A Comparative Analysis of the Measurement of BIST 30 Stock Performances with Radial and Non-Radial DEA Models

Creative Commons License

Isbilen L.

EKOIST-JOURNAL OF ECONOMETRICS AND STATISTICS, sa.44, ss.109-128, 2026 (ESCI)

Özet

In this study, the performance of stocks traded on the BIST 30 index was subjected to a comparative analysis using radial (CCR) and non-radial (SBM: Slacks-Based Model) data envelopment analysis models. The main objective of this study is to reveal to what extent the concept of full efficiency plays a role in identifying stock performance. Accordingly, the findings of the radial CCR model, which measures weak efficiency, and the non-radial SBM model, which aims for full efficiency, were compared. The inputs of the efficiency analysis are P/E, P/B, EV/EBITDA, and EV/ Sales; the output is the closing price of the stock. Initially planned as an output, the net income per share data could not be included in the analysis due to the limitations of DEA in handling negative values. The study uses data from September 2024, sourced from & Idot;& scedil; Yatirim. According to the analysis results, Ford Otosan, Migros, and Turkish Airlines were found to be efficient under both the CCR and SBM models. Although efficiency scores varied for other stocks, it was observed that their performance rankings remained similar. The presence of a single output in the model reduced the efficiency frontier of the output-oriented SBM model to that of the CCR. The findings for the output-oriented versions of both models were identical. For this reason, comparisons were conducted using input-oriented versions. In these versions, it has been demonstrated that SBM is more successful than CCR in identifying the sources of inefficiency. The study's hypothesis, "The full efficiency measurement performed by SBM, considered slack variables, produces more discriminatory results than the CCR model, which ignores these variables," has been confirmed. This study, which focuses on model sensitivities, does not claim that the stocks identified as efficient will provide higher returns. This innovative approach, based on the assumption that the prices of the identified efficient stocks are more rationally priced than their market peers and possess a more advantageous multiple structure, is considered a data-driven and more realistic guide in the portfolio selection process.