- 24 - d. Adjustment for Partial Year The experts also disagree on how to “annualize” the income from the partial fiscal year from July 1, 1993, through the April 10, 1994, valuation date for purposes of computing average income for the base period. Mr. Sliwoski extended the partial year income data pro rata to a full fiscal year, added this amount to the net income from the previous 5 years and divided the result by 6. Mr. Kramer, on the other hand, simply added the net income from the 9.33 months of the partial fiscal year to the income from the previous 5 years, divided the result by 69.33 months, and multiplied the result by 12 to compute the average. The estate finds fault with Mr. Sliwoski’s approach, and we agree. By simply extending the results of the 9.33 months of the partial fiscal year pro rata into 12 months, Mr. Sliwoski effectively postulates level income over each month of the fiscal year. We agree with the estate that this approach distorts Renier’s income. The first 9.33 months of Renier’s fiscal year include the holiday season, a period of high retail volume. The assumption that the average of the first 9 months of the fiscal year would be replicated in the last 3 is highly unlikely. In addition, both sides have conceded that 1994 income was anomalous, due to the correction of the inventory error. As a result, we believe a more accurate average is achieved by averaging the actual results of the first 9.33 months of fiscal 1994 with the preceding 5 fiscal years, as Mr. Kramer has done.Page: Previous 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Next
Last modified: May 25, 2011