The IRS' War on Captives; Humana, Part I

Welcome back.  I hope that everyone had a good holiday season.  

Before delving into the Humana case, note on the right side of the blog is a new box of links, titled "The Captive Cases."  I've added links to each of the sections I've written over the last few months on the captive cases and presented them in chronological order.  If you ever want to go back and review the legal history of captives, please do so.  

Just to provide some context before moving forward, we've been looking at the captive cases in chronological order.  Before the holidays we finished with the economic family doctrine.  In the conclusion of my analysis to those cases I noted that, most importantly, the IRS was very prepared for captive litigation while the taxpayers weren't.  As a result, the IRS gained a strategic foothold in captive jurisprudence.  With Humana, we see the first really good taxpayer push back against the IRS' efforts.  To that end,we'll be spending some time delving into the decision's details.  The following is an excerpt from my book, U.S. Captive Insurance Law.

Humana was a groundbreaking case because it was the first major victory for a taxpayer in the captive insurance area.   Humana was (and is) a publicly traded health care company.     By the mid-1970s, it was incredibly difficult for the company to find adequate insurance.   The company considered going uninsured but did not have enough funds to withstand a catastrophic risk.   They also considered setting up a reserve, but payments to a reserve fund are not deductible, and the trust fund would not allow Humana to access the third-party insurance market.   A third option was combining with other hospitals in a 5-year pooling arrangement, but Humana did not want to commit to a 5-year program and was unsure about the financial viability of other possible participants.   Finally, the company could set up a captive insurance company – an option which was accepted because 

it possessed none of the perceived disadvantages associated with the other options and it would provide a regulated method of insuring risks which would both isolate funds for the settlement of claims and satisfy interested lenders, mortgagees, and securities analysts.  In addition, Option (4) [establishing a captive] would provide access to world reinsurance and excess insurance markets. 

On August 5, 1976, Humana incorporated Health Care Indemnity under the Colorado Captive Insurance Act.   The Insurance Department of Colorado approved Humana’s establishment of a captive under Colorado law.   

Health Care Indemnity (the captive’s name) issued preferred and common shares.   Humana purchased all 250,000 shares of common stock by paying “$750,000 in the form of an irrevocable letter of credit issued in favor of the commissioner of insurance of the State of Colorado.”   Each common share of stock had 5 votes.   A Humana subsidiary in the Netherlands Antilles purchased all 150,000 shares of preferred stock for $250,000.   There were no further agreements among Humana, its Netherlands Antilles subsidiary and Health Care Indemnity for the injection of any additional funds into the captive insurance company.

Health Care Indemnity issued three policies which covered the vast majority of Humana’s hospitals.   All of these policies conformed with industry standard practices.   For years 1977 to 1979 Humana (the parent) paid total premiums of $21,055,575 to Health Care Indemnity.   These payments represented amounts for the parent and its subsidiaries.   The sole issue at trial was whether these amounts were deductible as insurance premiums.   However, there were two sets of premiums.  The first was from the parent company to the captive.  The second was from the subsidiaries to the captive.  It is important to remember this distinction going forward.

Humana lost the trial case but filed a petition for reconsideration.   The tax court withdrew its memorandum opinion and issued a full opinion after review by the 19-person court.   The written opinion contains a 12-member majority opinion, an 8-person concurrence, a 2- member concurring and a 7-member dissent.   The sole reason for Humana’s petition was to get a long opinion which the company could use for the basis of an appeal. 

End excerpt.

Before moving forward, let's make some observations.

Just as the IRS often waits for the most egregious fact pattern to prosecute, this fact pattern represents a a near perfect set-up for the taxpayer to defend.  First, business necessity -- not a tax angle -- forced the taxpayer to look at the possible formation of a captive program.  This brings Humana squarely in line with the business purpose requirement of the Frank Lyons case and prevents an IRS attack based on the then most prevalent anti-avoidance theory, the sham transaction doctrine.  In addition, the taxpayer carefully considered four possible alternatives, and rejected three for good business reasons: going uninsured would expose the company to too much risk; contributions to a reserve were not deductible and forming the then equivalent of a risk retention group would expose the company to possibly weaker plan participants.  As such, the best alternative -- again, after careful deliberation -- was to form a  captive. 

The captive formation process was also picture perfect.  First, the captive was formed domestically.  While there is nothing inherently wrong with using an offshore jurisdiction, it can have a negative taint when mentioned in court.  As such, the fact that the captive was formed in a US domicile makes this situation appear more "on the up and up."  Next, the captive was capitalized with $1 million -- a more than adequate amount of initial capital.  Additionally, there is no formal agreement for any plan participant to provide further capital -- a fatal flaw in an earlier case.  

The captive sold three master policies to the parent company that covered the vast majority of the parent companies' risks.  The policies confirmed to industry norms.  The amount of total premiums over a three year period -- $21,055,57 -- appears to be reasonable on its face (also note this issue was not contested at trial). 

Simply put, this is a great fact pattern to defend -- which we'll begin to explore in the next post.

---