First, the two special charges (listed on the spreadsheet and here by their footnote denotations) that supposedly occurred in the final quarter of 1996, if reversed, begin to bring the posted earnings of the two years closer in line. As posted, the cost of goods sold is 71% of sales for 1997 (70.6% for the fourth quarter), but a substantially larger 91.5% for 1996 (and a whopping 115% for the fourth quarter of 1996). Simply removing Special Charge A from the cost of goods sold brings 1996's cost of goods sold to 82.1% (80.7% for the fourth quarter). Selling expenses, which in 1996 were affected by Special Charge B, show a similar pattern. These expenses formed 11.2% of sales in 1997 (9.7% for the fourth quarter) but 21.7% in 1996 as reported (35% for the fourth quarter); removing the amount from Special Charge B. still leaves a significant gap between 1996 and 1997 but closes it substantially, with this ratio dropping 17.4% after the adjustment (with a more substantial drop, down to 19.2%, in the fourth quarter of 1996). The earnings (or loss) before taxes can also be adjusted simply by eliminating the special charges to show more similar performance between 1996 and 1997 than initially appears to be the case, though there is still a loss in 1996.

There is an additional restructuring charge of $154.9 million listed in 1996 as well, and removing this cost all but eliminates the loss that occurred in 1996, the majority of which still occurred in the fourth quarter when all of the special charges entered the picture. All of the above effects would be magnified still further if the reported charges in 1996 were removed from earnings in 1997, or were split between the two years. That is, it appears as though false losses were created in 1996 to make 1997 look better, and it would make sense that these false losses were then reported as part of earnings in 1997. If this is the case, performance in both years would be even closer in value and would be generally lower -- especially in 1997 -- than reported.

Taking the new earnings figures and adjusting for other discrepancies in the cash flow statement also leads to some significant differences in the perceived financial standing of the company in both years. While it is impossible to determine how much of the extreme difference in receivables and inventories for the two years comes down to the withholding in 1996 and the inflation of 1997's numbers, simply averaging out the values reported in both years and accounting for these differences (and the different earnings) in the net cash provided (or used ( by operating activities changes things substantially, making 1996 a much better year than reported and 1997 significantly worse. This case does not really suggest why Sunbeam's accountants or officers would have an incentive to engage in this behavior, yet their accounting discrepancies are glaring enough to be suspect even without a clear motive.