My parable about GM's earnings-management-driven pension decisions inspired a friend of this blog (let's call him Mark) to share a story from around the same time period.
I worked at J. I. Case, the farm and construction equipment manufacturer. As was the accounting policy for car companies, we recognized sales upon shipment out the plant to a dealer.
Case financed the dealers so it was relatively easy for the salesman (based on instructions from management) to tell every dealer in August (fiscal year end at end of October), "please order x number of tractors. They will arrive in early November, having shipped from the plant in late October. We know you don't need them until next spring, but we will give you interest-free loans until next June. Presumably the tractors will have been sold by then and you will not lose any sales in the meantime."
As a result, production was really ramped up in September and October, and production then really sank in November and December. Worse, the company had to finance its "sales" to dealers for six or more months. Knowing there would be an inflow of cash, even though the company was not doing well, Case told all its vendors: "We have to be out of the banks for six weeks in order to maintain our line of credit. So we will not pay you in October but you will get everything we owe you on December 15th." This satisfied the vendors because certainty of payment was sufficient. However, satisfied is not the same as liking it, and Case's leverage with vendors was reduced—which squeezed our margins even more.
Case had the choice of recognizing revenue at shipment, delivery or upon ultimate sale by the dealer; but once they had chosen the point of shipment to record revenue, the company (and industry for that matter) could never go back. I worked there at Case in the late '60s and this was only 10-12 years into the system, but even then the finance staff saw how a bad accounting rule really screwed up operations, financing and ultimately profitability, but there was nothing anyone could do.
In a couple of respects, this story is an even better object lesson to illustrate the principle that lousy accounting standards create costly perverse incentives for management. To begin with, pensions are a difficult financial problem, but machinery sales to dealers is relatively simple and a critical component of earnings for every company. Thus, it follows that schemes (legal and otherwise) to juice revenue abound; for example, my recollection is that the SEC general reports that about 60% of its enforcement actions for accounting violations involve revenue recognition.
Moreover, the main reason I decided to put Mark's story out there is that it is timely: revenue recognition standards are in the process of being revised by a joint FASB/IASB project. Would the same opportunities to scramble revenues and earnings have been available if the current proposal to replace extant U.S. GAAP had been in effect? YOU BETCHA!
In the revenue recognition project's nascence, the boards did solemnly aver that new standards would result in a much more faithful portrayal of economic reality: the driver of income would be changes to assets and liabilities, which would be measured at current values. But, once the boards got the message from issuers that they wouldn't easily give up the revenue recognition rules they knew and loved, the project has devolved into a (futile?) face-saving quest to put old wine in new bottles – i.e., to write one reasonably brief standard that would return the same (or better) values as hundreds of rules that comprise existing U.S. GAAP.
Ironically and tragically, the solution to the J.I. Case problem (and the general problem when cash is received after services are performed) is actually very simple. First, an entity should not be permitted to recognize a receivable from its customers (dealers) unless it owns a present legal right (conditional or unconditional) as of the balance sheet date to cash or another asset. Second, the asset (receivable) representing that economic right should be measured at its current value. Third, a detailed roll-forward of the balance of receivables from customers would be presented in the notes to the financial statements.
With those three simple guidelines, revenue recognition games for Case, and practically everyone else, would be over. Not coincidentally, financial statements would do a better job of communicating the effect of economic events on shareholder value. It is the only place that the boards could have ended up if they had stuck to their original commitments for producing a high quality revenue recognition standard.