At the heart of the problem was Medicis' policies for product returns. For the six months Dec. 31, 2005, the company utilized an exception to SFAS 48 that supports use of replacement costs for returned products in establishing a reserve. The exception exempts returns by endusers for similar products. However, Medici was found to be using this provision to accept returns on unsaleable products that were returned by resellers, not by endusers. Medicis granted credits equal to the original gross sale price on returns and the credit was often applied to purchase of differently priced new products. A 2006 quality review by the firm resulted in E&Y personnel coming up with a rational to support the Medici practice instead of remedying it.
While a team determined the exception did not apply, it decided that the return policy could be supported by analogy to warranty accounting which Medicis adopted under SFAS 5. According to the PCAOB report, a Consultation Memorandum dated June 1, 2006, pointed out the company itself had not invoked the warranty analogy. "... and it sure seems like this is a new policy, or at least a policy the company never knew it was following."
In 2006, Medici developed a new method of estimating a sales return reserve for new products which resulted in $17 million in unexpected returns for the fourth quarter of that year. The company had a return reserve of $35.2 million at the end of 2006 that was supposed to cover 18 months of future returns, but did not cover the $53.8 million in returns for the 12 months of 2007. The 2007 year-end reserve of $9.6 million was less than the value of just the returns for the fourth quarter of that year.
The PCAOB also found that the E&Y team failed to appropriately audit a units-in-channel method for non-legacy returns in calculating the reserve estimate. That was done in 2006 under Butler's encouragement. Butler, however, failed to test key assumptions including management's assumption about when returns would not occur. The company's management estimated that 12 weeks of project script inventory was the appropriate level of inventory in the channel and that sales below that level would not be returned.
According to the agency, Medicis had not studied proper channel inventory levels and Butler should have known evidence was not provided that the calculation was reasonable and did not consider whether the company needed to disclose the change in return policy in financial statements. If the company had stuck to the historical method of valuation, the return reserve would have been significantly higher. And when Medici used both methods for comparison, the historical method boosted the reserve by 81 percent.
It was also found that Anderson and Christie did not try to test whether management's estimates of channel inventory was reasonable, nor did they evaluate adequately the change to units-in-channel valuation, leading to the decline in the reserve to an amount that was not adequate to cover actual returns.