Estimating the Private Value of Financial Statement Statistics
ABSTRACT We develop a method for estimating the private value of knowing the future realization of some financial statistics, and then we apply the measure to the familiar ratios from the DuPont decomposition of return on equity. The estimation is grounded in the standard rational expectations model, adapted to accommodate relative risk aversion, and determines an investor's willingness to pay for information denominated in units of a riskless asset (e.g., money). The method can accommodate different levels of investable wealth, multiple assets, and any information system regarding those assets. We use this method to document several interesting contrasts. An investor with $1 million in wealth who already knows all the current DuPont ratios, as well as the most recent analyst forecast, would be willing to pay $2861 to know all of next year's ratios. This amount increases to $3158 for firms below the median size. We find that the value of financial information is higher in periods of high sentiment and for firms that are hard to value or hard to arbitrage (i.e., firms with high sentiment exposure). In practical terms, this suggests investors prioritize their information search to small, young, loss‐making, high‐growth, low book‐to‐market, non‐dividend‐paying, and illiquid firms. In terms of the DuPont components, we find that knowing next year's operating profit margin alone is worth $1226, whereas knowing next year's leverage is worth only $170 and knowing next year's sales growth is worth only $184. In practical terms, this suggests that investors allocate more attention to forecasting profitability and less attention to forecasting sales growth. Finally, we demonstrate that the estimated value of financial statistics increases substantially when we assume the investor can trade in the firm asset and can also hedge using an industry exchange‐traded fund (ETF). This result suggests that professional investors integrate their information search with their hedging behavior.
Authors
- Xin Zheng (ORCID: https://orcid.org/0000-0002-7327-0306)
- Russell Lundholm
Institutions
- University of British Columbia (CA)
Publication Details
- Journal
- Contemporary Accounting Research
- Published
- 2026-09-28
- DOI
- https://doi.org/10.1111/1911-3846.70087
- Primary Topic
- Auditing, Earnings Management, Governance
- Type
- article
- Field-Weighted Citation Impact
- 0.00