This article discusses issues involved with accounting for product warranties.
The fundamental idea of accrual accounting is that revenues are recognized when the earnings process is complete and not necessarily when the goods or services are paid for. Click here for more on bases of accounting. In many cases determining when the earnings process is complete is very easy. A bar that lets patrons run a tab has a completed earnings process every time a customer is poured a drink. However in many businesses the completion of the earnings process is not so clear cut. Take the case of product sellers who offer warranties to purchasers. Many if not most consumer goods are sold subject to some kind of product warranty. This includes everything from microwaves and computers to automobiles and trucks. Sales of products subject to warranties present yet another challenge to accrual accounting.
Example. Bust Buy sells silicon gel implants used in cosmetic surgery. Each pair of implants sells for $1,000. However they offer a ten year warranty. If their implants fail they will replace them with new ones. When is the earnings process complete? It would seem that the “firm“is on the “hook” to “support” their product for ten years. So from one standpoint they should not recognize a completed sale until ten years have elapsed.
But waiting until a product warranty expires seems like an extreme application of the accrual accounting concept. After all if manufacturers and producers thought that actual warranty costs were very significant they either would not sell the product at all, improve the product or go out of business.
A more realistic approach and the one sanctioned by GAAP is to allow warranty issuers to make reasonable estimates of the costs of meeting product warranties and recognize these warranty costs when sales are made.
Example. Say Bust Buy “implant” engineers believe that only one out of every one hundred pairs of implants will fail within the ten year warranty period. This is an implant failure rate of 1 percent. Using this percentage the firm would recognize $10 of warranty expense for every $1,000 sale of implants.
The bookkeeping for this is straightforward. When a sale is made an expense entry is also recorded. Say the company sells 10 pair of implants for $10,000. On the same day the sale was made the following entry would be made:
The first account is simply an expense while the second account is a liability account. What happens if an implant actually fails within the warranty period? Let’s say a pair of implants fails five years after sale. Let’s also suppose that the replacement costs are only $300. Let’s say all the replacement costs involve cash payments. This would be the entry:
Notice that no actual expense is recognized in 2010. This is because the expense was recognized in 2007 when the sale of the implants was consummated. From the point of view of the matching concept this makes sense.
Compared to variables such as useful lives of assets for depreciation purposes or collectability of accounts receivable, warranty costs can be wildly unpredictable. After all a major failure in the production process may require a complete recall of a product line which can be enormously expensive.
Estimating warranty expense cannot possibly be an exact science because with continuing product innovation past warranty cost experience with established products will not necessarily be a good guide for estimating warranty expense associated with a new untested product.
For recent developments in accounting for product and service warranties click here.