Russell 3000 Changes in Depreciation Estimates: Q2 2017

Whenever filing season comes around, we get an influx of interesting disclosures. One thing we keep an eye on is changes in accounting estimates, something we’ve discussed more than once in the past.

In this post we look at the seven Russell 3000 companies that disclosed a change in accounting estimate related to the estimated useful lives of certain depreciable assets during the recently concluded Q2 filing season.

What does it mean? Many depreciation methods involve an estimate of the “useful life” of the asset being depreciated. Take, for example, a piece of machinery in a factory. The asset cost $500K, and the company estimates that the machine will work for 10 years, with no salvage value. Assuming the machine was put into service on the first day of the fiscal year, then using the straight-line method, the company would recognize $50K in depreciation expense on this asset each year for the next 10 years.

Now, suppose at the beginning of the second year, the company determines that, due to advances in machine maintenance, the piece of equipment will now last for another 25 years. The asset’s book value after one year of depreciation at the old rate is $450K. With that change in the asset’s estimated useful life, the company would now recognize $18K in depreciation expense on this asset each year for the next 25 years. Although the total depreciation expense over the course of the asset’s life is the same ($500K), the impact to the period in which the change was made was a relative decrease in depreciation expense of $32K. That is to say, with very little substantive change to the company’s business, operating income is set to receive a nice bump in the year of the change compared to the year prior.

As we can see with the examples above, these changes can often have a substantial impact on a company’s results for a period. In this sample, there are more negative impacts than positive. This is somewhat unusual. Last time we looked at this issue, depreciation estimate changes were more likely to have a positive impact on earnings.

For our AQRM model, we assign severity based primarily on where the estimate, especially positive ones, affects the financial statement: the higher the line item, the more severe the flag. Adjustments to revenue recognition (sales return allowance, breakage, etc.), the allowance for doubtful accounts, or to inventory have the potential to inflate gross margin and give the impression of better operating results than the economic reality.

Depreciation estimates on the whole are not as significant, in large part because investors and analysts routinely exclude depreciation from their assessment of a company anyways (in the form of “EBITDA”). Nevertheless, we’ve found that the impact of these changes in estimates is often elided in a company’s earnings announcement, and therefore are often included in the “sticker price” of a company’s adjusted earnings for a quarter.