Gulf Oil Corp. v. Commissioner

Gulf Oil Corporation, Petitioner v. Commissioner of Internal Revenue, Respondent

Gulf Oil Corp. v. Commissioner

Docket No. 22499-82

United States Tax Court

89 T.C. 1010; 1987 U.S. Tax Ct. LEXIS 162; 89 T.C. No. 70;

November 24, 1987; As Amended November 30, 1987; As amended December 7, 1987 November 24, 1987, Filed

---

In 1971, Gulf incorporated Insco, a wholly owned foreign subsidiary, to conduct a general insurance business. During 1974 and 1975, Gulf and its affiliates insured risks with unrelated insurance carriers who, by prearrangement, reinsured with Insco and ceded premiums on such reinsurance to Insco. Insco paid claims on the reinsured risks. In 1975, Insco began insuring risks of unrelated parties. Net premium income from unrelated parties represented 2 percent of its total net premium income in 1975.

The Commissioner determined that premium payments made by Gulf and its domestic affiliates were not deductible to the extent those payments were ceded to Insco. He also recharacterized, as constructive dividends to Gulf, the premiums paid by the foreign affiliates which were ceded to Insco. In addition, he recharacterized, as constructive dividends to Gulf, the payments of claims by Insco to Gulf and its domestic affiliates. Held, the sums paid to Insco are not deductible as ordinary and necessary business expenses. Humana v. Commissioner, 88 T.C. 197">88 T.C. 197 (1987); Clougherty Packing Co. v. Commissioner, 84 T.C. 948">84 T.C. 948 (1985), affd. 811 F.2d 1297">811 F.2d 1297 (9th Cir. 1987); and Carnation Co. v. Commissioner, 71 T.C. 400">71 T.C. 400 (1978), affd. 640 F.2d 1010">640 F.2d 1010 (9th Cir. 1981), followed. Held, further, the premiums paid by the foreign affiliates to Insco, and the claims paid by Insco, do not constitute constructive dividends to Gulf. See Mobil Oil Corp. v. United States, 8 Cl. Ct. 555">8 Cl. Ct. 555 (1985).

J. Waddy Bullion, Emily A. Parker, Sean T. Crimmins, Buford P. Berry, J.Y. Robb III, and Margaret S. Alford, for the petitioner.

Joel V. Williamson and Joseph R. Goeke, for the respondent.

Whitaker, Judge. ^*Nims, Korner, Shields, Hamblen, Clapp, Swift, Jacobs, Gerber, Wright, Parr, and Williams, JJ., agree with the majority opinion. Wells, J., did not participate in the consideration of this case. Goffe, J., concurring. Chabot and Cohen, JJ., agree with this concurring opinion. Chabot, J., concurring. Sterrett, Goffe, and Cohen, JJ., agree with this concurring opinion. Nims, J., concurring. Whitaker, Korner, Hamblen, Swift, Jacobs, Gerber, Wright, Parr, and Williams, JJ., agree with this concurring opinion. Parker, J., dissenting.

---

WHITAKER

The Commissioner determined deficiencies in petitioner's Federal income tax for the taxable year 1974 in the amount of $ 80,813,428, and for the taxable year 1975 in the amount of $ 166,316,320. Petitioner, respondent, and the Court agreed that certain issues would be severed and tried at a special trial session. ¹ One of the issues tried was designated as the "Insco issue." It involves the following two questions:

(1) Whether petitioner may deduct as ordinary and necessary business expenses amounts paid as insurance premiums by Gulf Oil Corp. and its domestic affiliates to the extent that those payments were ceded to its wholly owned captive insurance company, Insco, Ltd.; and

(2) Whether the payments designated as premiums made by the foreign affiliates of Gulf Oil Corp. which were ceded to Insco, Ltd., and the claims paid by Insco, Ltd., to Gulf Oil Corp. and its domestic affiliates represent constructive dividends to Gulf Oil Corp..

FINDINGS OF FACT

Some of the facts have been stipulated. The stipulation of facts and accompanying exhibits are so found and incorporated by this reference.

Gulf Oil Corp. (hereinafter referred to as petitioner or Gulf) is a corporation organized under the laws of the Commonwealth of Pennsylvania with its principal office in Pittsburgh, Pennsylvania. During the taxable years at issue, Gulf and certain of its subsidiary corporations constituted an "affiliated group" as that term is defined in section 1504. ² Petitioner, directly and through its foreign subsidiaries and affiliates, is engaged in world-wide exploration, development, production, purchase, transportation, and marketing of petroleum products. Petitioner maintained its books of account for the taxable years in issue on the accrual method of accounting using the calendar year as its taxable year. Gulf, as the common parent of an affiliated group of corporations, timely filed consolidated Federal income tax returns for its taxable years 1974 and 1975 on behalf of itself, and certain of its subsidiary corporations, with the Office of the Internal Revenue Service at Pittsburgh, Pennsylvania.

By virtue of their world-wide operations, Gulf and its foreign and domestic subsidiaries and affiliates were exposed to various risks. These risks fall into three general categories: (i) those associated with domestic and foreign on-and off-shore properties, marine hull and machinery, and marine cargo which include, but are not limited to: fire, windstorm, flood, earthquake, spills, pollution, seepage, contamination, shipwreck, collision, seizure, explosion, hurricane, typhoon, cyclone, hail, lightning, and liability to third parties incident thereto; (ii) those associated with general casualty which include, but are not limited to, workmen's compensation, automobiles and other vehicles, and products liability; and (iii) those associated with disability, life, and accident of employees.

Until the late 1960's, commercial coverage for the risks of Gulf and its affiliates was available at acceptable rates and adequate coverages. Up until this time, the general policy of Gulf and its affiliates was to self-insure risks up to an amount of $ 1 million. Insurance coverage for risks in excess of $ 1 million, including catastrophic risks, which were generally considered to be those risks in excess of $ 10 million, was obtained from third-party primary insurance carriers and reinsurers in the United States and world-wide markets. However, in the late 1960's, several incidents occurred in the oil industry that caused commercial insurance carriers to increase rates and provide more limited coverage or to altogether eliminate coverage for certain risks. Some of these incidents included a refinery explosion at Lake Charles, Louisiana, an oil spill off the coast of Santa Barbara, California, and Hurricanes Betsy and Camille. Subsequently, Gulf decided that its excellent loss history was not being adequately reflected in the higher rates in the coverages that remained available in the primary and reinsurance markets. In response, Gulf participated in the creation of Oil Insurance Ltd. (OIL) with several other major and independent oil companies and also created its own insurance company subsidiary, Insco, Ltd.

OIL was formed in late 1970 as a petroleum industry mutual insurance company for the purpose of providing insurance coverage of catastrophic risks of its member-shareholders in excess of $ 10 million. OIL originally included Gulf and seven other shareholder companies. In 1971, OIL was restructured because of concern over the deductibility of premium payments. As part of the restructuring, shareholders could only withdraw from OIL over a 5-year period. The method of calculating premiums was also changed so that they were based upon both the loss experience over the previous 5 years and the gross assets of each shareholder. Gulf was dissatisfied with the new premium formula because it resulted in the subsidy of smaller, independent shareholders by larger shareholders, such as Gulf. Nevertheless, Gulf agreed to participate in OIL as restructured.

In further response to the changes in the third-party primary insurance and reinsurance markets, Gulf contemplated the formation of a captive insurance subsidiary to provide coverage for Gulf and its subsidiaries and affiliates. In 1968, Gulf conducted a feasibility study in concert with Marsh & McLennan, Inc., an insurance brokerage and consulting firm, with regard to the advisability of establishing an affiliated insurance company outside of the United States. This study, which updated a 1964 study, included consideration of the potential tax and financial benefits to be obtained through the operation of such an insurance affiliate. On June 10, 1971, the management of Gulf approved a plan to establish an affiliated insurance company in Bermuda. Under this plan, the insurance affiliate would initially insure only certain foreign risks of domestic subsidiaries. Later, the insurance affiliate was to provide further insurance, including coverage for Gulf's marine fleet and U.S. situs risks. Gulf also contemplated eventually offering insurance coverage to unrelated third parties.

On November 3, 1971, Gulf incorporated Insco, Ltd. (Insco), as a foreign wholly owned subsidiary authorized to conduct general insurance business under the laws of Bermuda. Although an initial capitalization of $ 10 million was authorized, Insco initially issued 1,000 shares with a value of $ 1,000 per share of which only 12 percent was paid up. Insco established its office in Hamilton, Bermuda, with a staff of two, consisting of A.W. Gregg, president, and a secretarial assistant. Gregg was a Gulf employee who had been in the insurance division of Gulf's treasury department. In addition, David B. Vaughn of Marsh & McLennan Management, Ltd., a Bermuda subsidiary of Marsh & McLennan, Inc., served as vice president, and R.S.L. Pearman of Conyers, Dill & Pearman, solicitors for Insco, served as assistant secretary. W.H. Burkhiser and J.C. Kelly, both employees of Gulf, served as treasurer and secretary of Insco, respectively. Pursuant to an agreement dated November 10, 1971, Marsh & McLennan Management, Ltd., agreed to provide Insco with all underwriting, premium rating, claims, reinsurance, record keeping, banking, and check disbursement services relative to its operations.

Generally, but not uniformly, the following depicts the relationship between Gulf, its affiliates, primary commercial carriers, Insco Ltd., OIL, and reinsurers of catastrophic risks. Gulf and its affiliates entered into insurance contracts with third-party commercial carriers relative to their respective risks. By prearrangement with these insurance carriers, a significant portion of the primary carrier's exposure was reinsured with Insco Ltd. Gulf and its affiliates paid premiums directly to the primary commercial carriers. To the extent the primary carriers reinsured the risks of Gulf and its affiliates with Insco. Ltd., premiums paid them by Gulf and its affiliates were ceded to Insco Ltd. The primary carriers retained a commission for acting as a fronting or ceding company for Insco Ltd. The risks assumed by Insco Ltd. from primary carriers relative to Gulf and its affiliates were limited to $ 10 million but did not include the first $ 1 million of loss. This latter amount was self-insured by Gulf and its affiliates. It was the practice of primary carriers and Insco Ltd. to reinsure catastrophic risks, i.e., those exceeding $ 10 million, with third-party reinsurers on the worldwide reinsurance market or place them in OIL. A portion of the premiums received by Insco Ltd. attributable to its insurance of a portion of the risks of Gulf and its affiliates was ceded to reinsurers who accepted the catastrophic risks of Gulf and its affiliates exceeding $ 10 million.

Relative to the risks of Gulf and its affiliates associated with on-shore properties, such as refineries and chemical plants, it was the general practice to self-insure such properties up to a $ 1 million deductible level. Risk exposures in excess of $ 1 million up to a $ 40 million level were placed with a primary carrier such as Oil Insurance Association, a third-party commercial insurance concern. By prearrangement, the first $ 10 million of coverage, exclusive of the $ 1 million deductible, was reinsured with Insco Ltd. Risk exposures between $ 10 million and $ 40 million were retained by the primary carrier. Risk exposures exceeding $ 40 million, up to a level of $ 150 million, were placed with OIL. Upon Gulf's phased withdrawal from OIL, risk exposures exceeding the $ 40 million limit were reinsured in the worldwide commercial reinsurance market.

Relative to risks associated with off-shore properties such as drilling rigs and production equipment, the practice of Gulf and its affiliates was to self-insure up to a level of a $ 1 million deductible. Risk exposures in excess of $ 1 million were insured by a primary carrier who was frequently a member of the American International Group. The limit of coverage in the case of off-shore properties was determined by the value of the properties involved. By prearrangement, generally 50 percent of the primary carrier's coverage limit, up to a $ 10 million level, was reinsured with Insco Ltd. The primary carrier retained coverage in excess of $ 10 million and often reinsured a portion of this coverage in the worldwide reinsurance market or with OIL.

Relative to the risks of Gulf and its affiliates associated with marine hull, i.e., oil tankers and other vessels, it was the practice of Gulf and its affiliates to self-insure up to a $ 1 million deductible level. Risk exposures in excess of the $ 1 million, deductible, were placed with a primary carrier such as Lloyd's of London, American International Group, or Argonau Mid-West Insurance Co. The limits of the liability of the primary carrier depended upon the value of the particular hull being insured. By prearrangement with the primary carrier, 50 percent of the primary carrier's coverage up to a $ 10 million level was reinsured with Insco Ltd. The primary carrier retained coverage in excess of $ 10 million and often reinsured a portion of this coverage in the worldwide reinsurance market, generally through the London reinsurance market. OIL did not assume any catastrophic coverage relative to the marine hull risk exposures of Gulf and its affiliates.

Relative to the risks of Gulf and its affiliates associated with marine cargo, it was the practice to self-insure up to a $ 1 million deductible level. Risk exposures in excess of the $ 1 million deductible level were placed with a primary carrier such as the Insurance Co. of North America. The amount of coverage held by the primary carrier was dependent upon the limit of the policy of insurance. By prearrangement with the primary carrier, the latter's coverage was reinsured with Insco Ltd. up to a level of $ 10 million. The primary carrier retained coverage in excess of $ 10 million and often reinsured a portion of this coverage in the worldwide reinsurance market or with OIL.

Gulf made available to its employees a group life insurance benefit plan in which employees could participate. This plan was written through a policy issued by Connecticut General. The benefit plan included a provision which continued life insurance for employees who were disabled, with Gulf paying a premium for such life insurance while they were so disabled. By prearrangement, 100 percent of Connecticut General's exposure was reinsured with Insco Ltd. while the employee was disabled. The premium for this exposure was ceded to Insco Ltd. Connecticut General received a fee for administrative services provided to Insco Ltd.

Insco's premium income received from Gulf, Gulf's affiliates, and third parties consisted of earned and unearned premiums. Earned premiums relate to amounts paid in a particular year for coverage in that year, while unearned premiums relate to amounts paid in a particular year for coverage in a subsequent year.

Set forth below are the gross premiums written and the net premium income reported by Insco on its financial statements for the years 1972 through 1983:

Year	Gross premiums written	Net premium income
1972	$ 1,064,000	$ 569,000
1973	8,118,000	2,505,000
1974	12,344,000	11,816,000
1975	21,300,000	16,177,000
1976	31,770,000	30,079,000
1977	40,269,000	41,238,000
1978	85,017,000	73,926,000
1979	112,630,000	75,812,000
1980	153,670,000	87,621,000
1981	154,022,000	94,758,000
1982	140,454,000	98,158,000
1983	114,090,000	82,264,000

Net premium income is gross premiums written, less the increase or decrease in unearned premiums relative to the prior year, minus commissions, expenses, and premium taxes.

Set forth below is a schedule of Insco's yearly and cumulative investment income (primarily consisting of interest income) for the years 1972 through 1980.

	Yearly	Cumulative
Year	investment income	investment income
1972	$ 8,197	$ 8,197
1973	85,433	93,630
1974 *	1,151,290	1,244,920
1975 *	2,439,525	3,684,445
1976 *	4,102,747	7,787,192
1977	6,886,287	14,673,479
1978	9,260,000	23,933,479
1979	14,188,000	38,121,479
1980	21,285,000	59,406,479

(Amounts are computed on a consolidated basis where appropriate. Years with asterisks end 11/30; all other years end 12/31.)

For the years 1972 through 1983, the amounts that Insco had available to meet insurance loss claims (that is, the value of Insco's assets minus its liabilities (other than estimated unpaid losses and loss expenses ³)) were as follows:

Year	Amount
1972	$ 0.7 million
1973	3.0 million
1974 *	1.5 million
1975 *	34.3 million
1976 *	61.2 million
1977	79.7 million
1978	124.7 million
1979	171.6 million
1980	223.9 million
1981	260.1 million
1982	285.4 million
1983	305.4 million

(Amounts are computed on a consolidated basis, where appropriate. Years with asterisks are years ending 11/30. All other years end 12/31.)

Set forth below is a schedule of Gulf and its foreign and domestic affiliates whose risks were reinsured with Insco during the taxable years 1974 and 1975 along with their respective country of incorporation:

	Country of
Name	incorporation
Gulf Oil Corp	U.S.
Mene Grande Oil Co	U.S.
Ecuadorian Gulf Oil Co	U.S.
Zaire Gulf Oil Co	U.S.
Key International Drilling Co., Ltd	Bermuda
Britama Tankers, Ltd	England
Belgulf Tankers, NV	Belgium
Nedgulf Tankers, NV	Belgium
Fuel Transport Co., Ltd	Liberia
Compannia Maritima Rio Gulf Co., SA	Spain
Gulf Oil Company (Nigeria), Ltd	Nigeria
Afran Transport Co	Liberia
Gulftankers, Inc	Liberia

Insco paid claims to primary carriers relative to all risks it reinsured totaling $ 1,001,444 and $ 3,107,212 for the years 1974 and 1975, respectively. No part of the claims paid in 1974, and only $ 48,018 of the claims paid in 1975 related to the risks of unrelated third parties.

Members of the American International Group, Inc. (AIG), served as primary insurers for a substantial amount of the risks of Gulf and its affiliates that were reinsured with Insco. On December 20, 1973, Gulf executed a guaranty in favor of AIG that obligated Gulf to indemnify AIG in the event Insco could not meet its obligations regarding the risks it reinsured. The AIG guaranty remained in effect throughout the relevant years. On December 20, 1973, Gulf also executed a similar guaranty relative to the risks reinsured by Oil Industry Association. The latter guaranty, or a substituted version thereof, was in effect throughout the relevant years. Gulf was never required to indemnify any primary insurers under these guaranties.

In 1975, several changes in the operation and structure of Insco were made. Among these changes, Gulf transferred ownership of Insco to Transocean Gulf Oil Co. (Transocean), a wholly owned Gulf holding company. At this time, Insco called its issued but non-paid-up shares and Transocean contributed capital of $ 880,000. Concurrently, Insco issued 9,000 new shares at a value of $ 1,000 per share. These shares were fully paid up by Transocean thereby increasing the paid-in capital of Insco to $ 10 million. Gulf and its affiliates no longer directly placed catastrophic coverage with OIL or other third-party reinsurers, but rather placed those risks with Insco, which, in turn, reinsured those risks.

In 1975, Insco first began insuring risks of unrelated parties. Net premium income from the insurance of non-Gulf risks represented 2 percent of the total net premium income of Insco for the taxable year 1975. Gulf also determined that it should withdraw from OIL and commenced doing so over the minimum 5-year period. In October 1975, J.M. Turnbull, of Gulf's financial department, replaced A.W. Gregg as president of Insco.

In a statutory notice of deficiency mailed to petitioner, the Commissioner determined that payments made by Gulf and its domestic affiliates to primary insurers were not deductible insurance premiums to the extent those payments were ceded to Insco. As a result, the Commissioner disallowed insurance expense deductions claimed by Gulf and its domestic affiliates as follows:

	Insurance expense disallowed
	1974	1975
Gulf Oil Corp.	$ 8,970,653	$ 9,426,308
Mene Grande Oil Co.	1,059,306	1,168,453
Ecuadorian Gulf Oil Co.	214,398	185,414
Total	10,244,357	10,780,175

With the consent of petitioner, the Commissioner amended his pleadings to disallow additional deductions claimed for insurance premiums ceded to Insco in the amounts of $ 40,973 and $ 119,906 for the taxable years 1974 and 1975, respectively.

The Commissioner also recharacterized the insurance premium payments made by foreign affiliates of Gulf as constructive dividends to Gulf to the extent those premiums were ceded to Insco. In the statutory notice of deficiency, the Commissioner determined that Gulf received dividend income in the amounts of $ 2,659,410 and $ 4,662,194 ⁴ for the taxable years 1974 and 1975, respectively. In addition, with the consent of petitioner, the Commissioner amended his pleadings to reflect that Gulf had additional taxable income from constructive dividends arising from the payment of insurance premiums to Insco by foreign affiliates of Gulf in the amount of $ 1,369,236 for the taxable year 1974. The parties agree that for the taxable year 1974, each of the Gulf foreign affiliates who paid premiums to Insco, through third-party primary insurers, had sufficient current or accumulated earnings and profits equal in amount to the constructive dividends determined by respondent. The parties agree that with the exception of Belgulf Tankers, NV, Britama Tankers, Ltd., Gulftankers, Inc., and Gulf Oil (Great Britain) Ltd., each of the Gulf foreign affiliates who paid premiums to Insco in 1975, through third-party primary insurers, had sufficient current or accumulated earnings and profits equal in amount to the constructive dividends determined by respondent.

In the statutory notice of deficiency, the Commissioner also treated claims paid by Insco in the taxable years 1974 and 1975, relative to the reinsurance of the risks of Gulf and its domestic affiliates, as constructive dividends directly to Gulf or to Gulf through Transocean. These payments of claims, which were treated as constructive dividends, totaled $ 1,001,441 and $ 3,059,194 for the taxable years 1974 and 1975, respectively. However, the Commissioner determined that Gulf and its domestic affiliates sustained deductible uninsured losses under section 165 for the taxable years 1974 and 1975 totaling $ 1,001,444 and $ 3,059,194, respectively.

From 1976 to the present, Insco increased its underwriting of third-party risks and continued to underwrite additional risks of Gulf and its affiliates. In April 1977, Insco hired Leslie Dew to guide the expansion of Insco's underwriting efforts. He was named vice president and chief operating officer for underwriting, and was appointed to Insco's executive committee and board of directors. Dew is well respected in the insurance industry. He has an extensive background as a non-marine insurance underwriter at Lloyd's of London, with one of the major underwriting syndicates. He served as deputy chairman of Lloyd's of London immediately prior to commencing work with Insco.

In 1978, Insco terminated its service agreement with Marsh & McLennan Management, Ltd. Subsequently, Insco systematically assembled the necessary staff to perform all the services previously performed by Marsh & McLennan Management, Ltd. Based upon the recommendation of Dew, Insco formed Britamco, Ltd., as a wholly owned Bermuda subsidiary. Britamco, Ltd.'s business activity was to act as an agent for Insco and other insurance companies in writing third-party insurance and reinsurance business. In order to underwrite third-party reinsurance risks in the United States, Insco entered into a trust agreement with Citibank, New York, New York, pursuant to which Insco set aside $ 10 million in liquid assets as security for its U.S. insureds and reinsureds relative to claims payable in U.S. currency. By establishing the $ 10 million trust fund, Insco was able to qualify as a non-admitted surplus line insurer in various States throughout the United States.

Set forth below is a schedule of the net premium income earned by Insco from reinsuring the risks of Gulf and its associates and third parties, as well as the amount of net premium income earned from reinsuring third-party risks as a percentage of total net premium income for the years 1976 through 1983:

			Percentage of
Year	Gulf-related risks	Third-party risks	third-party risks
1976	$ 29,049,000	$ 2,235,000	7 percent
1977	32,295,000	6,008,000	16 percent
1978	32,615,000	33,902,000	51 percent
1979	30,871,000	36,214,000	54 percent
1980	36,202,000	40,881,000	53 percent
1981	39,312,000	46,544,000	54 percent
1982	46,497,000	42,381,000	48 percent
1983	29,011,000	46,520,000	62 percent

It was the practice of Insco Ltd. to limit its individual loss occurrence exposure relative to third-party risk to $ 500,000 or less. Reinsurance was placed with unrelated reinsurers at appropriate terms and conditions to accomplish this.

The following is a schedule of claims paid by Insco during the years 1976 through 1980:

Year	Claims paid
1976	$ 4,718,000
1977	12,698,000
1978	5,989,000
1979	15,169,000
1980	25,947,000

Set forth below is a schedule of the dividends paid by Insco to Transocean during the years 1976 through 1983:

Year	Dividends
1976	0
1977	$ 10,000,000
1978	13,000,000
1979	10,000,000
1980	10,000,000
1981	10,000,000
1982	10,000,000
1983	11,000,000

During the year 1982, Insco advanced $ 30 million to Transocean without providing for the payment of interest. This advance was to be repaid out of anticipated future dividends of Insco.

OPINION

In this case, we must decide whether Gulf may deduct as ordinary and necessary business expenses amounts paid as insurance premiums to its wholly owned captive, Insco, in 1974 and 1975. In each of our prior opinions in which we have addressed the captive insurance issue, we concluded thatpayments to the captive subsidiary, designated as premiums, whether from the parent corporation or from other subsidiaries, did not represent payments for insurance. Carnation Co. v. Commissioner, 71 T.C. 400">71 T.C. 400 (1978), affd. 640 F.2d 1010">640 F.2d 1010 (9th Cir. 1981); Clougherty Packing Co. v. Commissioner, 84 T.C. 948">84 T.C. 948 (1985), affd. 811 F.2d 1297">811 F.2d 1297 (9th Cir. 1987); Humana v. Commissioner, 88 T.C. 197">88 T.C. 197 (1987).⁵ In each of those situations, the captive was wholly owned by its parent, and the captive insured risks only within the affiliated group. In this case, we are for the first time faced with a wholly owned captive that insures unrelated third-party risks as well as those of its affiliated group. The issue before us is whether the insurance of these risks produces a result different from our prior decisions.

Gulf argues that this case is distinguishable because Insco's intention to insure the risks of unrelated third parties had been established for years prior to the taxable years in issue, that in 1975, one of the years before the Court, Insco did in fact insure unrelated parties, and that the large amount of unrelated third-party risks insured in later years demonstrates that Insco is and should be treated as an unrelated insurance company. Respondent argues that Insco's writing of third party insurance is irrelevant to the existence of risk transfer, that the economic family concept precludes deductible insurance premiums to the extent that the captive insures any risks of an affiliated corporation.

In Carnation, Clougherty, and Humana, we were attempting to determine whether the arrangements between the affiliated parties were insurance as defined by the Supreme Court in Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531 (1941). We concluded in each situation that they were not. ⁶ Central to our holdings are two principles. First, to have insurance, risk transfer and risk distribution must be present. Second, payments designated as premiums for insurance, which are the equivalent of nondeductible payments to a reserve for losses, are not payments for insurance; there is no risk transference in such situations. Although technically, transfer of risk may occur when a captive is involved that is a separate, viable entity, financially capable of meeting its obligations, we simply decline to recognize it as such when the arrangement is merely, in substance, the equivalent of a reserve for losses or self-insurance.

To have a true transfer of risk, another risk-bearer must replace the insured; to speak of a transfer of risk to a fund or reserve established by the insured is merely to describe self-insurance in the jargon of insurance. [K. O'Brien & K. Tung, "Captive Offshore Insurance Corporations," 31 N.Y.U. Thirty-First Ann. Inst. on Fed. Tax. 665, 678-684 (H. Sellin ed. 1973). Fn. refs. omitted.]

Furthermore,

It is apparent that the nature of the captive-insurance device involves not only the element of insurance through "transfer" of risks, but also the notion of self-insurance since the "owners" of the risks insured therein are the "owners" of the insurer. The fortunes of the two entities are interlocked to the extent that the risks insured in the captive are not reinsured. In this sense, captive insuring can be considered a risk-retention device similar to self-insurance. In fact, if self-insurance involves the conduct of risk management "according to all the sound principles and practices employed by insurance companies" it might be argued that captive insuring is the epitome of the self-insurance device * * *. [Robert S. Goshay, "Captive Insurance Companies," Risk Management, Ch. VI, Richard D. Irwin, Inc., Homewood, Illinois, 1964, pp. 80-121, at p. 85, as referenced in Humana v. Commissioner, 88 T.C. at 211.]

In addition, we rejected the "economic family" theory espoused in Rev. Rul. 77-316, 2 C.B. 53">1977-2 C.B. 53. Under that theory, we could have reached the same result, but we would have foreclosed a wholly owned captive from ever being considered a separate insurance company, payments to which are deductible by its owners. ⁷ Recognizing that there may in fact be such situations, we declined to adopt this theory. We specifically reserved any discussion of the tax consequences of payments to captives with unrelated owners and/or unrelated insureds. Clougherty Packing Co. v. Commissioner, 84 T.C. at 960; Humana v. Commissioner, 88 T.C. at 210.

We are now faced with one of these situations. ⁸ The facts of this case are for all purposes identical to those in Carnation, Clougherty, and Humana (with respect to the parent-subsidiary relationship), except that Insco, Gulf's wholly owned subsidiary, began insuring risks of third parties in 1975, one of the years before the Court. The net premium income from the insurance of third-party risks in 1975 was only 2 percent of Insco's total net premium income. However, in the years 1978 through 1981, and 1983, net premium income from third parties exceeded 50 percent of the total. In 1975, Insco paid total claims of $ 3,107,212, of which $ 48,018 represented claims paid to third parties. The issue before us is the effect, if any, of this unrelated business, on a holding that would otherwise be controlled by our prior cases.

Under principles of the insurance industry, risk transfer and risk distribution occur only when there are sufficient unrelated risks in the pool for the law of large numbers to operate. ⁹ As the number of unrelated risks is increased,

protection is improved against the chance that the severity and number of harmful events will be spread over time or in other ways in groupings disproportionate to the overall risk. That is, with an increasing number of ventures in a combined pool, the unusually favorable and unusually harmful experiences tend to stay more nearly in balance * * * [R. Keeton, Insurance Law Basic Text 6-7 (1971). Emphasis added; fn. refs. omitted.]

In this instance "unrelated" risks need not be those of unrelated parties; a single insured can have sufficient unrelated risks to achieve adequate risk distribution. ¹⁰

When the law of large numbers operates, the risk of each policyholder is divided into small units and is transferred to and distributed among the policyholders. As the Supreme Court stated in Group Life & Health Ins. Co. v. Royal Drug Co., 440 U.S. 205">440 U.S. 205, 211 (1979):

It is characteristic of insurance that a number of risks are accepted, some of which involve losses, and that such losses are spread over all the risks so as to enable the insurer to accept each risk at a slight fraction of the possible liability upon it. [Citations omitted.]

Thus, risk transfer and risk distribution occur upon the issuance of each policy. ¹¹

The premium allocable to each insured contributes to the payment of aggregate losses whether or not that insured actually suffers any loss. The cost of future estimated losses is capable of reasonably accurate estimation; what is uncertain is the identity of the insureds (or properties) which will suffer losses. ¹² Thus:

Insurance allows the individual insured to substitute a small, definite cost (the premium) for a large but uncertain loss (not to exceed the amount of the insurance) under an agreement whereby the fortunate who may escape loss will help compensate the unfortunate few who suffer loss. [R. Mehr, Fundamentals of Insurance 33 (1983).]

If all of the insureds are related, the insurance is merely self-insurance because the group's premium pool is used only to cover the group's losses. By adding unrelated insureds, the pool, from which losses are paid no longer, is made up of only the affiliated group's premiums. When a sufficient proportion of premiums paid by unrelated parties is added, the premiums of the affiliated group will no longer cover anticipated losses of all of the insureds; the members of the affiliated group must necessarily anticipate relying on the premiums of the unrelated insureds in the event that they are "the unfortunate few" and suffer more than their proportionate share of the anticipated losses. ¹³

Thus, when the aggregate premiums paid by the captive's affiliated group is insufficient in a substantial amount to pay the aggregate anticipated losses of the entire group, the affiliates and unrelated entities, the premiums paid by the affiliated group should be deductible as insurance premiums and should no longer be characterized as payments to a reserve from which to pay losses. ¹⁴ Risk distribution and risk transfer would be present, and the arrangement is no longer in substance equated with self-insurance.

With these principles in mind, we turn to the facts of the present case. For the year 1974, the facts of this case are in all material respects the same as those in our prior decisions, and therefore those decisions control. The amounts paid to Insco by Gulf and its affiliates are not deductible as insurance premiums. With respect to the year 1975, we reach the same result for the same reasons, with one noted difference. We have considered Insco's unrelated business in order to determine whether it is sufficient to affect the premium pool such that risk transfer has occurred, and have concluded that we do not have enough facts to determine the same. ¹⁵ However, even without additional evidence we have concluded that the addition of 2 percent of unrelated premiums is de minimis and would not satisfy us that risk transfer has occurred. ¹⁶ We reserve judgment, however, on years other than those before the Court. ¹⁷

The second issue is whether the payments designated as premiums made by the foreign affiliates, and the payments of claims by Insco to Gulf and its domestic affiliates, represent constructive dividends to Gulf.

Under section 61(a)(7), gross income includes dividends. The term "dividend" is defined in section 316(a) as a distribution of property by a corporation to its shareholders out of its earnings and profits. There is no requirement that the dividend be formally declared or even intended by the corporation. See, e.g., Sachs v. Commissioner, 277 F.2d 879">277 F.2d 879 (8th Cir. 1960), affg. 32 T.C. 815">32 T.C. 815 (1959). Distributions by a corporation will be treated as dividends to the shareholder if the distributions are made for the shareholder's personal benefit; the funds need not be distributed directly to the shareholder. Rushing v. Commissioner, 441 F.2d 593">441 F.2d 593 (5th Cir. 1971), affg. 52 T.C. 888">52 T.C. 888, 893 (1969); Rapid Electric Co. v. Commissioner, 61 T.C. 232">61 T.C. 232, 239 (1973).

Because we have decided that the arrangement between Gulf and Insco does not represent insurance, respondent contends that payments of claims by Insco to Gulf represent constructive dividends to Gulf. We disagree. Our holding is limited to a finding that the arrangement between Gulf and Insco is not insurance and, therefore, the payments are not deductible for Federal income tax purposes. This does not alter the fact that a captive insurance company may be a useful and legitimate tool in risk management. Insco was organized and operated to provide Gulf with sufficient protection for certain risks of loss, as well as to obtain tax benefits which we have disallowed. The payments of claims by Insco to Gulf were made in consideration for the premiums paid by Gulf and ceded to Insco.

Respondent also contends that the payments designated as premiums made by the foreign affiliates, and the payments of claims by Insco to the domestic affiliates, represent constructive dividends to Gulf. It is clear that a transfer of property from one corporation to another corporation can constitute a constructive dividend to their common shareholder. Sammons v. Commissioner, 449">472 F.2d 449 (5th Cir. 1972), affg. in part, revg. in part, and remanding a Memorandum Opinion of this Court. However, before we will characterize a transfer of property from one corporation to another as a constructive dividend to a common owner, we must first examine the transfer under a two-part test set forth in Sammons v. Commissioner, supra.See Gulf Oil Corp. v. Commissioner, 87 T.C. 548 (1986).

The first part of the test is objective, i.e., did the transfer cause funds or other property to leave the control of the transferor corporation, and did it allow the stockholder to exercise control over such funds or property either directly or indirectly through some instrumentality other than the transferor corporation. The first part of the test has been satisfied. The funds were transferred from foreign affiliates to Insco in the form of insurance premiums, and from Insco to the domestic affiliates as payments of claims. As is typical in these cases, the critical inquiry is whether the second part of the Sammons test has been met.

The second part of the test is subjective -- whether the transfer was prompted by a business purpose of the transferor corporation or a shareholder purpose of the common owner. The transferor corporations had a business purpose for making the transfers and, therefore, we hold that the second part of the test has not been satisfied. Although payments made by the foreign affiliates to Insco are not classified as deductible insurance premiums, nevertheless, such payments were for the benefit of the affiliates because they provided coverage for their risks as separate entities. While the payments made to this captive insurance company are equivalent to additions to a reserve for losses, Insco, nevertheless, represents a useful and legitimate tool in risk management. The same rationale applies to the payments of claims by Insco to the domestic affiliates. The payments were for the primary benefit of the affiliate which received them, not for the benefit of the parent, Gulf.

Accordingly, we hold that the payments designated as premiums made by the foreign affiliates, and the payments of claims by Insco to Gulf and its domestic affiliates, do not represent constructive dividends to Gulf. See Mobil Oil Corp. v. United States, 8 Cl. Ct. 555 (1985).

An appropriate order will be issued.

APPENDIX

Discussion of Prior Captive Insurance Cases

In Carnation the taxpayer corporation insured its risks with an unrelated company, which in turn reinsured 90 percent of those risks with the corporation's wholly owned captive. The captive was capitalized with $ 120,000, the requisite legal minimum in Bermuda. Before agreeing to this insurance and reinsurance arrangement, the unrelated insurer required Carnation to agree to increase the captive's capital up to $ 3 million, if necessary, for the captive to meet its reinsurance obligations. We determined that this arrangement was not insurance, and that therefore premiums paid to the third-party insurer which were ceded to the captive were not deductible. Relying on the Supreme Court decision in Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531 (1941), we held that the contracts, when taken together, were void of risk because the agreement to increase the captive's capital bound the taxpayer to an investment risk that was directly tied to the loss experience of its captive, which was, in turn, wholly contingent on the taxpayer's losses. The Ninth Circuit affirmed our decision agreeing that as in Le Gierse, the agreements effectively neutralized risk for purposes of insurance. Carnation Co. v. Commissioner, 640 F.2d at 1013.

Le Gierse is based on the principle that an agreement may resemble insurance in form, yet lack an essential ingredient to insurance. Carnation Co. v. Commissioner, 71 T.C. at 415. This principle applies regardless of whether the parent and captive are considered separate entities for tax purposes. Carnation Co. v. Commissioner, 71 T.C. at 408.

In Clougherty, the operative facts were essentially the same as in Carnation, although no collateral capitalization agreement was present. Clougherty Packing Co. v. Commissioner, 84 T.C. 948">84 T.C. 948 (1985), affd. 811 F.2d 1297">811 F.2d 1297 (9th Cir. 1987). Regardless of this distinction, we held that the arrangement was not insurance; the risk of loss was not shifted from the parent to its captive. We stated "to deduct insurance premiums it is essential to decide whether the relationship giving rise to the payment constitutes insurance." Clougherty Packing Co. v. Commissioner, 84 T.C. at 957. We held that the payments to the captive for premiums, which were used to pay losses of the parent and its affiliates only, were not deductible; they were essentially the same as nondeductible payments to a reserve for losses. We reasoned that payments are not insurance premiums unless they are paid to shift the risk away from the taxpayer who seeks to deduct them. Clougherty Packing Co. v. Commissioner, 84 T.C. at 958. We declined to decide how the result might be affected, however, if the captive had business from unrelated customers. Clougherty Packing Co. v. Commissioner, 84 T.C. at 960.

The Ninth Circuit affirmed our result, but on the theory that the risk of loss was not shifted, because the losses paid by the subsidiary reduce the value of its stock, and thus, correspondingly reduced the value of the corporation's assets. Clougherty Packing Co. v. Commissioner, 811 F.2d 1297">811 F.2d 1297 (9th Cir. 1987). ¹

Humana v. Commissioner, 88 T.C. 197">88 T.C. 197 (1987), presented a slightly different fact pattern, but we again reached the same result. In Humana, both the parent and its affiliates paid premiums to the parent's captive, again a wholly owned subsidiary that insured only the risks of the affiliated group. We held that the tax consequences of the payments by the parent to the captive were controlled by our prior decisions in Carnation and Clougherty. Humana v. Commissioner, 88 T.C. at 207. With respect to the affiliates, the situation was no longer parent-subsidiary but instead was brother-sister. We reached the same result, however, extending the rationale of Carnation and Clougherty, and reclassifying the payments as nondeductible additions to a reserve for losses. Humana v. Commissioner, 88 T.C. at 209. We stated "If we decline to extend our holdings in Carnation and Clougherty to the brother-sister factual pattern, we would exalt form over substance and permit a taxpayer to circumvent our holdings by simple corporate structural changes." Humana v. Commissioner, 88 T.C. at 213. We thus concluded that there was not the necessary shifting of risk from the operating subsidiaries of the parent to the captive, and therefore the arrangement was not insurance. Humana v. Commissioner, 88 T.C. at 214.

GOFFE; CHABOT; NIMS

Goffe, J. concurring: In this case, there are two taxable years before the Court, 1974 and 1975. The issues in each of the years is whether petitioner may deduct amounts paid as insurance premiums by Gulf and its domestic affiliates, to the extent that those payments were ceded to its wholly owned captive insurance company, Insco, and whether payments designated as premiums paid by the foreign affiliates of Gulf, which were ceded to Insco, and the claims paid by Insco to Gulf and its domestic affiliates represent constructive dividends to Gulf. The majority holds for petitioner on the second issue, and I agree with the reasoning of the majority. The majority holds for respondent on the first issue, and with respect to the taxable year 1974, bases its holding upon our prior decision, with which I agree. However, with respect to the taxable year 1975, the majority adopts a new and novel theory, not argued, briefed, or even contemplated by the parties, with which I do not agree.

The theory of the majority can be stated as follows: "adequate" risk distribution, created by insuring the risks of independent third parties, somehow translates into risk transfer. No court has ever suggested such a theory, nor have the parties advanced such an argument. The majority holds that the addition of 2 percent of unrelated premiums is de minimis and would not satisfy the majority that the risk was transferred. However, in dicta, the majority goes on to state that if the premium income from unrelated parties is at least 50 percent, there would be sufficient risk transfer so that the arrangement would constitute insurance.

The theory of the majority is in direct conflict with all of the case law which holds that the arrangement of insurance requires a shifting of the risk and a distribution of the risk. The novel theory of the majority blurs the distinction between the concepts of shifting of risk and distribution of risk. However, such concepts cannot be equated because they have historically been held to be separate and distinct concepts.

In Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531 (1941), the Supreme Court first stated the rule as follows:

We think the fair import of subsection (g) is that the amounts must be received as the result of a transaction which involved an actual "insurance risk" at the time the transaction was executed. Historically and commonly insurance involves risk-shifting and risk-distributing. That life insurance is desirable from an economic and social standpoint as a device to shift and distribute risk of loss from premature death is unquestionable. That these elements of risk-shifting and risk-distributing are essential to a life insurance contract is agreed by courts and commentators. See for example: Ritter v. Mutual Life Ins. Co., 169 U.S. 139">169 U.S. 139; In re Walsh, 19 F. Supp. 567">19 F. Supp. 567; Guaranty Trust Co. v. Commissioner, 16 B.T.A. 314">16 B.T.A. 314; Ackerman v. Commissioner, 15 B.T.A. 635">15 B.T.A. 635; Couch, Cyclopedia of Insurance, Vol. I, § 61; Vance, Insurance, §§ 1-3; Cooley, Briefs on Insurance, 2d edition, Vol. I, p. 114; Huebner, Life Insurance, Ch. 1. Accordingly, it is logical to assume that when Congress used the words "receivable as insurance" in § 302(g), it contemplated amounts received pursuant to a transaction possessing these features. Commissioner v. Keller's Estate, supra; Helvering v. Tyler, supra; Old Colony Trust Co. v. Commissioner, 102 F.2d 380">102 F.2d 380; Ackerman v. Commissioner, supra. [Helvering v. Le Gierse, 312 U.S. at 539-540.]

The distinction between risk shifting and risk distributing is most succinctly described in the following quotation from an early case, Commissioner v. Treganowan, 183 F.2d 288">183 F.2d 288 (2d Cir. 1950), which has been uniformly followed by the Tax Court and other courts:

"Insurance" is not defined by statute, and Treasury interpretation of the term has never gone beyond a statement that § 811(g) is applicable to "insurance of every description, including death benefits paid by fraternal beneficial societies operating under the lodge system." Treas. Reg. 105, § 81.25, 26 C.F.R. 81.25. In deciding whether the $ 20,000 paid to decedent's widow is insurance within the statutory meaning, the Tax Court looked, rather naturally, to the test announced in the leading case of Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531, 539, 61 S. Ct. 646">61 S.Ct. 646, 649, 85 L. Ed. 996">85 L.Ed. 996, that "historically and commonly insurance involves risk-shifting and risk-distributing." * * *

* * * As a critical commentator on the decision below well states it: "Risk shifting emphasizes the individual aspect of insurance: the effecting of a contract between the insurer and insured each of whom gamble on the time the latter will die. Risk distribution, on the other hand, emphasizes the broader, social aspect of insurance as a method of dispelling the danger of a potential loss by spreading its cost throughout a group. By diffusing the risks through a mass of separate risk shifting contracts, the insurer casts his lot with the law of averages. The process of risk distribution, therefore, is the very essence of insurance." Note, The New York Stock Exchange Gratuity Fund: Insurance That Isn't Insurance, 59 Yale L.J. 780">59 Yale L.J. 780, 784.

[Commissioner v. Treganowan, 183 F.2d at 290-291.]

The distinction between risk shifting and risk distributing was also recognized in Crawford Fitting Co. v. United States, 606 F. Supp. 136">606 F. Supp. 136 (N.D. Ohio 1985). That is the only captive insurance case in which a court analyzed both concepts. The District Court applied the definitions contained in Commissioner v. Treganowan, supra.It held that the payments were deductible as insurance premiums because there were both a shifting of risk and a distribution of risk. Some of the more pertinent quotations from that opinion are as follows:

"Insurance" is not defined under the Internal Revenue Code. The Supreme Court, however, has stated that elements of risk-shifting and risk-distributing, historically and commonly, have been considered ordinary indicia of insurance. Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531, 61 S. Ct. 646">61 S.Ct. 646, 85 L. Ed. 996">85 L.Ed. 996 (1941).

Where elements of risk-shifting and risk-distributing are absent from an insurance contract, the fundamental aspects of an insurance agreement are lacking. See Steere Tank Lines, Inc. v. U.S., 577 F.2d 279">577 F.2d 279 (5th Cir. 1978), cert. denied, 440 U.S. 946">440 U.S. 946, 99 S. Ct. 1424">99 S.Ct. 1424, 59 L. Ed. 2d 634">59 L.Ed.2d 634 (1979). In that case, the Court stated that "a taxpayer who establishes a reserve in an effort to ensure against future losses is not entitled to a deduction at the time the reserve is established or funded. A deduction is proper when the liability becomes fixed." Id. at 282. Accordingly, the Court found that where there was no shifting or genuine pooling of risks under the insurance agreement before it, plaintiff's payment of money into a contract premium account, from which were paid all accident claims against the taxpayer, was not an insurance premium deductible as a business expense, but rather a nondeductible reserve for accident claims. See also Spring Canyon Coal Co. v. Commissioner, 43 F.2d 78">43 F.2d 78 (10th Cir. 1930), cert. denied, 284 U.S. 654">284 U.S. 654, 52 S. Ct. 33">52 S.Ct. 33, 76 L. Ed. 555">76 L.Ed. 555 (1931) (Corporation's payments into self-insurance reserve fund held not deductible as an "ordinary and necessary business expense"). A taxpayer cannot "deduct as an expense an amount which he fears he may some day be called upon to spend," such as self-insurance. See Id. at 80, citing Appeal of Pan-American Hide Co., 1 B.T.A. 1249">1 B.T.A. 1249 (1925).

The fundamental issue in the case at bar, then, is whether the payment made by the plaintiff Crawford Fitting Company to the Constance Insurance Company, characterized by the plaintiff as an insurance premium, was in fact an insurance premium deductible under § 162, or was, as the defendant contends, a reserve held by the plaintiff Crawford Fitting Company as self-insurance to cover contingent losses.

[Crawford Fitting Co. v. United States, 606 F. Supp. at 140. Fn. refs. omitted.]

The Court of Appeals for the 10th Circuit in Beech Aircraft Corp. v. United States, 797 F.2d 920">797 F.2d 920 (10th Cir. 1986), described the difference between risk shifting and risk distributing as follows:

In Helvering v. Le Gierse, the Supreme Court observed that "[historically] and commonly insurance involves risk-shifting and risk-distributing." 312 U.S. 531">312 U.S. 531, 539, 61 S. Ct. 646">61 S.Ct. 646, 649, 85 L. Ed. 996">85 L.Ed. 996 (1941) (citing, inter alia, Ritter v. Mutual Life Insurance Co., 169 U.S. 139">169 U.S. 139, 18 S. Ct. 300">18 S.Ct. 300, 42 L. Ed. 693">42 L.Ed. 693 (1898)). "Risk-shifting" means one party shifts his risk of loss to another, and "risk-distributing" means that the party assuming the risk distributes his potential liability, in part, among others. An arrangement without the elements of risk-shifting and risk-distributing lacks the fundamentals inherent in a true contract of insurance. See Steere Tank Lines, Inc. v. United States, 577 F.2d 279">577 F.2d 279, 280 (5th Cir. 1978), cert. denied, 440 U.S. 946">440 U.S. 946, 99 S. Ct. 1424">99 S.Ct. 1424, 59 L. Ed. 2d 634">59 L.Ed.2d 634 (1979); Commissioner v. Treganowan, 183 F.2d 288">183 F.2d 288, 291 (2d Cir.), cert. denied, 340 U.S. 853">340 U.S. 853, 71 S. Ct. 82">71 S.Ct. 82, 95 L. Ed. 625">95 L.Ed. 625 (1950). * * * [Beech Aircraft Corp. v. United States, 797 F.2d at 922.]

The Court of Appeals for the Ninth Circuit, in affirming our opinion in Clougherty Packing Co. v. Commissioner, 84 T.C. 948 (1985), affd. 811 F.2d 1297">811 F.2d 1297 (9th Cir. 1987) discussed both concepts as follows:

Shifting risk entails the transfer of the impact of a potential loss from the insured to the insurer. If the insured has shifted its risk to the insurer, then a loss by or a claim against the insured does not affect it because the loss is offset by the proceeds of an insurance payment. See Beech Aircraft, 797 F.2d at 922;Treganowan, 183 F.2d at 291; O'Brien & Tung, Captive Off-Shore Insurance Corporations, 31 N.Y.U. Inst. 665, 683-84 (1973). Distributing risk allows the insurer to reduce the possibility that a single costly claim will exceed the amount taken in as a premium and set aside for the payment of such a claim. Insuring many independent risks in return for numerous premiums serves to distribute risk. By assuming numerous relatively small, independent risks that occur randomly over time, the insurer smoothes out losses to match more closely its receipt of premiums. See Beech Aircraft, 797 F.2d at 922; Treganowan, 183 F.2d at 291. Risk distribution incorporates the statistical phenomenon known as the law of large numbers. This law is reflected in the financial world by the diversification of investment portfolios and in the day-to-day world by the adage "Don't put all your eggs in one basket." See T. Wonnacott & R. Wonnacott, Introductory Statistics for Business & Economics 31-32, 54-55 (1972). ⁵ [Clougherty Packing Co. v. Commissioner, 811 F.2d at 1300.]

The majority in the instant case should have followed the example of the Court of Appeals for the Ninth Circuit when it affirmed this Court in Clougherty, i.e., there has been no shifting of risk; therefore, an analysis of distribution of risk is unnecessary.

The most recent discussion by the Tax Court emphasizing the distinction between risk shifting and risk distributing is Anesthesia Service Medical Group v. Commissioner, 85 T.C. 1031">85 T.C. 1031 (1985), affd. 825 F.2d 241">825 F.2d 241 (9th Cir. 1987). We described the distinction as follows:

Expenditures are deductible as insurance expenses, however, only if a true insurance arrangement exists. Clougherty Packing Co. v. Commissioner, 84 T.C. 948">84 T.C. 948 (1985); Carnation Co. v. Commissioner, 71 T.C. 400">71 T.C. 400 (1978), affd. 640 F.2d 1010">640 F.2d 1010 (9th Cir. 1981). An insurance arrangement, in turn, necessarily involves both the shifting and the distributing of risk. Helvering v. Le Gierse, 312 U.S. 531">312 U.S. 531, 539-540 (1941); Commissioner v. Treganowan, 183 F.2d 288">183 F.2d 288, 290 (2d Cir. 1950); Tighe v. Commissioner, 33 T.C. 557">33 T.C. 557, 564 (1959).

"Risk shifting emphasizes the individual aspect of insurance: the effecting of a contract between the insurer and insured each of whom gamble on time the * * * [loss] will * * * [occur]. Risk distribution, on the other hand, emphasizes the broader, social aspect of insurance as a method of dispelling the danger of a potential loss by spreading its costs throughout a group. By diffusing the risks through a mass of separate risk shifting contracts, the insurer casts his lot with the law of averages. * * * [Commissioner v. Treganowan, 183 F.2d at 291.]"

Thus, there can exist no risk distribution without risk shifting. * * *

[Anesthesia Service Medical Group v. Commissioner, 85 T.C. at 1038-1039.]

The case law set forth above has uniformly held that, in order for insurance to exist, risk transfer and risk distribution must be present. The case law has consistently treated risk transfer and risk distribution as two separate and distinct concepts. Nonetheless, the majority, citing no authority, concludes that risk distribution which includes third-party risks may be equivalent to risk transfer. Such a holding does violence to the principles of the entire line of cases emanating from Le Gierse, decided by the Supreme Court in 1941.

The majority opinion, in addition to being contrary to all of the case law, is contrary to the record made in this case. The transcript on this issue was 416 pages in length, 215 of which covered the testimony of expert witnesses whose opinions are contrary to the theory of the majority. Petitioner offered no expert witnesses. Respondent's expert witnesses were Dr. Irving Plotkin and Mr. Richard Stewart. They submitted written reports and testified at length on direct examination. There was considerable cross-examination. As the trier of fact, I found their testimony to be credible, convincing, and unassailable. Furthermore, I asked numerous questions which they answered to my satisfaction. I can find nothing in the entire record to cast any doubt on their testimony. Their opinions unequivocally draw the distinction between risk distribution and risk transfer, and emphasize that risk distribution cannot be equated with risk transfer.

The majority, nevertheless, states in note 14 on page 1027:

Without expert testimony we decline to determine what proportion of unrelated premiums would be sufficient for the affiliated group's premiums to be considered payments for insurance. However, if at least 50 percent are unrelated, we cannot believe that sufficient risk transfer would not be present. This anticipated sharing of premiums paid by unrelated insureds is similar in concept to a mutual insurance arrangement. See, e.g., Rev. Rul. 80-120, 1 C.B. 41">1980-1 C.B. 41; Rev. Rul. 78-338, 2 C.B. 107">1978-2 C.B. 107.

I am at a loss to understand how such a statement can be made. The footnote implies that there was no expert testimony presented to me, which implication is contrary to the record in this case. Not only was expert testimony offered, but the experts who testified were Dr. Plotkin and Mr. Stewart. They are the only experts whose testimony this Court and other courts have adopted.

Dr. Plotkin is a vice president of Arthur D. Little, Inc., and a director of its valuation subsidiary. He holds a B.S. degree in economics from the Wharton School (University of Pennsylvania) and a Ph.D. in mathematical economics from the Massachusetts Institute of Technology, where he taught computer science and finance. He has presented numerous papers in several fields of economics and regulation before various academic societies, professional organizations, and industrial associations, and has served as an editorial reviewer for the Journal of the American Statistical Association, the Journal of Industrial Economics, and the Journal of Risk and Insurance. His published papers and monographs include "Risk/Return: U.S. Industry Pattern" (Harvard Business Review); "Rates of Return in the Property and Liability Insurance Industry: A Comparative Analysis" (Journal of Risk and Insurance); The Consequences of Industrial Regulation on Profitability, Risk Taking, and Innovation; and Torrens in the United States.

Dr. Plotkin's expert report contains the following:

the concept and operation of the principles of risk distribution are independent of and unrelated to the concept of risk transfer.

* * * *

To have a true transfer of risk, another risk-bearer must replace the insured. To speak of a transfer of risk to a fund or reserve established by the insured is merely to describe "self-insurance." A captive insurance subsidiary, such as Insco, represents a recognized form of funding risk retention or self-insurance. * * * there is no dispute that a captive insurance company represents a device for funding retained risks.

* * * *

* * * whether * * * insurance was provided [to unrelated parties] does not in any way affect the fact that the transactions between the parent, its subsidiaries, and the captive did not provide insurance. The fact that a captive insurance company accepts the financial consequences of the risks of persons or firms unrelated to it by ownership, does not change the relationship between it and its affiliates from formalized, funded risk retention to insurance (risk transfer). By accepting third party business, the captive and its owners do not relieve themselves of the uncertainty stemming from their retention of their own risks. Rather they accept the uncertainty inherent in the third parties' risks in addition to those inherent in their own risks.

Dr. Plotkin's expert report clearly states that the existence of third-party risks does not translate into transfers of risk from Gulf and its domestic affiliates to Insco. Simply put, a transfer of risk does not occur by reason of adequate distribution of risk created by the presence of the insured risks of independent third parties. The majority directly contradicts Dr. Plotkin's expert report and offers absolutely no explanation for its complete disregard of it.

We adopted the opinion of Dr. Plotkin in Humana v. Commissioner, 88 T.C. 197">88 T.C. 197 (1987), and he likewise rendered similar opinions which were adopted in Beech Aircraft Corp. v. United States, an unreported case ( D. Kan. 1984, 54 AFTR 2d 84-6173, 84-2 USTC par. 9803), affd. 797 F.2d 920">797 F.2d 920 (10th Cir. 1986); Stearns-Roger Corp. v. United States, 577 F. Supp. 833">577 F. Supp. 833 (D. Colo. 1984), affd. 774 F.2d 414">774 F.2d 414 (10th Cir. 1985); and Mobil Oil Corp. v. United States, 8 Cl. Ct. 555">8 Cl. Ct. 555 (1985). He testified before me that, with respect to the subject matter involved, he had spent 500 to 1,000 hours in research. The bibliography accompanying his expert report lists 22 books, 7 monographs, and 55 articles. The majority cites 1 Supreme Court case, ¹ 4 books, 4 articles, and l private letter ruling for various statements. However, the majority does not cite any of these as authority for its theory.

Mr. Richard Stewart also testified on behalf of respondent. We likewise adopted his opinion in Humana and he has imposing credentials. Mr. Stewart was graduated from West Virginia University in 1955, where he was President of the student body and the holder of the highest academic record in the history of the institution. As a Rhodes Scholar at Queen's College, Oxford, he was president of the student body, and of the law society, and received a B.A. degree in jurisprudence, with congratulatory first class honors, in 1957. In 1959, he was graduated, cum laude, from Harvard Law School. From 1967 to 1970, he was Superintendent of Insurance of New York State, and in 1970, he was president of the National Association of Insurance Commissioners. From 1971 to 1972, he was senior vice president and general counsel of First National City Bank and its parent, First National City Corp., now Citibank and Citicorp, and a member of the policy committee of both corporations. He was also a director of General Reinsurance Corp. From 1973 to 1981, he was senior vice president and chief financial officer of the Chubb Corp. and of its subsidiaries and senior vice president and director of Chubb & Son, Inc. His responsibilities included personal lines underwriting, surety bonding, financial institutions underwriting, corporate finance, investments, and various staff functions. From 1979 to mid-1981, he was also a founding governor of the New York Insurance Exchange. Since mid-1981 he has been chairman of Stewart Economics, Inc., a consulting firm specializing in the insurance business. He is the author of Reason and Regulation (1972) and Insurance and Insurance Regulation (1980) and co-author of Automobile Insurance . . . For Whose Benefit? (1970), and Medical Malpractice (1977).

Mr. Stewart's expert report, submitted to me at trial, contains the following:

For there to be insurance, in both its textbook and its practical business sense, there must be transfer of risk, that is, a shifting of the financial consequences of an event, where the event or its financial consequences, or both, are uncertain at the time the insurance arrangement is made. In short, insurance involves the transfer of financial risk.

* * * *

* * * what it all comes down to is whether there was transfer of financial risk. There was not. The reason is that Gulf owned Insco.

In response to questions at trial concerning an example of a parent which owned 100 percent of a captive, which insured third-party risks, Mr. Stewart testified that there could be no transfer of risk from the parent to the wholly owned captive.

The only two experts who testified before me in the instant case agreed that there was no shifting of risk and that the presence of third-party risks was not relevant in determining whether there had been a shifting of risk from Gulf and its domestic affiliates to Insco. Because there was no transfer of risk, the case should be decided on that basis under the authority of our prior opinions in Humana, Clougherty, and Carnation Co. v. Commissioner, 71 T.C. 400">71 T.C. 400 (1978), affd. 640 F.2d 1010">640 F.2d 1010 (9th Cir. 1981). The new and novel theory of the majority reaches the same result with respect to the taxable year 1975. However, its application will lead to incorrect results when there is a higher percentage of insured risks of independent third parties. More importantly, this theory is in direct conflict with the written reports and testimony of the eminently qualified experts who testified before me and in other trials.

Furthermore, the theory of the majority, in addition to being contrary to the case law and the expert reports and testimony, is seriously flawed in its reasoning. The majority attempts to explain how risk distribution, created by the presence of the insured risks of third parties, is somehow transformed into risk transfer on pages 1026-1027:

The premium allocable to each insured contributes to the payment of aggregate losses whether or not that insured actually suffers any loss. * * *

* * * When a sufficient proportion of premiums paid by unrelated parties is added, the premiums of the affiliated group will no longer cover anticipated losses of all of the insureds; the members of the affiliated group must necessarily anticipate relying on the premiums of the unrelated insureds in the event that they are "the unfortunate few" and suffer more than their proportionate share of the anticipated losses.

Thus, when the aggregate premiums paid by the captive's affiliated group is insufficient in a substantial amount to pay the aggregate anticipated losses of the entire group, * * * [risk] distribution and risk transfer would be present * * *

[Fn. refs. omitted.]

It appears that the majority equates risk transfer with the ability of the affiliated group to use the premium income of the unrelated parties to pay for its losses. If the premium allocable to each insured contributes to the payment of aggregate losses, it would seem that the logical conclusion should be that the affiliated group has shifted its risk to the extent of unrelated premiums. For example, if the unrelated parties account for 25 percent of the premium income, it would seem that the affiliated group has shifted its risk to the extent of 25 percent. However, under the majority's theory, if the aggregate premiums paid by the corporations affiliated with the captive are "sufficient" in a "substantial" amount to pay the aggregate anticipated losses of all of the insureds there is no transfer of risk, but if the aggregate premiums paid by the corporations affiliated with the captive are "insufficient" in a "substantial" amount to pay the aggregate anticipated losses of the entire group, there is 100-percent transfer of risk. I question why, under the majority's theory, some magical percentage of unrelated premium is critical to transform risk distribution into risk transfer.

Finally, the analysis of the majority is incomplete in that it fails to consider the possibility that the unrelated parties may be "the unfortunate few" and suffer more than their proportionate share of the anticipated losses. In this instance, premium income of the affiliated group may be used to cover the losses of the unrelated parties. For example, assume the affiliated group pays $ 50 of premiums to the captive and that the captive also receives $ 50 of premiums from unrelated parties. If the affiliated group incurs $ 60 of losses, the affiliated group will be able to use $ 10 of the premiums of the unrelated parties to cover its losses. On the other hand, if the unrelated parties incur $ 60 of losses, the unrelated parties will be able to use $ 10 of premiums of the affiliated group to cover its losses. This scenario illustrates the point contained in Dr. Plotkin's expert report that by accepting third-party business, the captive and its owners accept the uncertainties inherent in the third parties' risks. If we examine the whole picture and not just a portion of it, the theory of the majority does not withstand scrutiny.

I do not question the power of the Court to adopt a new theory. However, even ignoring my analysis above, I question the wisdom of adopting such a new and novel theory, which heretofore has never been considered, when neither party has been given an opportunity to brief the merits, if any, of the theory. Respondent has ably refuted petitioner's claim that the payments in issue constituted insurance premiums. We should not assume that he has no argument against the new and novel theory advanced by a majority of the Court. More importantly, the majority can only define and describe its theory in generalities. The theory of the majority will undoubtedly have a major impact, and I question the sagacity of adopting such a theory sua sponte. Indeed, this Court should heed the admonition of the Court of Appeals for the Fifth Circuit in Brooks v. Commissioner, 424 F.2d 116">424 F.2d 116 (5th Cir. 1970), revg. 50 T.C. 927">50 T.C. 927 (1968), when it said:

The Tax Court based its decision on a novel theory * * * which was not raised, briefed or argued by either party. * * * In the usual circumstances we would reverse and remand so that * * * both parties may be given an opportunity to brief and argue the merits of the new theory. * * * [424 F.2d at 119.]

Chabot, J., concurring: Both the majority and Judge Goffe agree with the result that petitioner is not entitled to any deduction for insurance premiums for the years before us in the instant case.

If, as appears to be the case, the majority are unwilling to agree with Judge Goffe's extension of prior case law as the justification for this result, then it would be the more prudent course of action to (1) note the "law of large numbers" analysis, (2) observe that that analysis would lead to the same result in the instant case but perhaps different results in other cases, and (3) indicate that the parties in other cases may wish to argue in their cases the strengths and weakness of that analysis and present the appropriate expert testimony. This approach would have preserved our opportunity to carefully evaluate a well-grounded exposition of both sides regarding the "law of large numbers" analysis. See Concord Consumers Housing Cooperative v. Commissioner, 89 T.C. 105">89 T.C. 105, 106 n. 3 (1987), and the concurring opinion by Judge Korner, 89 T.C. at 125.

Instead, the majority have committed us to certain decisions in cases not before us (p. 1027), decisions that might not be justified by the records made in those cases. As Judge Parker has indicated, we do a disservice to other taxpayers when we so boldly announce the result of future cases under these circumstances.

Nims, J., concurring: I concur with the majority in both reasoning and result. However, I find it necessary to comment separately with respect to several of the concerns expressed by my colleague, Judge Goffe.

In essence, Judge Goffe would have us determine that a group of affiliated corporations can never deduct premiums paid to a wholly owned captive insurer. The insurance of unrelated third parties, he concludes, is irrelevant because there is never shifting of risk from a parent and its affiliates to their wholly owned captive. Because there is no transfer of risk, the arrangement can never be considered insurance. Judge Goffe rests his conclusion on the expert testimony and reports of Dr. Plotkin and Mr. Stewart, both of whom relied not only upon insurance principles, but also upon economic theory to arrive at their conclusion that Gulf and its affiliates did not transfer their risk to Insco.

Because the majority rejected the conclusions of Dr. Plotkin and Mr. Stewart with respect to the effect of the presence of unrelated insureds, Judge Goffe concludes, I believe incorrectly, that the majority "directly contradicts Dr. Plotkin's expert report and offers absolutely no explanation for its complete disregard of it." The majority did not completely disregard the expert reports in this case. In fact, the majority agrees with the experts with respect to their description of the necessary elements of insurance ¹ and their conclusion that a captive insurer represents a form of risk retention or self-insurance -- when the captive insures only risks of its affiliates. The majority even quotes from the same insurance text relied upon by Dr. Plotkin for the proposition that a captive is a hybrid possessing characteristics of insurance and self-insurance. The majority disagrees, however, with Dr. Plotkin and Mr. Stewart's ultimate conclusion that a wholly owned captive is always a risk-retention device and that the addition of third party risks does not change the relationship between the captive and its affiliates from one of risk retention to risk transfer. ²

One is compelled to conclude that Judge Goffe, Dr. Plotkin, and Mr. Stewart can only reach their conclusion by adopting the economic family theory, an approach not based on insurance principles, but rather on the principle that the capital of the affiliates remains at risk, regardless of the identity of the captive's insureds, because the affiliates and the captive are related. ³ As the majority points out, we have rejected this theory in the past and we do not, despite Judge Goffe's view to the contrary, adopt that theory here simply because the experts rely on it.

Instead, the majority relies on principles of the insurance industry to distinguish between arrangements which are in essence self-insurance and arrangements which will be recognized as insurance for tax purposes. The majority recognizes that there are circumstances under which a captive should be treated as a separate corporation providing insurance to its affiliates and premiums paid by the affiliates should be deductible.

I also disagree with Judge Goffe that the theory of the majority is "new and novel" and is in direct conflict with case law holding that insurance requires risk shifting and risk distribution.

Gulf's principal arguments throughout its briefs are that Insco should be treated as a separate corporate entity and that the agreements entered into by Gulf, Gulf's affiliates, and Insco are insurance, under the definition set forth in Helvering v. Le Gierse, for which premiums paid should be deductible. ⁴Gulf argues that the agreements resulted in a transfer of risk because Insco was financially capable of meeting its obligations and that therefore Insco, rather than Gulf and its affiliates, bore the risk of loss. Gulf further argues that risk distribution was achieved both by the types of risks Insco insured and the parties who were insured. Gulf maintains that this should be true regardless of whether the captive insured unrelated parties. Alternatively, it contends that the insurance of unrelated third parties is relevant because --

Insco's intention coupled with its carrying through on such intention support the conclusion that Insco was providing insurance in the taxable years in issue, even for the portion of such insurance which related to risks of Gulf and its affiliates.

As an example of how the insurance of unrelated third parties makes a difference, Gulf refers to O.M. 19167, which was transmitted by G.C.M. 38136 and which was withdrawn by G.C.M. 39247:

O.M. 19167 considered the insurance of significant third party risks * * * to compel a conclusion that the transactions entered into by the insurance subsidiary should be considered "insurance" in such years even with respect to those transactions involving risks of companies which were related to the insurance affiliate. Furthermore, while O.M. 19167 expressed the opinion that there was not "insurance" in [earlier years] due to the fact that the percentage of third party risks that were insured in such years were minimal, O.M. 19167 considered that such years might be viewed by a court, not "in isolation, but as the start-up phase of what will become an insurance business in the fullest sense * * *"

Respondent's "one economic family" theory does not take account of the fact that an insurance subsidiary may intend to insure (and may in fact insure) the risks of third parties. * * *

It is apparent from the above that one aspect of Gulf's argument is that the existence of unrelated insureds should make a difference in our determination of whether the arrangements with Insco were insurance. Gulf contends that the existence of unrelated insureds is evidence that Insco is a separate insurance company. The majority, agreeing with Gulf on this point, explains further that the existence of unrelated insureds allows the Court to distinguish between arrangements which are in substance self-insurance and arrangements which are insurance. It is with respect to this point that the majority offers its reasoning that when a substantial percentage of the captive's insureds are unrelated, risk transfer is present and will be recognized because the risks of the affiliated group are transferred to the unrelated insureds through the captive's premium pool.

Contrary to Judge Goffe's conclusion, the majority does not blur the distinction between the concepts of shifting of risk and distribution of risk. The majority requires that both be present. The majority simply recognizes that both may be present when the captive insures a substantial percentage of unrelated insureds and when the law of large numbers operates so that the premiums are reasonably calculated to cover losses. The majority's reasoning is not unlike that set forth in G.C.M. 38316, which was cited by Gulf on brief:

In short, while it is settled that for an "offshore captive" to be truly an insurer it must serve to shift and distribute risks outside of the corporate group in a substantive way, we must look to both policyholders and shareholders to determine if the requisite risk shifting and distribution is present. Under insurance theory, risks are shifted and distributed not through the capital structure of the company, but rather through the premiums (and resulting surplus and investment income) paid by the policyholders. Thus, we believe that, although there may be no ownership of an insurer outside of an affiliated group, we must still look to the percentage of nonaffiliated policyholders that the "captive" has, and/or the relative dollar value of premiums paid by nonaffiliated policyholders to determine the presence or absence of insurance.

With regard to the instant case, your memorandum suggests that although risk is distributed by the presence of nonaffiliated policyholders, the element of risk shifting is absent, because the parent is shifting its insurable risk to an "insurance" company for which the parent provided the basic capitalization. Looking to the capital structure of the subsidiary alone to determine whether risks are shifted from the insured to the insurer, however, ignores a basic premise of the insurance business, that in a normal insurance situation risk shifting and distributing is accomplished through the premium-based underwriting operation. In other words, claims are paid by a solvent insurer not from paid in capital, but rather from premium income, investment income, and surplus. See D. Gregg and V. Lucas, Life and Health Insurance Handbook 140 (3d ed. 1973) and S. Huebner and R. Black, Life Insurance 6 (9th ed. 1976), both of which discuss the relationship of insurance benefits and premiums in the context of life insurance; 1 W. Freedman, Richards on the Law of Insurance 2, 83 (1952). Therefore, we believe that when, as in the instant case, risks have been distributed to other policy holders then it necessarily follows that the risk has also been shifted to those other policy holders. Thus, because in this case the other policyholders are not members of the affiliated group, there has been a shifting and distributing of risks outside the group, through the medium of the "captive" insurance company which, although largely owned by the parent, receives (at least in the latter two years) approximately half of the money from which it will pay claims from unrelated parties. [Emphasis added.]

Because Gulf's primary argument was that Insco should be treated as a separate corporation providing insurance it did not go to great lengths to argue that the unrelated insureds made the difference between an arrangement that lacked risk transfer and one that did not. However, it is apparent that Gulf did present and argue, albeit as an alternative, this theory.

For the foregoing reasons I respectfully disagree with Judge Goffe's conclusions.

PARKER

Parker, J., dissenting: I agree with Judge Goffe's views on the merits of the case, and normally would simply join in his concurring opinion. Here, however, I must dissent from what I perceive as the majority's unwarranted arrogation of judicial power.

I am mindful that this Court, although a trial court, functions as a collegial body. Nonetheless, such untrammeled exercise of majoritarian power as has occurred in this case could, I fear, threaten the necessary and wholesome judicial independence of the individual trial judge of this Court. Rather than serving as an instrument to assure uniform application and/or development of the tax law, the Court Conference could well become a committee of revision to try to force the individual trial judge to address some pet legal theory not presented by the case and not tried, argued, or briefed by the parties.

Judge Goffe, as the trial judge who heard this case, proposed to dispose of the case by following a well-established line of authority of this and other courts in the captive insurance area. He should have been permitted to do so. The majority does not disagree with that case law nor with the result flowing from application of that case law to the facts of this case. On the contrary, the majority reaches the same result but does so by a most unfortunate route. The majority wrested the case from the presiding trial judge, seizing it as a vehicle in which to expound a novel legal theory not presented by the pleadings, trial, arguments, and briefs of the parties, and without giving the litigants an opportunity to address this theory.

In my opinion, the majority is reaching out to decide an issue that is not before it. I view such practice as particularly mischievous in this case where there may well be little practicable incentive to seek appellate review. This novel theory does not necessarily come into operation on the facts and for the years before the Court, and the bottom-line result would be the same either way. A litigant may have little desire to incur the expense to appeal the basis on which he won or lost his case, where the outcome remains unchanged.

---

Footnotes

• *. This case was reassigned by order of the Chief Judge dated Nov. 24, 1987.↩

• 1. This is the final opinion on the severed issues. The other opinions are reported at: 87 T.C. 548">87 T.C. 548 (1986) (constructive dividends and payables); 87 T.C. 324">87 T.C. 324 (1986) (intangible drilling and development costs); 87 T.C. 135">87 T.C. 135 (1986) (worthless properties); 86 T.C. 937">86 T.C. 937 (1986) (Kuwait nationalization); 86 T.C. 115">86 T.C. 115 (1986) (Iranian foreign tax credit); 84 T.C. 447">84 T.C. 447↩ (1985) (North Sea farmout).

• 2. All section references are to the Internal Revenue Code of 1954 as amended and in effect during the taxable years in issue, and all Rule references are to the Tax Court Rules of Practice and Procedure.↩

• 3. Insco made provision for the estimated unpaid amounts of losses, and loss expenses, arising from incidents reported to Insco during the year, together with an allowance for losses incurred, but not yet reported.↩

• 4. This figure includes a reduction of $ 5,917 in the amount actually determined in the statutory notice of deficiency resulting from the concession of respondent. Premiums in the amount of $ 5,917 were ceded by Insco to an unrelated third-party reinsurer.↩

• 5. Other courts that have confronted the captive insurance company issue have also disallowed deductions for purported premium payments to a captive insurance subsidiary, albeit on theories not necessarily adopted by us. Beech Aircraft Corp. v. United States, 797 F.2d 920 (10th Cir. 1986); Stearns-Roger Corp. v. United States, 774 F.2d 414">774 F.2d 414 (10th Cir. 1985); Mobil Oil Corp. v. United States, 8 Cl. Ct. 555">8 Cl. Ct. 555 (1985); contra Crawford Fitting Co. v. United States, 606 F. Supp. 136">606 F. Supp. 136↩ (N.D. Ohio 1985).

• 6. See Appendix at pp. 1030-1032 for discussion of our prior decisions.↩

• 7. The "economic family" concept is based on the theory that when a captive receives a dollar, its net worth and its parent's net worth increases by that amount, and that when the captive pays out a dollar the converse occurs. Recent articles illustrate the conflict between this and petitioner's positions: Barker, "Federal Income Taxation and Captive Insurance," 6 Virginia Tax Review 267 (1986)↩; Bradley & Winslow, "Self-Insurance Plans and Captive Insurance Companies -- A Perspectived on Recent Tax Developments," 4 American Journal of Tax Policy 217 (1986).

• 8. In Mobil Oil Corp. v. United States, 8 Cl. Ct. 555">8 Cl. Ct. 555, 563 (1985), the Claims Court found that one of the captive insurance subsidiaries insured risks of unrelated third parties. In Beech Aircraft Corp. v. United States, an unreported case (D. Kan. 1984), 54 AFTR 2d 84-6173, 84-2 USTC par. 9803, affd. 797 F.2d 920">797 F.2d 920 (10th Cir. 1986), the District Court indicated that the captive insurance subsidiary insured risks of unrelated third parties in the years subsequent to the year before the court. In Stearns-Roger Corp. v. United States, 577 F. Supp. 833">577 F. Supp. 833, 834 (D. Colo. 1984), affd. 774 F.2d 414">774 F.2d 414↩ (10th Cir. 1985), the District Court found that the captive insurance subsidiary, in addition to insuring the risks of its parent and brother sister subsidiaries, insured certain risks of the parent's customers. In holding that the premium payments from the parent to the captive insurance subsidiary were not deductible, the courts did not discuss the effect of insuring unrelated third-party risks.

• 9. As a theoretical matter, risk distribution or pooling requires:

  (i) Mass -- * * *,

  (ii) Homogeneity -- * * *, and

  (iii) Independence -- * * *

  If these requirements are met to some minimum extent, the principles of average and large numbers operate so that the risk carried by an insurer is far less than the sum of the risks of the insured. * * * To the extent that these requirements are not satisfied, the insurer can offer security to the insured only through sheer financial power. * * * If there is neither adequate distribution of risk nor the financial power to withstand the simultaneous occurrence of all or a significant portion of the insured risks, there can be no transfer of risk, and hence no insurance. [K. O'Brien & K. Tung, "Captive Offshore Insurance Corporations," 31 N.Y.U. Thirty-First Ann. Inst. on Fed. Tax 665, 678-684 (H. Sellin ed. 1973). Fn. refs. omitted.]↩

• 10. See, e.g., Bradley & Winslow, "Self-Insurance Plans and Captive Insurance Companies -- A Perspective on Recent Tax Developments," supra at 233-234 n. 59, and at 255 n. 121↩; K. Tucker & D. Van Mieghen, Federal Taxation of Insurance Companies, par. 20.10, at 2012, P-H Services (1983).

• 11. This concept has been recognized by the Internal Revenue Service. See Ltr. Rul. 8111087, which states in part:

  "Under insurance theory risks are distributed and shifted not through the capital structure of the company, but rather through the premiums (and resulting surplus and investment income) created by the policyholders. * * *"↩

• 12. R. Goshay, Corporate Self-Insurance and Risk Retention Plans 23 (1964); R. Keeton, Insurance Law 7 (1971).↩

• 13. See, e.g., Bradley & Winslow, "Self-Insurance Plans and Captive Insurance Companies -- A Perspective on Recent Tax Development," supra at 237 n. 69, and at 242-243 n. 87↩.

• 14. Without expert testimony, we decline to determine what proportion of unrelated premiums would be sufficient for the affiliated group's premiums to be considered payments for insurance. However, if at least 50 percent are unrelated, we cannot believe that sufficient risk transfer would not be present. This anticipated sharing of premiums paid by unrelated insureds is similar in concept to a mutual insurance arrangement. See, e.g., Rev. Rul. 80-120, 1 C.B. 41">1980-1 C.B. 41; Rev. Rul. 78-338, 2 C.B. 107">1978-2 C.B. 107↩.

• 15. In order to determine whether a substantial proportion of unrelated premiums are paid to a captive, we must assume that reliable actuarial estimations of risk of loss have been used to arrive at the amount of each insured's premium payment. The parties in this case did not directly address the issue of whether premiums paid to Insco were negotiated on this basis, and at arm's length. It appears that during 1974 and 1975, the premiums paid were greatly in excess of the actual losses paid. While this could be due to an extraordinarily low loss year, it could also be a result of excessive premiums. Without concluding which may have been the case here, we note that excessively large premium payments can be indicative of what is in substance a risk-retention arrangement, which would support our conclusion that risk transfer was not present in those years.↩

• 16. Petitioner argues on brief that the years 1974 and 1975 constituted a "startup phase" for Insco, and that payments made in these years should be treated similarly to payments made subsequent to the substantial increase in third-party business for the years 1978-83. We recognize that the startup phase of a business enterprise may not reflect the ultimate intentions and future prospects of the venture. For instance, a determination as to whether an activity constitutes a "trade or business" may require a review of activities over time, as might a determination as to whether an activity is engaged in for profit. See Allen v. Commissioner, 72 T.C. 28">72 T.C. 28, 33-34 (1979); Boyer v. Commissioner, 69 T.C. 521">69 T.C. 521, 537 (1977) (has been appealed); Benz v. Commissioner, 63 T.C. 375">63 T.C. 375, 383 (1974). Similarly, the deduction or amortization of startup costs requires a review of expenditures incurred prior to the active conduct of a trade or business. See sec. 195; Richmond Television Corp. v. United States, 345 F.2d 901">345 F.2d 901 (4th Cir. 1965), vacated and remanded on other issues 382 U.S. 68">382 U.S. 68↩ (1965). In this instance, however, we are concerned with determining the character of payments made to Insco based upon the degree of risk shifting and risk distribution present in the years before the Court. While it may not be necessary to make this determination on an annual basis, we are confident that the 2 percent of unrelated business is not sufficient to satisfy the minimum level of risk shifting and risk distribution necessary to qualify for the deduction.

• 17. In our findings of fact, we have set forth the extent of third-party risks for 1976 through 1983, although those years are not before the Court in this case. At trial, respondent objected to the admissibility of such evidence on the ground of relevancy. We overruled his objection, but admitted the evidence for the limited purpose of demonstrating the consistency or lack of consistency of petitioner in carrying out its avowed purpose described as phase two of its captive insurance program, i.e., to solicit insurance business from third-party sources.↩

• 1. We have not adopted and do not adopt this theory.↩

• 5. While the Supreme Court in Le Gierse expressly stated that insurance must exhibit both the shifting and↩ the distributing of risk, the resolution of that case depended only on the absence of risk shifting. Here, too, we need not consider whether Clougherty's captive insurer arrangement exhibited risk distribution because we conclude that Clougherty did not shift its risk.

• 1. The quotation is not original with the Supreme Court, but is, instead, from G. Couch, Cyclopedia of Insurance Law, sec. 1:3 (2d ed. 1959).↩

• 1. For example, Dr. Plotkin's report provides:

  "The essential element of an insurance transaction from the standpoint of the insured firm, is that no matter what insured perils occur, the financial consequences are known in advance. Thus, the insured, for the price of the premium, is protected, within the limits of its policy, from such financial consequences and from having to worry about and provide for them. By reason of its contract, the insured is indemnified against loss from a defined hazard or risk. In essence, the premium represents the substitution of a small, but certain "loss," for a potentially large and uncertain loss. It provides piece [sic] of mind and the ability to devote all of one's financial resources to other concerns and objectives."

  These basic principles are identical to those set forth by the majority.↩

• 2. Dr. Plotkin specifically concludes that by accepting third-party risks, the captive and its owners accept the uncertainty inherent in the third parties' risks in addition to those inherent in their own risks.↩

• 3. Dr. Plotkin concluded:

  "So long as the firm does not transfer to another the ultimate responsibility for the financial consequences of its risks, it remains the risk bearer and faces the uncertainty of each year's actual financial losses. The attempted placing of a firm's risks, directly or indirectly, in its "insurance affiliates" does not accomplish a transference of risk, or constitute an insurance transaction as a matter of insurance theory or practice or as a matter of economic reality. We find our conclusion in complete accord with the clear theoretical and applied teachings of the economics, insurance theory, risk management, and captive self-insurance literatures and the documented practices of corporate risk managers."

  Mr. Stewart concluded:

  "what it all comes down to is whether there was transfer of financial risk. There was not. The reason is that Gulf owns Insco."↩

• 4. Alternatively, Gulf argued that the agreements should be treated as insurance regardless of the presence of risk transfer,

  "When one considers the large number of risks to which Gulf and its affiliates were subject, the pooling of these diverse independent risks in a single corporate entity such as Insco achieves risk distribution which, standing alone, should result in a determination that "insurance" existed. * * *"

  Gulf did not pursue this argument because we held in Carnation Co↩. that there cannot be insurance without transfer of risk.