Strategies for Handling Negative PEG Ratios in Equity Cost Research
In equity cost research, when forecasted EPS is unavailable, researchers often use the PER/earnings growth rate as a substitute for the PEG ratio. However, this alternative formula can generate numerous negative values and outliers, affecting result accuracy. This article explores how to handle negative PEG, including whether it can be set to zero, and suggests referencing relevant journal literature to establish methodological justification.
When conducting research on the cost of equity, researchers often rely on model frameworks from prior literature. However, when forecasted earnings per share (EPS) data are unavailable, making it impossible to directly calculate the standard PEG ratio, some researchers instead adopt an alternative formula: the price-to-earnings ratio (PER) divided by earnings growth. This alternative method can generate a large number of negative PEG values in practical applications, accompanied by significant outliers, posing challenges to the robustness of research findings.
Regarding the handling of negative PEG values, there is currently no unified consensus in academia. One common practice is to replace all negative PEG values with 0, but whether this operation is reasonable requires careful evaluation. Setting negative values to zero may distort the sample distribution and introduce systematic bias, especially when the proportion of negative values is high, potentially affecting the validity of subsequent regression analyses or factor tests. Therefore, it is recommended that researchers first diagnose the root causes of negative values—for example, negative earnings growth (such as in loss-making firms) or negative P/E ratios (such as in loss-making or marginally profitable firms)—and consider whether to exclude these observations or adopt other adjustment methods (such as Winsorization) to mitigate the impact of outliers.
To obtain correct results, researchers may refer to the following strategies:
- Sensitivity analysis:Report results from the raw data, after excluding negative values, and after setting negative values to zero, respectively, to test the robustness of conclusions.
- Alternative variables:Consider using other growth indicators (such as historical growth rates, analyst long-term growth forecasts, if available) or alternative valuation ratios (such as P/B, EV/EBITDA) for cross-validation.
- Literature support:Consult relevant journal articles to find precedents for similar handling methods. For example, some studies exclude observations with negative growth or negative earnings when calculating implied cost of equity, or employ truncation. It is recommended to search for papers on the application of PEG ratios in journals such as the Journal of Financial Economics and Review of Accounting Studies to obtain methodological references.
Regarding the question of whether all negative PEG values can be set to 0, it should be clarified: setting to zero implicitly assumes that negative PEG represents no growth or zero growth, but in reality, negative PEG may reflect declining earnings or cyclical fluctuations, not zero growth. Therefore, simply setting to zero may introduce measurement error. A more prudent approach is to try setting to zero in robustness checks and compare whether results change significantly; if results are consistent, it is acceptable; if not, the handling method needs to be reconsidered.
In summary, handling negative PEG values requires consideration of research design, sample characteristics, and literature conventions, avoiding mechanical operations. It is recommended that researchers document the processing steps in detail and transparently disclose them in the paper to enhance the credibility of the results.