An Algorithm for the Detection of Revenue and Retained Earnings Manipulation

This paper presents a statistical analysis confirming the former empirical findings that positive differences between the growth rates of P-Score and Z-score appears in financial statement data of companies involved in major financial fraud. The paper examines firms that engaged in fraud in the late 1990's through early 2000's. The paper reports the results of regression analysis, using ratios, from financial statement data used in the calculations of P-Score and Z-Score. The results show that positive values of the difference between the growth rates of P-Score and Z-Score correlate with Net Income, Revenue, Retained Earnings and Total Equity ratios. Both ratios represent the financial statement areas where most identified fraud occurred. The findings imply that positive differences between the rates of growth suggest financial statement manipulation. The standard error of the estimate shows the early linear regression to be coarse. The final part of the paper optimizes the linear regression formula and discusses its limits. The paper shows the potential uses of Extensible Business Reporting Language (XLRB) for getting the necessary values for algorithm calculations