Shareholders often ultimately judge company managements by how well they generate returns on capital. For example, two companies may generate net income of $2 million, but if one required $10 million of shareholder capital to do it while the other required $100 million, the former company's management team will be lauded while the latter may have its job on the line. But in investigating a company's return history, shareholders must keep in mind the effects accounting standards can have on the results. In some industries, accounting requirements can seriously skew calculations such as ROE and ROIC, making several industries look more attractive than they otherwise are.
For example, consider a company in an industry where capital expenditure requirements are low, but research and development (R&D) requirements are high. (These properties are common in software development companies or pharmaceutical companies, for example.) Because of how GAAP works, R&D spend is expensed as incurred (i.e. expensed immediately), in contrast with investments in say equipment, which are capitalized and expensed over several years.
There is nothing wrong with these accounting policies. Conservatism dictates that R&D cannot be capitalized (i.e. placed on the balance sheet) because the benefits from R&D are difficult to estimate. But spending on R&D is still an investment, with risky returns that could be several years out, just like a capital investment.
But consider how this accounting practice affects the companies that fall under the above category. Because these investments are not capitalized, the book values of these companies are lower than they are for companies in industries where investments are capitalized. As a result, returns look better in some industries than they do in others, but only because of the accounting.
Consider the example of Pfizer (PFE), a company with little in the way of capital expenditure requirements, but that must spend big money every year on R&D to stay competitive. Its ROE over the last several years (depicted below) suggests a a company with decent returns:
But if you treat Pfizer's R&D spend as if they are capital expenses, the returns look very different. The chart below compares Pfizer's ROE to an adjusted ROE level in which R&D expenses over the last three years are capitalized:
Suddenly, the returns look a lot more average! If we then capitalize R&D over a period of more than three years, which is not unreasonable considering the amount of time it takes a newly developed drug to get to market, returns are even lower.
When an investor is considering firms in the same industry, these industry-specific attributes will of course weigh on all firms. But for bottom-up value investors who look for value across industries, recognizing the role that accounting plays in determining return-on-capital metrics is of the utmost importance; it will result in more accurate comparisons of profitability in companies separated by accounting standards that treat various investments differently.