The Relationship Between Credit Referencing and the Level of Non - Performing Loans of Commercial Banks in Kenya

Abstract

The concept of credit referencing was introduced in Kenya by the Central Bank in February 2009 as one way of providing vital information for credit assessment in an attempt to reduce incidences of loan default and therefore reduce nonperforming loans. A greater portion of nonperforming loans in Kenya were as a result of information asymmetry that enabled serial defaulters to borrow from multiple commercial banks with no intention of repaying as their past credit history was not available to prospective lenders. The main reason for the introduction of credit referencing was that it would enable sharing of credit information by lenders and thus reduce loan default through elimination of information asymmetry. The objective of this study was to investigate the effect of credit referencing on the level of NPLs as measured by the ratio of NPLs over total loans and advances of commercial banks in Kenya. The hypothesis was that credit referencing eliminates information asymmetry and therefore reduces loan default rates. The study employed descriptive as well as correlation research designs. A multiple regression analysis was used to analyze time series empirical data to examine the relationship between credit referencing and the level of NPLs by establishing correlation coefficients between the aggregate number of credit reports obtained by forty two commercial banks and customers and the aggregate NPLs ratio over total loans and advances of the commercial banks. The study period was February 2009 to May 2013. The findings were that the ratio of NPLs over total loans and advances is negatively related with the growing number of credit referencing reports obtained from credit referencing bureaus. The analysis showed that 33% of variations in the ratio of NPLs over total loans and advances were explained by credit referencing. The coefficient of multiple determinations, as measured by adjusted R-square, was 29% implying that 29% of variations in the ratio of NPLs over total loans and advances were explained uniquely by credit referencing. These findings were in line with the expectations of the CBK for launching the concept of credit referencing and the findings of Udell (2003), Brown (2007), Pagano (2009) and Mutangili (2011). As a result, the concept of credit referencing was observed to have had a positive impact on reducing NPLs. Therefore, the study recommended that commercial banks should fully adopt and practice the concept of credit information sharing at every stage of credit appraisal so as to enable them weed out potential loan defaulters.