Objective: The objective of the paper is to verify the Grey relation analysis methods and compare them with known models in financial management. The first part is devoted to MADM (multi-attribute decision methods) and the second one to grey forecasting methods. Method: Localization GRA and globalization GRA multi-criteria models are described and compared with the TOPSIS method. The forecasting GRA(1,1) model based on cumulative data is described, derived and compared with the linear growth model. The models are verified on the data set. Results and Discussion: The globalisation model based on a pair-wise comparison of alternatives is different, and the ranking of alternatives is fully different in comparison with the TOPSIS model. It was found that the localisation GRA model and TOPSIS, both constructed on similarity measures, present similar and comparable results. So, the grey relation approach considers the uncertainty of input data. The forecasting model GRA(1,1) is computed and compared with the growth model. Results show that the growth model is the trendy model, and the non-trendy series is overestimated or underestimated, so the GRA model looks more robust. Research Implications: The problem of imprecise data is investigated. And it is a realistic condition for practical decision-making. The problem of lack of data is often a phenomenon in decision-making. Traditional statistical methods cannot be applied. Therefore, grey relation methodology is one approach to overcoming the problem of imprecise data and lack of data. Uncertainty and lack of data problems are topics important for research and the managerial community. Originality/Value: Sometimes, in financial decision-making, high-quality data or long-term time series is not at our disposal, and on the other hand, decisions have to be made. This problem is often neglected and solved by not-relevant traditional statistical models. Therefore, methods that allow the solving of such problems are useful. It was shown that a grey relation methodology for multicriteria financial performance evaluation and forecasting shows more robust and stable results and could be more relevant for imprecise data or short time series data in financial decision-making problems.
Read full abstract