Financial reporting is, at its core, an exercise in summarization. For example, the income statement condenses millions of transactions into only a handful of line items. This act is not merely convenient but necessary, since preparing and disclosing every underlying transaction would impose substantial costs on firms. However, summarization also comes at a cost to investors. When a single line item combines activities with distinct economic properties, aggregation can obscure information that investors would otherwise find useful.
This tradeoff raises a fundamental question about how much detail companies should be required to disclose. In a new paper, we examine this issue in the context of selling, general, and administrative expense, or SG&A.
SG&A is a natural place to examine this question. Nearly every public company reports it, and for many firms it is one of the largest expenses after cost of goods sold. Yet the label explicitly combines two fundamentally different kinds of spending: selling and marketing (S&M) costs with general and administrative (G&A) costs. S&M includes advertising, promotion, sales force, and customer-acquisition costs. These expenditures often resemble investment, as firms incur them today to build demand, brand, and revenue in future periods. G&A spending, by contrast, often includes executive salaries, professional fees, compliance costs, and other overhead expenses. These expenditures are more naturally viewed as the cost of maintaining the organization rather than an investment in future growth. When S&M and G&A are combined into a single SG&A figure, these differences become difficult to observe. Accordingly, if separating aggregated expenses provides useful information, SG&A should be an especially informative case.
An Important Measurement Problem
Part of the reason this question has gone unanswered is data. Commercial databases typically report SG&A as a single figure, often because the underlying filings do not provide more detail and because aggregation makes firms easier to compare. Our approach instead uses XBRL-tagged filings submitted to the Securities and Exchange Commission, which preserve the detail a company chose to report. This allows us to identify firms that separate SG&A on the face of the income statement and to measure the magnitude of each component.
Who Disaggregates?
Across roughly 34,700 firm-years of U.S. data from 2011 through 2024, we find that about a quarter of companies voluntarily separate their selling costs from their administrative costs on the income statement. Among those that do, the two components are similar in size on average but vary widely across firms, industries, and life cycle. Marketing-intensive industries, such as business services, computers, and food and drink, break out selling costs far more often than capital-intensive industries such as oil, utilities, mining, and steel. Younger firms also report relatively more selling and marketing expenses, consistent with building a customer base, while older firms report relatively more general and administrative costs. Even before any formal statistical analysis, these descriptive patterns suggest that S&M and G&A expenses capture economically different activities.
What We Found
We then tested whether the two components behave differently in ways that matter to investors. They do, in three areas.
Stock prices. S&M spending exhibits a significantly stronger positive association with stock price than does G&A spending. This suggests that investors treat S&M as more closely tied to future value creation, while viewing G&A spending less favorably. Aggregation washes out this distinction: Combined SG&A has no significant relation to stock price.
Future revenue and profit. If S&M is aimed at building future demand, it should predict future sales better than administrative spending does. Our evidence is consistent with this prediction. G&A is also associated with lower future earnings and cash flow, whereas S&M shows no such drag, consistent with spending whose future payoff roughly offsets its cost.
Cost behavior in good times and bad. S&M and G&A also respond differently when business conditions change. S&M budgets are often noted to be the first expenditures trimmed in a downturn. Administrative commitments, such as personnel, systems, and office infrastructure, are harder to reduce in the short run. The data bear this out: S&M falls more sharply when revenue declines, while G&A is considerably stickier.
Taken together, the three results support the conclusion that the combined SG&A line obscures economically distinct costs that aggregation collapses into a single number.
Additional Analyses
Analyst forecasts provide another way to assess whether disaggregation is useful. Professional analysts may already infer differences between S&M and G&A from other disclosures, or they may simply not incorporate the breakdown even when it is provided. Among matched firms, however, we find that analysts produce more accurate earnings forecasts when S&M and G&A are reported separately, consistent with the breakdown helping users distinguish investment-like spending from overhead.
One important caveat is that companies that disclose the breakdown are not a random sample. They tend to be smaller and more research-intensive, among other differences. Because we observe the split only when a company voluntarily provides it, our results speak to the usefulness of the disaggregation companies actually report, not necessarily to what would happen if every firm were required to disaggregate. To address the concern that our findings simply reflect differences between disclosing and non-disclosing firms, we reweight the disclosing firms so their observable characteristics mirror those of non-disclosing firms. Our main inferences are unchanged.
Why It Matters
This issue is no longer abstract. In 2024, the International Accounting Standards Board issued IFRS 18, and the Financial Accounting Standards Board issued ASU 2024-03, both addressing aggregation and disaggregation in financial reporting. As standard setters weigh how much detail to require and whether that detail belongs on the face of the income statement or in the footnotes, our study offers timely evidence from SG&A. Separating SG&A into its two functional components reveals expenses with different relationships to stock prices, future performance, and cost behavior. These findings highlight how aggregation choices matter because some line items combine activities that investors would evaluate differently if reported separately.
Bradley E. Hendricks is an associate professor at the University of North Carolina at Chapel Hill’s Kenan-Flagler Business School, Jed Neilson is an associate professor at Penn State’s Smeal College of Business, and F. Dimas Peña-Romera is an assistant professor at Arizona State University. This post is based on their paper, “When Aggregation Obscures: Evidence from SG&A Disaggregation,” available here.