Securitas Holdings, Inc., is a large corporation that provides a variety of services through its operating subsidiaries, including armored truck and securities services. In 1999, Securities acquired Pinkerton's and the following year, Burns', both companies that provide security services.
To make a long story short, Securitas' second-largest expense was insurance. In 2002, to better manage its insurance costs, Securitas formed a captive insurance company domiciled in Vermont. in 2003, Securitas paid over $77 million in premiums to its captive, and nearly $57 million in 2004.
Securitas took deductions for these amounts paid to its captive, and the IRS wasn't pleased. While conceding that the premium amounts were reasonable, the IRS disallowed over $47 million of the 2003 deductions, and over $41 million of the 2004 deductions. The balances for both years that the IRS did not challenge constituted either actual claims that Securitas had paid in those years, or were the fees paid by the captive for reinsurance.
If the premium amounts were reasonable, why did the IRS deny the deductions? Here, we get to the meat of the dispute. To give its captive additional financial strength, and to allow some of the money in the captive to be loaned out for use of one of the operating subsidiaries, Securitas guaranteed the payment of the captive's claims. This is known as a "Parental Guarantee", i.e., Securitas as the parent company guaranteed the liabilities of its subsidiary captive insurance company.
The IRS, citing some cases where the presence of a Parental Guarantee put the hex on the captive arrangements and resulted in an IRS victories, argued that the same result should apply here: The presence of Securitas' Parental Guarantee destroyed the captive arrangement because there was a circularity going on -- Securitas' captive was insuring Securitas (well, its operating subsidiaries), and Securitas was basically insuring the captive in case it failed. Therefore, according to the IRS, the Securitas captive wasn't a real insurance company, because at the end of the day it didn't stand behind its losses, but rather its de facto insured (Securitas) did.
The Tax Court rejected the IRS's arguments, distinguishing those previous cases involving parental guarantee from Securitas' captive for basically two reasons: (1) The Parental Guarantee was never actually called, and thus the IRS's circularity argument was at best theoretical; and (2) Securitas' captive was adequately capitalized to stand on its own and pay claims.
Notably, this is the second Tax Court case this year involving a parental guarantee. In the Rent-A-Center case, the IRS similarly challenged Rent-A-Center's Parental Guarantee of its captive, and similarly lost.
ANALYSIS
This Opinion is probably to be limited in its importance to the issue of the Parental Guarantee. Since Parental Guarantees are nothing like a "best practice" in the captive industry, and are actually pretty rare, this decision probably will have nothing like a broad impact.
Notably, the decision does not validate all Parental Guarantees, but like that in Rent-A-Center, only validates that Parental Guarantees are not ipso facto bad, and instead are subject to a facts-and-circumstances analysis that looks to whether the captive was adequately capitalized such that the Parental Guarantee really wasn't needed and in fact was never actually called. In other words, and somewhat ironically, if the captive doesn't need a Parental Guarantee, then the Parental Guarantee is just fine with the Tax Court.
Otherwise, it is worth noting that the Tax Court spent a goodly amount of time discussing Securitas' need for the captive, and noted that the captive paid very significant amounts of claims in 2003 and 2004. These are important facts that helped the Tax Court distinguish Securitas' captive as a bona fide risk management vehicle, as opposed to a tax shelter disguised as a captive.
The Tax Court also put great weight upon the decisions of the Vermont captive regulators, which Securitas followed. It is one thing for a captive to do something edgy unilaterally, and quite something else to do so after due deliberation and approval by the Vermont captive regulator (I am skeptical that the Tax Court would extend the same deference to smaller captives formed in obscure jurisdictions).
While everybody in the captive industry likes to see "a victory for captives", I'm not sure that this Opinion will be of much comfort other than to those involved with personal guarantees. As in the Rent-A-Center opinion, the Tax Court has indicated its acceptance (as has the IRS) of properly formed and operated captives. However, it is important to note that what might work for Securitas and Rent-A-Center, being very large companies with substantial risk and substantial claims, might not work at all for smaller captives with much lesser risk and fewer if any claims.
Which is to say that the sorting between "real captives" that are primarily meant to fulfill an insurance and risk management function, and "tax shelter captives" that are primarily meant to generate a tax deduction, will continue and the Securitas opinion will not affect that.
CITE AS
Securitas Holdings, Inc. v. CIR, U.S. Tax Court Dkt# 21206-10, Opinion dated October 29, 2014. Full Opinion at http://www.captiveinsurancecompanies.com/cases/securitas/securitas_taxctop_29oct14.pdf
This article at http://onforb.es/1o39OZ5 and http://goo.gl/PA9RTP
