Revenue recognition when product return and pricing adjustment uncertainties exist
My forthcoming paper in Accounting Horizons (Rasmussen 2013; “Revenue Recognition, Earnings Management, and Earnings Informativeness in the Semiconductor Industry”) examines the implications of revenue recognition for companies with product return and pricing adjustment uncertainties. Although these uncertainties are typically minimal for sales to end customers, they can pose large risks for sales to distributors. The reason being is that distributors’ product return and pricing adjustment rights often do not lapse until the distributor resells the product to an end customer. In the midst of these risks, companies recognize revenue upon delivery of product to distributors (sell-in), when the distributor resells the product to end customers (sell-through), or under some combination (sell-in for some distributors and sell-through for others).
I examine two implications of revenue recognition for companies with product return and pricing adjustment uncertainties. First, I examine whether the incidence of earnings management is higher for companies that recognize revenue before their product return and pricing adjustment uncertainties are resolved. This expectation is motivated by the fact that more opportunities exist to manage earnings when revenue is immediately recognized under the sell-in method compared to when at least some revenue recognition is deferred under the sell-through and combination methods. Specifically, managers using the sell-in method (1) maintain (and have opportunities to manipulate) product return and pricing adjustment accruals, and (2) can boost earnings through channel stuffing activities.
Second, I examine whether earnings informativeness (proxied for with the earnings response coefficient) differs among the revenue recognition methods used by companies with product return and pricing adjustment uncertainties. On one hand, immediate revenue recognition more quickly incorporates new accounting information into the financial statements. If this new information is useful to the market, earnings should be more informative under the sell-in method compared to the other revenue recognition methods. On the other hand, more opportunities exist for both intentional performance manipulations and unintentional estimation errors when revenue is immediately recognized. Thus, if earnings are (or are perceived to be) more inaccurate under the sell-in method, earnings informativeness should be higher when revenue recognition is deferred until distributors have resold products to end customers.
In order to study these research questions, I limit my sample to semiconductor companies because they sell to distributors and naturally face product return and pricing adjustment uncertainties due to rapid product obsolescence and declining prices over product life cycles. I find that sell-in companies are more likely to meet or beat analysts’ consensus earnings forecast compared to sell-through and combination companies, suggesting that earnings management is more likely when companies immediately recognize revenue for sales to distributors. I also find that the earnings response coefficient is significantly larger (meaning the returns-earnings relationship is stronger) for sell-through companies compared to sell-in and combination companies. This finding suggests that earnings are more informative when revenue recognition is deferred until the distributor has resold the product to end customers. Collectively, these results suggest that revenue recognition should be deferred until all product return and pricing adjustment uncertainties are resolved.
This study should be of interest to the FASB and IASB as they finalize a joint revenue recognition standard. The current exposure draft of the new standard states that revenue recognition should occur when the customer obtains control of the product or service. Control, as described in the exposure draft, is likely to be transferred when a manufacturer delivers product to a distributor except for cases where a consignment agreement exists. At the time control is transferred, the standard directs the manufacturer to estimate variable consideration (e.g., product returns and pricing adjustments), determine the transaction price, and recognize revenue so long as receipt of the estimated transaction price is reasonably assured. Recent technical briefs from the Big 4 accounting firms suggest that the new standard’s provisions regarding variable consideration may require many manufacturers that have historically used the sell-through method to change to the sell-in method. Such a shift is concerning as my findings suggest that earnings management is more likely and earnings informativeness is lower when revenue is recognized at sell-in.