The Bancroft case has been working its way through various federal jurisdictions over the last few years. In theory, it involves a captive insurance scenario. In reality, it's tax evasion, plain and simple.
For me, the dead giveaway is the the "round trip" nature of the transaction. Consider this explanation of the program from the decision:
Sigel and Barros formed Bancroft. They didn’t know much about insurance so they outsourced the underwriting function, actuarial responsibility, claims handling, accounting function, due diligence inquiries, and routine paperwork chores. They also outsourced much of the investment operation. One of Bancroft’s primary investment vehicles was to make commercial loans back to the various participants who gave their money to Bancroft in the first place. Not coincidentally, Bancroft would loan back 70% of the premium dollars that had been committed for “coverages.” Bancroft would outsource responsibility for securing the loans and perfecting the security to the borrower. Perhaps not surprisingly, the perfection of the security was, on occasion, “forgotten.” The insuring end of the business similarly went lacking. Scolari was asked how much he wanted to pay in premium dollars. In 2006 he responded: $2.6 million, and in 2007, $5 million. In return, Scolari received a tax deduction for the full premium, and insurance coverage that he and his company didn’t need. He also received a promise that, if his claims were low and the investments were successful, he would receive a refund of his premium dollars after five years.
Anti-avoidance law is littered with cases involving this type of fact pattern: parent makes payment to a controlled company. The payment provides some tax benefit to the parent (usually an outsized deduction). The controlled company holds the money for a short period of time and then funnels the money either directly back to the parent or to another entity that somehow allows the parent to still control or gain benefit from the funds. This is one of the primary reasons the tax code has section 482, which is the US codes transfer pricing section (the law's section is short; the accompanying Treasury Regulations are extensive).
At the heart of the IRS’ concern regarding captives is the captive is nothing more than an accounting reserve rather than an independent stand-alone insurance company. Defined broadly a reserve is a, “separation of retained earnings to provide for such payouts as dividends, contingencies, improvements, or retirement of preferred stock.” Several early tax cases (tried to the Bureau of Tax Appeals, the predecessor of the Tax Court) specifically disallowed the deduction of payments into a reserve fund, based on the following reasons:
1.) The tax code allowed a deduction for business expenses, but not for amounts paid into an internally held reserve. This is supported by a strict reading of the statute (currently 26. U.S.C. 162(a) and the accompanying Treasury Regulations).
2.) Moving funds internally – from cash to a reserve or from one corporate “pocket” to another – does not shift the risk as required by insurance. This is essentially an analysis based on a strict reading of this occurrence at the balance sheet level.
3.) Preventing the manipulation of gross income through the use of “reserves” and “contingency funds” as outlined in Spring Canyon Coal.
4.) Both accrual and cash accounting methods require the taxpayer to deduct specific “realized” amounts. A taxpayer cannot deduct a speculative amount. As the ultimate amount of the payout from the reserve is speculative, a deduction is not allowed.
Realistically, the service is ultimately concerned with point number 3 above – using the establishment, maintenance and eventual dissolution of a reserve to manipulate earnings. As an example of the fact pattern, suppose Acme Corp. has strong yearly earnings this year. In order to lower their gross income (and hence taxable income) they establish a reserve for some type of reasonably foreseeable contingency. However, five years later when earnings are lower, the company declares the contingency no longer exists and hence dissolves the contingency fund, bringing the reserves back onto the company’s income statement at a time when the tax burden is lower.
Regardless of the feasibility – or lack thereof – of the preceding hypothetical fact pattern, it does provide prospective captive owners with a clear compliance mandate: after establishing the captive, the longer the captive’s funds remain separate and non-distributed to the parent the better. Regrettably, there is no formal guidance providing a minimum amount of time during which the captive’s funds must remain entirely separate from the parent. However, we do have several examples from non-captive case law wherein a company’s corporate existence was too short to create sufficient corporate substance. The corporation established by the taxpayer in the landmark anti-avoidance case Gregory c. Helvering, existed for a mere three days. Several anti-avoidance cases from the 1990s involved partnerships which were in existence for periods up to one fiscal year. Beyond the period of a year, however, guidance becomes far murkier. However, common sense informs us that the longer a captive is in existence with its funds separate from the parent, the better. Ideally, the minimum time which the parent should commit to running the captive without payments from the captive going to the parent is three to five years. This length of time will allow the captive to fully develop initial capital and surplus relative to the original underlying risks being underwritten by the captive.
 Barron’s Finance and Investment Handbook, page 446 © 1990 Barron’s Educational Series, Inc.
 See also Appeal of William J. Ostheimer 1.B.T.A. 18, 21 (“The statute specifies what deductions are allowable and, except in the case of in insurance companies, no provision is made in the 1918 Act for the deduction of a reserve as such.”).
 See also Appeal of Consolidated Asphalt, 1 B.T.A. 79, 81 (“When estimating the reserve to set aside for a construction contract, the appellant’s accountant doubled the amount set aside for the years in question.”).
 See General Counsel Memorandum 35340, 05/15/1973 (“However, because anticipated casualty losses are contingent in nature, it is a firmly established principle of tax accounting that even as accrual basis taxpayer may not deduct amounts it adds to a reserve for insuring its own risks.”).
 Gregory v. Helvering, 293 U.S. 465 (1935).
 ACM Pshp. v. Commissioner, T.C. Memo 1997-115 (T.C. 1997), ASA Investerings Partnership v. Commissioner, 201 F.3d 505 (DC Cir. 2000),
Link From Our Previous Blog