Let’s Assume a Can Opener
A look at some recent changes in accounting assumptions that improved the earnings picture for several companies
There are many jokes based on the comical outcome of unrealistic assumptions. We are fond of the classic one poking fun at economists featuring a physicist, a chemist, and an economist stranded on a desert island with no way to open a can of soup. The physicist and chemist both offer potentially workable solutions to the problem based on their respective areas of expertise while the economist's contribution is simply “let’s assume a can opener.”
Analysts can never lose sight of the fact that like economics, accrual accounting is based on assumptions. While accountants have a reputation for being very precise, the fact is, two accountants starting with the same transaction data may very well arrive at materially different figures for the profit a company earned during a given period. This is due to the fact they may be using different assumptions for factors such as how long should it take the company to “earn” the cash it received from its customers in the period or how many warranty claims will be filed in the next year. The degree to which a company’s earnings are sensitive to changes in assumptions will vary depending on the company’s industry and business model as well as the company’s accounting policies. In many cases, a company may have one very large revenue or expense item that is very sensitive to assumptions that involve considerable management discretion. Changing assumptions in this assumption can have a material impact on the level of earnings.
We remember talking with the management of America’s Car-Mart (CMRT) several years ago about its credit loss reserves. The company sells and finances used cars and expected that one-third of its sales would require a modification to collections. A customer defaulting on payments leading to repossession was a primary cause, but so were car accidents or a large repair that was not justified by the car’s value nor did the customer have the ability to pay for it. CMRT held a credit loss reserve of about 20% of receivables for years. Suddenly the reserve started to decline which was driving earnings growth simply because it was booking lower credit losses. In most years, the credit loss figure was 2-3x the amount of net income CMRT was earning. The president was very upfront about it saying he believed the 20% level would return, but because scrap steel prices were up and there had been a two-year period of lower defaults – the auditors did not think the 20% figure was still justified. The reserve percentage fell as low as 18%. It did rise in the future back to 27% in some years. This was a company that was earning about $1.30-$1.40 per share and only a 5% change in credit losses was worth about 10 cents of EPS.
Many companies have large expense items like this. United Rentals (URI) depreciation on equipment runs $1.85 billion and the company notes that changing the useful life of that equipment by only one year would change depreciation by $215-$275 million. That would change United Rentals’ earnings by about 8%
Not only does management often have considerable discretion in making these assumptions, but they can also change their assumptions if it is decided that the old range is no longer appropriate. We note here that in our experience, these changes almost always result in earnings increasing. We don’t remember the last time we heard a company say “in the name of more realistic earnings, we are voluntarily decreasing the time over which we are recognizing our costs.” Still, in most cases, companies are very upfront about a change in accounting policy. They may even quantify what the change in policy did to EPS. However, the problem is that as time moves forward, investors may forget the change and just accept a higher level of earnings overall and in turn a higher stock price even if the P/E ratio is flat.
Analysts need to be able to identify what key costs are subject to assumption, how they compare to the company’s peers, how they are calculated, and how sensitive they are to changes in assumption. We will examine a few recent accountings changes at some high-profile companies below.
Note that since we put up our paywall, our reports digging into company earnings quality are now largely limited to paid subscribers. However, if you would like to get a feel for what our unique brand of analysis involves, you can sign up for a free trial here:
For an overview of earnings quality analysis and how we look at companies, we recommend reading An Introduction to Peek Behind the Numbers.
If you have any questions or are interested in hearing about our institutional research service, please contact firstname.lastname@example.org.