establishing the right to a refund.”). While “[t]he characterization
of a transaction for tax purposes is a question of law,” the underlying facts are subject to
factfinding. Salem Fin., Inc. v. United States, 786 F.3d 932, 940 (Fed. Cir. 2015); see also Frank
Lyon Co. v. United States, 435 U.S. 561, 581 n.16 (1978) (explaining that “[t]he general
11
characterization of a transaction for tax purposes is a question of law” although “[t]he particular
facts from which the characterization is to be made are not”).
III. DISCUSSION
To determine whether the payments at issue constituted insurance premiums, the Court
must apply binding precedent considering: (1) whether FinServ and PoolRe were separate, valid
insurance companies or whether they were shams; (2) whether the transactions under the captive
insurance arrangement transferred risk from Capstone to FinServ and PoolRe; and (3) whether
FinServ and PoolRe distributed risk. The Court finds that genuine issues of material fact prevent
Capstone from establishing that FinServ and PoolRe were not shams as a matter of law.
Additionally, genuine issues of material fact regarding risk transfer and risk distribution similarly
preclude summary judgment in Capstone’s favor.
A. Under Supreme Court and Federal Circuit Precedent, a Transaction Constitutes
Insurance, Rather than Loss Reserves, if the Captive Insurer Is Not a Sham, the
Transaction Transfers Risk from the Insured to the Insurer, and the Insurer
Distributes Risk.
As an initial matter, the parties dispute the appropriate legal test for evaluating whether
payments to a captive insurer constitute insurance premiums for federal income tax purposes.
Capstone cites the Federal Circuit’s decision in Ocean Drilling & Exploration Co. v. United States,
988 F.2d 1135, 1150–51 (Fed. Cir. 1993), which, based on the Supreme Court’s decisions in
Helvering v. Le Gierse, 312 U.S. 531, 539 (1941), and Moline Properties v. Commissioner of
Internal Revenue, 319 U.S. 436, 438–39 (1943), applied a three-part test evaluating: (1) whether
the captive insurer is a sham; (2) whether risk is transferred from the insured to the captive insurer;
and (3) whether the captive insurer distributes risk. See ECF No. 181 at 31. The Government
largely relies on cases from the Tax Court, which—similarly interpreting Le Gierse and Moline—
set forth a four-part test: (1) existence of an insurable risk; (2) risk shifting; (3) risk distribution;
12
and (4) hallmarks of commonly accepted notions of insurance. ECF No. 188 at 6 (citing Rsrv.
Mech. Corp. v. Comm’r of Internal Revenue, 34 F.4th 881, 904 (10th Cir. 2022); Avrahami, 149
T.C. at 190–91).
Since “[t]here can be no question that the Court of Federal Claims is required to follow the
precedent of the Supreme Court, [the Federal Circuit], and [the Federal Circuit’s] predecessor
court, the Court of Claims,” Coltec Indus., Inc. v. United States, 454 F.3d 1340, 1353 (Fed. Cir.
2006), the Court must apply the test articulated by the United States Court of Appeals for the
Federal Circuit in Ocean Drilling, consistent with the Supreme Court’s decisions in Le Gierse and
Moline. See Ocean Drilling, 982 F.2d at 1150–51. However, there can also be no dispute that
some facts relevant to the Ocean Drilling analysis are equally relevant when assessing a captive
insurance arrangement under the Tax Court factors, and vice versa. See id. at 1151 (noting that
the facts found to establish the insurer was not a sham under Moline also satisfied the Tax Court’s
“presence of insurance risk” and “commonly accepted notions of insurance” factors).
The first factor under Ocean Drilling requires the Court to analyze the captive insurer’s
operations and the validity of its business existence. Although related companies must generally
“be considered as separate entities in evaluating whether the transactions between the two
companies resulted in risk shifting and risk distributing” for purposes of constituting insurance, if
“the business operations of [the insurer] are a sham, [its] ‘corporate form may be disregarded.’”
Id. at 1150 (quoting Moline, 319 U.S. at 439). To determine whether the captive insurer is a sham,
courts look at whether—all facts considered—the insurer was created for a legitimate business
purpose and is carrying on a legitimate business activity, or whether it was instead created or is
operated for purposes of tax avoidance. See id. at 1151 (finding, after trial, that “[s]everal factors
contribute to recognizing [the insurer] as a valid insurance company”); Kidde Indus., Inc. v. United
13
States, 40 Fed. Cl. 42, 51 (1997) (finding, after trial, that “the totality of the evidence indicates
that . . . [the parent insured] had legitimate business reasons for creating a captive insurer
subsidiary that had nothing to do with the tax advantages that would result”), dismissed, 194 F.3d
1330 (Fed. Cir. 1999). This facts-and-circumstances analysis may consider, among other things,
(1) whether the common owner had a legitimate business reason for creating the captive insurer;
(2) whether the common owner operated the captive insurer in a manner consistent with a separate,
legitimate business; (3) whether the captive insurer had other business unrelated to that of the
common owner; (4) the capitalization and financial condition of the captive insurer; (5) whether
the insured faced true hazards; (6) whether insurance contracts were written and premiums were
paid; (7) whether the market played a role in how the captive insurer determined its premiums;
(8) whether the validity of claims were established before payments were made; (9) whether funds
were maintained separately; and (10) whether, overall, the captive insurer was operated separately.
Ocean Drilling, 982 F.2d at 1151; Kidde, 40 Fed. Cl. at 51.
Under the second Ocean Drilling factor, the court must analyze whether the captive
insurance arrangement transferred risk. Risk is transferred when the transaction reduces the
insured’s variability of loss such that the level of risk that the insured initially transferred exceeds
the level of risk it ultimately shoulders. Ocean Drilling, 982 F.2d at 1152; see Le Gierse, 312 U.S.
at 539 (“[I]nsurance involves risk-shifting and risk-distributing[.]”). “The risk to which [this
factor] refers is the risk that in the future a corporation will face uncertain and variable claims
against it.” Kidde, 40 Fed. Cl. at 53 (citing Ocean Drilling, 988 F.2d at 1151). Risk transfer occurs
when that risk shifts from the insured to the captive insurer. Id. Because the captive insurer agrees,
in return for the receipt of the premium payments, to pay all claims covered under the policy, “in
14
effect, the insured shifts the risk to the insurer that the total amount of the claims against it
(including the costs of processing) will exceed the premium paid.” Id.
The third factor in Ocean Drilling requires determination of whether the captive insurer
distributed its risk by spreading the risk among a sufficient number of unrelated policyholders.
Ocean Drilling, 982 F.2d at 1153; see Le Gierse, 312 U.S. at 539. In other words, “[b]y increasing
the total number of independent, randomly occurring risks that a [captive insurer] faces . . . , the
[insurer] benefits from the mathematical concept of the law of large numbers in that the ratio of
actual to expected losses tends to approach one.” Kidde, 40 Fed. Cl. at 53.
While the Tax Court cases that the Government cites also consider risk transfer and risk
distribution, the Tax Court cases additionally analyze whether the captive insurance arrangement
involves insurable risks and whether the captive insurer satisfies “commonly accepted notions of
insurance,” including whether: (a) the captive insurer “was organized, operated, and regulated as
an insurance company;” (b) the captive insurer “was adequately capitalized;” (c) the insurance
“policies were valid and binding;” (d) “the premiums were reasonable and the result of an arm’slength transaction;” and (e) “claims were paid.” ECF No. 188 at 16 (quoting Rsrv. Mech., 34 F.4th
at 907). The Federal Circuit did not adopt these Tax Court factors as independent factors in Ocean
Drilling, but it did acknowledge that some of the same facts relevant to analyzing whether
insurable risks exist and whether a captive insurer abides by “commonly accepted notions of
insurance” are relevant to the sham analysis under Moline. See 988 F.2d at 1151; Kidde, 40 Fed.
Cl. at 51 (analyzing whether the parent insured “had legitimate business reasons for creating a
captive insurer subsidiary that had nothing to do with the tax advantages that would result,” before
analyzing “whether the arrangement . . . is consistent with commonly accepted notions of
insurance”). As noted above though, the Tax Court factors are not binding requirements for
15
establishing insurance in this Court. Instead, the Court must determine whether the premiums
Capstone paid to FinServ and PoolRe were insurance under a totality of the circumstances. See
Ocean Drilling, 988 F.2d at 1150–53; Kidde, 40 Fed. Cl. at 51, 56–58; Moline, 319 U.S. at 439;
Le Gierse, 312 U.S. at 539.
B. Genuine Issues of Material Fact Preclude the Conclusion that FinServ and PoolRe
Were Not Shams as a Matter of Law.
1. Genuine Issues of Material Fact Preclude the Conclusion that FinServ Was Not a
Sham.
Whether FinServ was a separate, valid insurance company or was instead a sham is a
question of fact on which Capstone bears the burden of proof. N. Ind. Pub. Serv. Co. v. Comm’r
of Internal Revenue, 115 F.3d 506, 512 (7th Cir. 1997) (citing Bass v. Comm’r of Internal Revenue,
50 T.C. 595, 602 (1968)). As such, Capstone must affirmatively demonstrate that “no reasonable
trier of fact could find other than for the moving party.” Torres Vargas, 149 F.3d at 35–36 (quoting
Calderone, 799 F.2d at 258); see also Quapaw Tribe of Okla., 120 Fed. Cl. at 615; Albee Tomato,
155 F.3d at 618; Guzman, 18 F.4th at 160. Capstone has not met its burden as numerous genuine
issues of material fact prevent the conclusion that FinServ was not a sham.
Capstone argues that FinServ was a separate corporation from Capstone and was not a
sham because: (1) FinServ maintained adequate capital under Anguilla law and satisfied all
Anguilla regulatory requirements; (2) FinServ maintained separate financial accounts, issued
written policies that covered recognized risks, relied on third-party quotations and independent
actuaries in pricing its policies, and paid and documented claims after determining that the claims
were proper; and (3) the IRS did not make any adjustments to FinServ’s 2016 tax return after audit.
See ECF No. 181 at 38–40. While it does not appear that the Government disputes FinServ’s
corporate legal status, its capitalization, or compliance with Anguilla law, these facts alone are
insufficient to establish that FinServ’s separate status from Capstone was not a sham. See Ocean
16
Drilling, 988 F.2d at 1151 (citing adequate capitalization as but one of “[s]everal factors
contribut[ing] to recognizing [the captive insurer] as a valid insurance company”). Moreover, the
IRS’s previous action regarding FinServ’s 2016 tax return—which is not at issue here—is not
binding in this de novo proceeding involving Capstone’s premium deductions. See Stobie Creek
Invs., LLC v. United States, 82 Fed. Cl. 636, 663 (2008) (“The court tries factual issues de novo in
tax refund suits; no weight is given to the factual findings made by the IRS during administrative
proceedings.” (citing George E. Warren, 135 Ct. Cl. at 314)), aff’d, 608 F.3d 1366 (Fed. Cir. 2010).
Because the Court must assess the totality of the circumstances, Capstone cannot satisfy its burden
at the summary judgment stage by relying on a few undisputed or immaterial facts in isolation.
Indeed, there are several genuine disputes of material fact regarding Capstone’s payment
of premiums, FinServ’s processing and payment of claims, and whether FinServ’s loans to
affiliates support a finding that FinServ was a sham. For example, the Government identifies a
genuine dispute of evidentiary fact as to whether Capstone actually paid the purported premiums
to FinServ and in what amount. See ECF No. 188 at 35–36. To establish that Feldman Law Firm
paid FinServ $147,595 in premiums on behalf of Capstone during the 2016 tax year, Capstone
relies on a declaration from Feldman Law Firm’s controller, Paul Franks, interpreting Feldman
Law Firm’s general ledger. See ECF No. 180-1 ¶¶ 35–38 (Franks Declaration); id. at 134–35,
139–40 (excerpts from Feldman Law Firm’s general ledger attached to the Franks Declaration).
The Government, however, counters with its expert’s anticipated testimony that Feldman Law
Firm’s general ledger does not reflect that the premiums were paid in the manner that Mr. Franks
indicates. See Decl. of Michael D. Pakter ¶¶ 20–27, ECF No. 189-2.
Further, Feldman Law Firm’s general ledger only indicates an aggregate quarterly amount
paid from Feldman Law Firm to FinServ on behalf of the SSA signatories—all of which are
17
Capstone-affiliated companies. ECF No. 180-1 at 134–35, 139–40 (excerpts from Feldman Law
Firm’s general ledger indicating quarterly payments to FinServ for “captive premium[s]”); id. ¶ 38
(Franks Declaration indicating these “represent[] payments made by Capstone and other SSA
signatories”); see also ECF No. 184-9 at 44 (SSA noting that all signatories have “overlapping
ownership” with Feldman Law Firm). Capstone presents no contemporaneous evidence
establishing the breakdown of these aggregate payments between the SSA signatories and instead
relies on a memorandum to file drafted by Mr. Franks during the course of discovery in this
litigation. See ECF No. 180-1 at 29 (April 23, 2024 memorandum to file indicating that “[t]he
captive insurance premiums were allocated to the operating entities under [the SSA] on an agreed
upon allocation of 50% of the costs born[e] by The Feldman Law Firm LLP, 25% by Capstone
Associated Services, Ltd., and 25% of the costs by RSL Funding LLC”); see also id. ¶ 22 (Franks
Declaration explaining that the April 23, 2024 memorandum was produced to the Government on
or about April 30, 2024). At deposition, Mr. Franks conceded that the allocations were
“determined prior to [his] being there” and that he knew of no other documentation of that
determination aside from the memorandum. Dep. Tr. of Paul K. Franks at 19:7–20:12, 33:8–22,
ECF No. 189-13. The SSA also does not identify a premium allocation, and the allocation
described in Mr. Franks’s April 23, 2024 memorandum to file—50 percent to Feldman Law Firm
and 25 percent each to Capstone and RSL Funding LLC (“RSL Funding”)—is inconsistent with
other allocation methods used by the SSA signatories, which were based on per capita and per
space calculations. ECF No. 203-25 at 2–3 (2016 Feldman Law Firm Allocation of Expenses).
Taken together, these conflicting pieces of evidence create a genuine factual dispute as to
whether Capstone paid premiums to FinServ and in what amount. A reasonable factfinder could
find that the Government’s expert’s testimony regarding Feldman Law Firm’s ledger, along with
18
the ambiguity surrounding the allocation of premiums among the SSA signatories, indicates that
Feldman Law Firm did not make the premium payments in the manner in which Capstone claims.
See Anderson, 477 U.S. at 250 (explaining that a “genuine factual issue” is one that “may
reasonably be resolved in favor of either party”).
In reply, Capstone argues that the Government does not appear to dispute that Feldman
Law Firm received sufficient funds on Capstone’s behalf in 2016 to pay the premiums. See ECF
No. 205 at 28 (citing the United States’ Response in Opposition to Plaintiff’s Statement of
Undisputed Facts at 335–38, ECF No. 199). This argument, however, misses the point, as the
Government disputes whether Capstone paid the premiums to FinServ in the amount that Capstone
claims and in a manner that reflects that the two companies were engaged in an insurance
transaction. See ECF No. 188 at 35–36. Thus, even accepting that Feldman Law Firm took in
sufficient receipts to pay the premiums on Capstone’s behalf, whether Capstone can demonstrate
the payment and allocation of the premiums based on Feldman Law Firm’s records remains
disputed. Further, Capstone’s reliance on Mr. Franks’s declaration and memorandum to file to
establish the premium payments and allocations—rather than contemporaneous documentary
evidence—creates credibility issues that the Court cannot resolve at summary judgment.
Anderson, 477 U.S. at 255. This genuine factual dispute is material as, in conducting Moline’s
sham analysis, courts consider whether “premiums were paid” and whether such premiums “were
based on the commercial rates.” Ocean Drilling, 988 F.2d at 1151. Accordingly, the conflicting
evidence regarding Capstone’s payment of premiums to FinServ creates a genuine dispute of
material fact precluding summary judgment.
Another genuine issue of material fact exists as to the manner in which FinServ processed
and paid claims. Capstone relies on policy and procedure documents (many of which it drafted in
19
its role as a captive insurance manager) and the Feldman Declaration to demonstrate that
Capstone’s staff, acting on FinServ’s behalf, reviewed claims and established their validity prior
to paying them. See, e.g., ECF No. 180-3 ¶ 151 (Feldman Declaration indicating that “[i]n
assessing FinServ’ s exposure on a claim, Capstone insurance staff would review the claim against
the policy and make a recommendation to the insurer and the insured [about whether] coverage
exists for the claim and the amount to be paid . . . in accordance with the claims procedure”); ECF
No. 184-2 at 9–11 (Capstone Procedures Manual’s claims handling procedure); id. at 13–16
(Capstone Procedures Manual’s “Claims Overview”); ECF No. 184-8 at 282 (“Claims Handling
and Processing” document labeled with “Copy: Capstone procedures”). Yet, other documents and
testimony suggest that these policies and procedures were not followed in practice. See ECF No.
184-8 at 274–75 (FinServ documenting, on Capstone letterhead, the 2016 Ward Creative
Communications claim in a single paragraph without any explanation of why the claim was
covered by the applicable policies, how the split among policies was determined, what was being
paid, or how the amounts were calculated); id. at 277–80 (similar documentation for the 2016
Genitempo claim); Dep. Tr. of Stewart A. Feldman (Personal Capacity) at 75:4–76:20, ECF No.
189-11 (Feldman’s personal deposition testimony that he is “not sure it matters whether [the
Genitempo claim] was a ’15 or ’16 claim”); Decl. of David L. Eckles, Ph.D. ¶¶ 43–51, ECF No.
189-1 (Government’s expert opining that FinServ’s claim processing was flawed and inconsistent
with FinServ’s Capstone-drafted policies and procedures); Dep. Tr. of Daniel Calderon (Rule
30(b)(6) Designee) at 12–13, 17–19, ECF No. 189-7 (deposition testimony of one of Capstone’s
30(b)(6) witnesses, Daniel Calderon, inconsistently identifying a claims memorandum to file as
either written by Feldman-as-Capstone-the-insured or Feldman-as-Capstone-the-captivemanager).
20
This competing evidence creates a genuine issue of fact as to the manner in which FinServ
processed and paid claims, as a reasonable factfinder could conclude that in practice FinServ’s
processing and payment of claims, via Capstone’s staff, deviated from its policies and procedures,
which were also themselves drafted by Capstone, because FinServ did not sufficiently review and
verify the validity of claims prior to paying them. Additionally, Capstone’s reliance on the
Feldman Deposition to establish that Capstone and FinServ followed the procedures contained in
the policy documents, alongside the Government’s evidence that the companies did not in fact
follow the procedures, creates a credibility issue not suitable for resolution at this stage. Anderson,
477 U.S. at 255. This genuine dispute is also material. Although Capstone argues that there is no
requirement that FinServ have “standardized claims processes,” ECF No. 205 at 24, to demonstrate
that it was not a sham under Moline, courts can and do consider if “[t]he validity of claims was
established before payments were made on them.” 6 Ocean Drilling, 988 F.2d at 1151. Thus,
whether and how FinServ established the validity of claims prior to paying them is a genuine
dispute of material fact precluding summary judgment.
Finally, there exists a genuine issue of material fact regarding the competing inferences
that can be drawn from FinServ’s investments. The Government presents facts showing that
FinServ’s investments consisted of loans: (1) for real property constructively owned by Mr.
Feldman and Dr. Matz, ECF No. 190-8 at 3, 8–9 (mortgage loan from FinServ to Mr. Feldman and
6
Capstone argues that the Government’s insistence that FinServ maintain “standardized claims processes” seeks to “invalidate IRC § 831(b) insurance companies based on the proposition that they need to mimic commercial insurance carriers.” ECF No. 205 at 24. The Court takes Capstone’s point that, in imposing a $1.2 million ceiling on premiums to limit the tax benefit to small insurers, Congress likely did not intend for § 831(b) insurers to mimic large commercial insurers. See id. at 13–15. But Ocean Drilling recognized that whether a captive insurer verifies the validity of claims before paying them is one factor relevant to Moline’s sham analysis. See 988 F.2d at 1151. Thus, the Court’s finding of a genuine dispute of material fact as to whether FinServ did so does not hold Capstone to an improper standard.
21
Dr. Matz for property owned by Mr. Feldman, Dr. Matz, and The Feldman Children’s Irrevocable
Family Trust); ECF No. 190-9 at 3, 8–9 (mortgage loan from FinServ to 2075 Vermont Associates,
LLC, signed by Mr. Feldman as manager, for property beneficially owned by Mr. Feldman and
Dr. Matz); (2) to Mr. Friedman, the owner of PoolRe, ECF No. 190-5 at 3, 7–8, 16 (loan from
FinServ to Mr. Friedman secured by “[a]ll shares of common stock of PoolRe”); and (3) to other
affiliated business interests of Mr. Feldman and Dr. Matz, ECF No. 190-4 at 3–8 (loan from
FinServ to Feldman Law Firm, RSL Funding, and other Feldman-Matz companies); ECF No. 190-6 at 2–6 (promissory note between FinServ and RSL Funding). The Government maintains that
FinServ engaged in “self-dealing” rather than “traditional investments.” ECF No. 188 at 36–38.
While Capstone correctly argues that there is no genuine dispute as to the fact of FinServ’s
investments, see Oral Arg. Tr. at 35:11–22, ECF No. 208, there is a genuine issue of material fact
as to what inferences or conclusions can be drawn from those evidentiary facts in deciding the
ultimate issue of whether FinServ was a sham. See Gordon, 971 F.2d at 1093 (explaining that
summary judgment is inappropriate “even where there is no dispute as to the evidentiary facts but
only as to the conclusions to be drawn therefrom” (quotation omitted)). In other words, even
though there is no genuine dispute as to what investments FinServ made, a reasonable factfinder
could conclude either (a) that those investments support the conclusion that FinServ was a sham
or (b) that those investments support the conclusion that FinServ is a separate company with a
valid business existence.
This genuine issue of fact is also material because FinServ’s loans to Mr. Feldman, Dr.
Matz, and other related entities could suggest that FinServ was a sham under Moline. See Ocean
Drilling, 988 F.2d at 1141 (noting that the insurer “never loaned money to plaintiff or its
subsidiaries”). Although Capstone argues that there is no legal requirement for FinServ to engage
22
in “traditional investment strategies” and that there is nothing improper about loaning money from
one corporate affiliate to another, see ECF No. 205 at 23–24, whether the captive insurer’s
investment activity is consistent with what would be expected in the market, or instead reflects
that the captive insurer was a sham, is a factor that a factfinder may consider under Ocean
Drilling’s broad facts-and-circumstances analysis. See ECF No. 208 at 15:10–18 (Capstone’s
counsel emphasizing that Ocean Drilling and Moline contemplate “a broad-based analysis . . . of
all the facts and circumstances”). Thus, the significance of FinServ’s loans to Capstone-affiliated
entities and individuals is a genuine issue of material fact.
In light of the foregoing genuine disputes of material fact regarding Capstone’s payment
of premiums and FinServ’s processing of claims, as well as the conflicting inferences and
conclusions that can be drawn from FinServ’s investments, which consisted of loans to affiliates,
Capstone cannot establish that FinServ was not a sham based solely on FinServ’s corporate legal
status, as well as capitalization and compliance with Anguilla regulations.
2. Genuine Issues of Material Fact Preclude the Conclusion that PoolRe Was Not a
Sham.
As with FinServ, Capstone identifies similar facts with respect to PoolRe that purportedly
entitle it to summary judgment, specifically that: (1) PoolRe was owned by a third party (Mr.
Friedman), maintained adequate capitalization, and satisfied Anguilla regulatory requirements;
(2) PoolRe maintained separate financial accounts and issued written insurance policies to
unrelated entities; and (3) the IRS accepted PoolRe’s 2016 tax return, which it filed as an insurance
company. 7 See ECF No. 181 at 40–41. Under the totality of the circumstances, however, Capstone
has not established, as a matter of law, that PoolRe was not a sham because there are genuine
7
Like with FinServ, the IRS’s previous treatment of PoolRe’s tax return is immaterial to establishing that PoolRe was not a sham in this de novo proceeding concerning Capstone’s premium deductions. Stobie Creek Invs., 82 Fed. Cl. at 663.
23
disputes of material fact regarding whether PoolRe, in practice, operated as a separate entity
engaged in a legitimate insurance arrangement.
For example, there is a genuine issue of material fact as to how PoolRe calculated the
premiums it assessed and whether the market played a role in that determination. The Government
argues that there is a circular flow of funds among PoolRe, FinServ, and Capstone. See ECF No.
188 at 28–29. Admittedly, there is no genuine dispute as to the amount of premiums PoolRe
received and the premiums PoolRe ceded to the quota-share risk pool participants in 2016. See
ECF No. 184-10 at 502 (PoolRe’s 2016 income statement showing it received $28,321,610 in
direct written premiums and $11,167,000 in reinsurance assumed from the pooling arrangement
while ceding $28,321,610 of the direct written premiums and $11,102,691 of the reinsurance
premiums). Yet, there is a genuine dispute as to the inferences and conclusions that can be drawn
from the fact that the premiums PoolRe received were nearly identical to the premiums that PoolRe
ceded. That is, a reasonable factfinder could infer that PoolRe’s premiums were not determined
by the market and did not reflect the actual cost of potential risk but instead were strategically set
to shift funds in a way that achieved tax avoidance for the entities involved.
Capstone suggests that this dispute is immaterial, as the Government’s contention that
PoolRe’s premiums demonstrate a circular flow of funds relies on the impermissible economic
family argument. 8 See ECF No. 205 at 21–22; Rev. Rul. 2001-31, 2001-1 C.B. 1348 (indicating
8
Under the now-rejected economic family argument, the IRS historically maintained that risk could not be shifted or distributed between entities sharing common ownership because the risk remained within the same economic family. See Rev. Rul. 77-316, 1977-2 C.B. 53 (finding absence of risk transfer and risk distribution between taxpayer and captive insurer, a wholly-owned subsidiary of the taxpayer, because risk was retained within the economic family); Clougherty Packing, 811 F.2d at 1302 (explaining that “the economic family theory appears to treat a corporation and its affiliates as a single overall entity and then assess whether the insurance arrangement shifts the risk of loss away from that entity”).
24
the IRS “will no longer invoke the economic family theory with respect to captive insurance
transactions” but may “continue to challenge certain captive insurance transactions based on the
facts and circumstances of each case”); Rent-A-Ctr., Inc. v. Comm’r of Internal Revenue, 142 T.C.
1, 25 (T.C. 2014) (“Because Rev. Rul. 2001-31 has not been revoked, we could treat the economic
family argument as conceded.” (Buch, J., concurring)). But the IRS’s concession of the economic
family argument does not undermine the principle that “the corporate form may be disregarded
where it is a sham or unreal.” Moline, 319 U.S. at 439. And one factor that courts consider in
determining whether a captive insurer is a sham is whether it set its premiums based on the market.
See Ocean Drilling, 988 F.2d at 1151 (finding the fact that premiums “were based on the
commercial rates” indicated the captive insurer was not a sham).
To establish that PoolRe’s calculated premiums were consistent with what would be
expected in the market, Capstone relies on, among other things, a 2018 actuarial report analyzing
the stop-loss attachment point. See ECF No. 184-6 at 2–11. The Government challenges the
applicability and validity of that report for purposes of establishing that PoolRe’s premiums were
based on commercial rates. See, e.g., Decl. of George Zanjani, Ph.D. ¶¶ 14–23, ECF No. 189-4
(Government’s expert describing anticipated testimony that the actuarial analyses Capstone relies
upon rested on flawed assumptions and thus do not justify the premiums); ECF No. 189-1 ¶ 23
(Eckles Declaration describing anticipated testimony that PoolRe’s premiums and losses “did not
exhibit outcomes expected of a properly functioning risk pool”). Capstone also relies on a 2012
report prepared by actuary Steven Glicksman and titled “Discussion of PoolRe’s Coverage
Options.” See ECF No. 184-1 at 14–22. But the report indicates that the coverage options were
“designed to result in approximately 40.0% to 50.0% of premiums ceded to PoolRe” because
PoolRe communicated that “this is considered a desirable range of ceded premiums.” Id. at 15.
25
In other words, the report apparently did not set out to determine a premium rate based on the
market but instead set out to create coverage options that would result in a specific percentage of
ceded premiums PoolRe found to be advantageous. Moreover, Mr. Glicksman stated at deposition
that the report was specific to the year 2012 and could not be relied on for the 2016 tax year. Dep.
Tr. of Steven Allan Glicksman at 8:4–9:10, ECF No. 189-14. Overall, the conflicting evidence
creates a genuine dispute of fact as to whether PoolRe’s premiums were actuarially determined for
the 2016 tax year because a reasonable factfinder could instead conclude that the premiums were
deliberately set to achieve Capstone’s desired tax avoidance. This dispute is material because
courts consider whether the captive insurer’s premiums “were based on the commercial rates” in
determining whether the captive insurer is a sham. Ocean Drilling, 988 F.2d at 1151.
Likewise, whether PoolRe faced a risk that it would actually be liable to pay out claims is
a genuine dispute of material fact. PoolRe’s liability was explicitly limited to the amount that each
captive invested in the quota-share risk pool. See, e.g., ECF No. 184-3 at 63 (FinServ and PoolRe’s
Joint Underwriting Stop Loss Endorsement providing that PoolRe’s liability “is expressly limited
to the amount collected by PoolRe under the reinsurance policy from each of the quota share
reinsurers listed” (capitalization omitted)); see also ECF No. 189-23 at 12 (Insurance Pool
Administration Agreement between Capstone and PoolRe explaining that PoolRe “is acting in the
capacity of a facilitator and, . . . does not retain any material liability for coverages”). There also
exists a genuine dispute of material fact as to the amount of claims PoolRe paid out prior to 2016.
See ECF No. 189-20 at 14:3–22 (deposition testimony of Mr. Snyder, one of PoolRe’s directors,
that he could not recall an instance in which PoolRe had paid a claim prior to 2016); ECF No. 189-4 ¶ 25 & n.25 (Zanjani Declaration noting that the documentary evidence “contain[s] conflicting
data on the specific amount of losses paid by PoolRe”); e.g., ECF No. 192-5 at 6 (PoolRe’s 2011
26
income statement showing losses of $59,851). A reasonable factfinder could conclude from the
evidence that there was little likelihood that PoolRe would have been liable to pay out claims and
that PoolRe only paid out a de minimis number of claims. Such a dispute is material because
whether and how a captive insurer could be liable for claims is one fact courts consider in
addressing whether a captive insurer is a sham. See Ocean Drilling, 988 F.2d at 1151 (finding that
the insured faced true hazards and that the captive insurer paid out valid claims from separate funds
as facts indicating that the insurer was a not a sham); see also Le Gierse, 312 U.S. at 539 (holding
that the relevant transaction must involve actual “insurance risk” to constitute insurance).
Similarly, although it is undisputed that PoolRe had a separate corporate legal status,
genuine issues of material fact exist as to PoolRe’s independence from Capstone. While there is
no genuine dispute that Mr. Friedman formally owned PoolRe in 2016, see ECF No. 182-5 at 7
(Attachment to PoolRe’s Foreign Insurance Company Election Under Section 953(d) listing Mr.
Friedman as the 100 percent shareholder of PoolRe), several other facts suggest a relationship
between PoolRe and Capstone. For example, Mr. Freidman pledged as collateral on a loan from
FinServ “[a]ll shares of common stock of PoolRe.” ECF No. 190-5 at 16. And Mr. Snyder, one
of PoolRe’s directors, testified at deposition that he simultaneously provided “overlapp[ing]”
consulting services to Capstone during his time as a PoolRe director. ECF No. 189-20 at 11:3–
12:25. A services agreement between PoolRe, Capstone, and Feldman Law Firm further suggests
that Capstone possessed extensive control over PoolRe, including the ability to dissolve and
liquidate it. ECF No. 189-23 at 7. Indeed, PoolRe’s own conflict-of-interest policy described
PoolRe as “a wholly owned subsidiary of Capstone.” ECF No. 190-1 at 3; see also ECF No. 190-3 at 2 (Capstone’s Summary of Third-Party Reinsurance Programs describing PoolRe as a
“Capstone-administered Risk Pool” that “[l]imit[s] [its] pool participants to Capstone27
administered companies”). The totality of this evidence creates a genuine dispute of fact regarding
whether PoolRe and Capstone are unrelated, as Capstone argues, because a reasonable factfinder
could conclude that, despite Mr. Friedman’s formal ownership, in practice, the two companies did
not operate independently. Such a genuine dispute is material as courts conducting a sham analysis
consider whether “unrelated parties purchased reinsurance from [the insurer].” Ocean Drilling,
988 F.3d at 1151. Therefore, the evidence creates a genuine dispute of material fact as to PoolRe’s
independence from Capstone, FinServ, and other entities owned by Mr. Feldman and Dr. Matz.
Finally, there is a genuine dispute of material fact as to what amount of risk was ultimately
ceded under the Wesco vehicle service contracts. Capstone maintains that PoolRe reinsured thirdparty vehicle service contracts originating with Wesco, through Aria. ECF No. 180-3 ¶¶ 86–88
(Feldman Declaration describing the purported Wesco contractual liability insurance related to
vehicle service contracts assumed by PoolRe). But Capstone has not presented evidence tracing
any particular vehicle service contract insurance account from Wesco to Aria to PoolRe. Compare
ECF No. 184-1 at 40 (November 23, 2010 Reinsurance Treaty Trust Agreement between Aria,
Wesco, and Comerica Bank & Trust, N.A. referring to segregated account “SA 063”), with id. at
66 (Attachment to January 1, 2015 Coinsurance Contract issued to Aria by PoolRe referencing
“SA 010” and “SA 020”); see also id. at 73 (Quota Share Reinsurance Agreement issued to Wesco
by Aria referencing “SA#19”); id. at 89 (Quota Share Reinsurance Agreement issued to Wesco by
Aria referencing “SA#63”); 9 id. at 105 (example of vehicle insurance contract); id. at 185
9
While the Reinsurance Treaty Trust Agreement between Aria and Wesco also referenced SA 63, ECF No. 184-1 at 40, Capstone has not shown that PoolRe ever reinsured or coinsured SA 63, nor any other individual account. Compare id. at 70 (Attachment to 2015 Reinsurance Contract between FinServ and PoolRe indicating FinServ reinsured “the 2016 insurance exposures on all policies of vehicle service contracts ceded to [Aria] under its treaty dated January 1, 2015”), with id. at 64–66 (indicating that the January 1, 2015 Coinsurance Contract between Aria and PoolRe was expressly limited to SA 10 and SA 20).
28
(purported excerpt of individual policies in Wesco-Aria risk pool). Without the ability to trace the
individual policies from Wesco to Aria to PoolRe, Capstone has not established what amount of
risk such contracts ultimately ceded to PoolRe. Further, the Government has offered anticipated
expert testimony opining that the arrangement did not transfer any risk. See ECF No. 189-1 ¶¶ 28–
31 (Eckles Declaration opining that PoolRe’s alleged external reinsurance arrangements did not
transfer risk); ECF No. 189-4 ¶¶ 28–36 (Zanjani Declaration opining that the Wesco arrangement
was not proper reinsurance). Altogether, this creates a genuine issue of fact regarding what extent,
if any, the Wesco vehicle reinsurance contracts ceded unrelated third-party risk to PoolRe. This
genuine issue of fact is material, as it could change the overall conclusion about whether PoolRe
was a sham. See Ocean Drilling, 988 F.2d at 1151 (considering whether the captive insurer also
insured unrelated third parties in determining whether the insurer was a sham).
In sum, several genuine disputes of material fact prevent Capstone from establishing that
PoolRe was not a sham as a matter of law. For instance, there are genuine disputes of material
fact regarding whether PoolRe’s operations reflected a circular flow of funds, whether the market
played a role in how PoolRe calculated its premiums, whether PoolRe faced a risk that it would
actually be liable for claims of loss, whether PoolRe was independent from Capstone, and what
amount of risk, if any, was ultimately ceded under the Wesco vehicle service contracts.
* * *
As the sham determination is a factual issue on which Capstone bears the burden of proof,
the Court cannot grant summary judgment in Capstone’s favor unless Capstone has established
that, drawing all inferences in the Government’s favor, no reasonable factfinder could find in favor
of the Government. It has not done so. Rather, as numerous genuine disputes of material fact
prevent the Court from concluding that either FinServ or PoolRe was a separate, valid insurance
company—rather than a sham. Because Capstone cannot transfer risk to a sham, Capstone’s
29
inability to establish FinServ and PoolRe’s separateness as a matter of law is sufficient for the
Court to deny Capstone’s motion. See Kidde, 40 Fed. Cl. at 50 (“If the court determines that the
corporate arrangement is a sham, then the inquiry ends and the payments to the captive insurer are
treated as nondeductible reserves rather than ‘insurance premiums.’”).
C. Genuine Issues of Material Fact Exist Regarding Risk Shifting and Risk
Distribution.
Even though the existence of genuine issues of material fact regarding whether FinServ
and PoolRe were shams is sufficient to preclude summary judgment, additional grounds support
the denial of Capstone’s motion because genuine disputes of material fact also prevent the
conclusion that the transactions between Capstone, FinServ, and PoolRe transferred, and then
distributed, risk. The overall characterization of the transactions is a question of law but the
underlying factual issues from which that characterization is made are subject to factfinding. See
Salem Fin., 786 F.3d at 940; Frank Lyon, 435 U.S. at 581 n.16; Rent-A-Ctr., 142 T.C. at 38
(“Whether the facts and circumstances, evaluated in the aggregate, give rise to ‘insurance’ presents
a question of proper characterization. It is thus a mixed question of fact and law.” (Lauber, J.,
dissenting)).
1. Genuine Issues of Material Fact Preclude Summary Judgment on the Issue of Risk
Transfer.
To determine whether risk shifting has occurred, courts consider whether “the risk
ultimately shouldered by the [insured] is less than the risk transferred” to the insurer. Kidde, 40
Fed. Cl. at 57 (citing Ocean Drilling, 988 F.2d at 1152–53). For purposes of this factor, the
relevant risk is the insured’s risk of variability of future loss, and the Court’s analysis of whether
that risk was transferred from the insured to the captive insurer is based on the transaction as a
whole. See Le Gierse, 312 U.S. at 539–41. Here, numerous genuine disputes of material fact
30
prevent the conclusion that a reasonable factfinder could only find in Capstone’s favor on the issue
of risk transfer.
For example, there is a genuine dispute of material fact as to whether FinServ would have
been reimbursed by PoolRe in all realistic loss scenarios. As previously discussed, although there
is no genuine dispute as to the circumstances that dictated when PoolRe would be required to pay
a claim, there is a dispute regarding the inferences and conclusions that can be drawn therefrom.
See supra § III.B.2; see also ECF No. 184-3 at 63 (FinServ and PoolRe’s Joint Underwriting Stop
Loss Endorsement providing that PoolRe’s liability is expressly limited to the amount each insurer
pays for its quota share reinsurance premiums); ECF No. 189-23 at 12 (Capstone and PoolRe’s
Insurance Pool Administration Agreement indicating PoolRe, as a facilitator, retained no material
liability for coverages). Moreover, there is a genuine evidentiary dispute regarding which claims
PoolRe paid out prior to 2016. See ECF No. 189-20 at 14:3–22 (Mr. Snyder’s deposition testimony
that he could not recall any time when PoolRe had paid a claim before 2016); ECF No. 189-4 ¶ 25
& n.25 (Zanjani Declaration opining that the conflicting documentary evidence indicates PoolRe
only paid out a minimal number of claims); ECF No. 192-5 at 6 (PoolRe’s 2011 income statement
showing losses of $59,851). Just as these facts create a genuine dispute regarding whether PoolRe
was a sham, the same facts lead to a genuine dispute of material fact as to whether risk was
ultimately transferred from FinServ to PoolRe (and thus from Capstone to FinServ). See Le Gierse,
312 U.S. at 540–41 (finding risk shifting was not present where annuity “counteracted” life
insurance policy and thus “neutralize[d] the risk”); Ocean Drilling, 988 F.2d at 1152 (“If the level
of risk plaintiff initially had transferred and the level of risk plaintiff ultimately had shouldered
were virtually the same, plaintiff’s risk effectively would not have transferred.”); Steere Tank Line
Inc. v. United States, 577 F.2d 279, 280 (5th Cir. 1978) (finding no risk transfer where claims
31
would only be paid up to the amount of premiums); cf. R.V.I. Guar. Co., Ltd. & Subsidiaries v.
Comm’r of Internal Revenue, 145 T.C. 209, 225–26 (2015) (finding risk transfer where the insurer
“not only was financially capable of satisfying claims made against it, but it in fact paid such
claims” (quoting Harper Grp. & Includible Subsidiaries v. Comm’r of Internal Revenue, 96 T.C.
45, 59 (1991), aff’d, 979 F.2d 1341 (9th Cir. 1992))).
Thus, in addition to Capstone’s failure to demonstrate that the undisputed material facts
establish both FinServ and PoolRe were not shams, Capstone is not entitled to summary judgment
given the existence of genuine disputes of material fact regarding risk shifting.
2. Genuine Issues of Material Fact Preclude Summary Judgment on the Issue of Risk
Distribution.
Additional genuine disputes of material fact prevent the conclusion that the transactions
distributed risk as a matter of law. Insurers distribute risk by “spreading the risk of loss among
policyholders.” Ocean Drilling, 988 F.2d at 1153. Put differently, “[b]y increasing the total
number of independent, randomly occurring risks that a[n] [insurer] faces . . . , the [insurer]
benefits from the mathematical concept of the law of large numbers in that the ratio of actual to
expected losses tends to approach one.” Kidde, 40 Fed. Cl. at 53 (emphasis added). Accordingly,
to be entitled to summary judgment, Capstone must establish that the undisputed material facts
demonstrate that FinServ and PoolRe’s “amount of unrelated business sufficiently spreads risk so
as to constitute risk distribution.” Ocean Drilling, 988 F.2d at 1153. Capstone, however, has not
made such a showing. Instead, at a minimum, the record reflects numerous factual disputes
surrounding whether FinServ’s other insureds constitute unrelated risks and whether any such
amount of unrelated risk was sufficient to constitute risk distribution under the circumstances.
While Capstone contends that it has established risk distribution as a matter of law because
FinServ and PoolRe’s percentages of unrelated business in 2016 exceeded those found to distribute
32
risk in Ocean Drilling, 988 F.2d at 1153, this argument both rests on an overly broad reading of
Ocean Drilling and ignores genuine disputes of material fact regarding whether the other risks that
FinServ and PoolRe insured were independent from Capstone. First, Ocean Drilling did not hold
that 44 percent of unrelated business is sufficient, as a matter of law, to establish risk distribution.
Instead, the Federal Circuit affirmed the lower court’s finding, after trial, that risk distribution was
present under the circumstances of that case, which included the fact that the captive insurer’s
unrelated business constituted between 44 percent and 66 percent of the captive insurer’s overall
business during the years at issue. See id.; see also ECF No. 208 at 28:5–23 (Capstone’s counsel
conceding at oral argument that “there is no magic number” of unrelated business that
demonstrates risk distribution because the inquiry is instead based on the “facts and
circumstances”).
Second, there are genuine issues of material fact regarding whether the other risks that
FinServ and PoolRe insured, and which Capstone claims are unrelated, were truly independent and
randomly occurring such that they would distribute risk. As one example, there is a genuine
dispute of material fact regarding whether FinServ’s policies were spread among independent and
randomly occurring risks as many, if not all, of FinServ’s insureds were located in Houston, Texas
(with some even in the same office building), and shared or contracted the same employees. See
ECF No. 180-3 ¶ 99 (Feldman Declaration explaining that POC Lease Holdings, LLC, a named
insured of FinServ, was the landlord for Capstone and other related companies at a shared office);
id. ¶ 174 (noting that “all companies [under the SSA] operat[ed] under the same master office
lease”); id. ¶ 64 (indicating Capstone was “an administrative co-manager” of FinServ); id. ¶¶ 120–
21 (explaining that “Capstone’s internal underwriters” in addition to “independent third-party
underwriters” hired by “Capstone’s staff” assisted in pricing FinServ’s policies).
33
The fact that the companies shared the same office and employees means that a reasonable
factfinder could conclude that they do not represent unrelated risks. For instance, if an employee
who works for multiple policyholders is unavailable due to a covered loss, that covered loss affects
multiple policyholders. See ECF No. 208 at 56:6–25. This dispute is material to establishing risk
distribution because Capstone must demonstrate that FinServ spread risk among a sufficient
number of independent, randomly occurring risks. Ocean Drilling, 988 F.2d at 1153; Kidde, 40
Fed. Cl. at 53. Thus, there is a genuine dispute of material fact regarding whether FinServ’s
insureds constitute independent risks given their common geographic location and shared
employees.
The existence of genuine disputes of material fact regarding whether FinServ and PoolRe
distributed risk among their insureds precludes summary judgment in Capstone’s favor. For this
additional reason, Capstone’s motion must be denied.
IV. CONCLUSION
For the foregoing reasons, Capstone’s Motion for Summary Judgment (ECF No. 180) is
DENIED. The parties shall submit a joint status report proposing a schedule for further
proceedings by no later than October 16, 2026.
SO ORDERED.
Dated: September 25, 2026 /s/ Kathryn C. Davis
KATHRYN C. DAVIS
Judge
34