Novel variants of the TOPSIS algorithm to select and rate the bank counterparties

Credit rating agencies (CRAs) assign ratings to banks using the through-the-cycle (TTC) approach, which often fails to reflect the current condition of banks. Selecting bank counterparties is crucial in the derivatives market, with credit ratings typically guiding this choice. This study introduces two innovative variants of the technique for order of preference by similarity to the ideal solution (TOPSIS) for selecting and rating bank counterparties. These variants, TOPSIS1 and TOPSIS2, depart from the traditional TTC approach by using point-in-time analysis. We analyse the TOPSIS scores and rankings using statistical measures like Spearman's rank correlation coefficient. The results show that TOPSIS2 is a practical, interpretable method for rating unrated banks, predicting upgrades/downgrades, and mitigating counterparty credit risk (CCR).