Revenue cost efficiency deviation
factor.formula
Revenue-cost regression model:
Revenue cost efficiency deviation factor:
in:
- :
The operating income of the i-th quarter has been Z-Score standardized to eliminate the impact of dimension and improve the robustness of the regression model.
- :
The operating costs in the i-th quarter are also Z-Score standardized to maintain the same dimension and distribution characteristics as the operating income data.
- :
The intercept term of the regression model represents the expected operating income level when the operating cost is 0, which is often ignored in actual regression.
- :
The slope coefficient of the regression model indicates the change in expected operating income for every unit change in operating costs, reflecting the marginal relationship of historical revenue costs.
- :
The residual term of the regression model represents the deviation between actual revenue and expected revenue based on historical relationships, that is, the forecast error of the model. The residual term here reflects the excess performance of current operating activities and is the core component of this factor.
- :
The residual term of the latest period (period t) reflects the latest deviation of revenue cost efficiency and is the final factor value.
- :
i ∈ {0, 1, 2, ..., N-1}, which indicates the serial number of each quarter in the lookback period, where 0 represents the most recent quarter and N is the length of the lookback window. The default value is N = 8, which means the most recent 8 quarters are looked back.
factor.explanation
This factor first selects the operating income (Revenue) and operating cost (Cost) data of the most recent N quarters, and performs Z-Score standardization processing respectively to eliminate the impact of data dimension and distribution differences and improve the robustness of subsequent regression. Then, the standardized operating income and operating costs are linearly regressed using the ordinary least squares (OLS) method to construct a revenue-cost relationship model based on historical data. The residual term (ε) obtained from the model regression reflects the degree of deviation between the actual revenue and the model prediction value. The residual value (εt) of the most recent quarter is extracted as the final RROC factor value. A positive residual indicates that the current revenue is higher than expected, which may indicate an improvement in operating efficiency, and vice versa. It may indicate a decrease in operating efficiency. The absolute value of the factor value represents the degree to which the actual revenue deviates from the expected revenue.