The Quality and Drivers of Own Credit Risk Measurement Disclosures for Fair Value Liabilities by European Banks

2014 
This research paper studies the quality of own credit (OC) risk disclosures in European banks by analyzing their annual reports. By defining and applying a self-developed disclosure index that incorporates both the (i) relative quantity and (ii) richness of the information provided, it is found out that disclosure in annual reports about own credit risk tends to be low on average. Furthermore, there is a great difference between disclosure qualities of the analyzed banks. By using a multiple linear regression, various variables are tested for their influencing impact on OC disclosure quality. The analyzed variables were selected on a holistic basis and relate to a bank’s earnings situation (profit and loss statement view), risk situation (balance sheet view), return on invested capital situation (investors’ view), size and potential information asymmetries between the bank and investors. Our results support the overall hypothesis that own credit disclosure is a function of firm-specific variables. A statistically significant impact is found for variables related to the impact of the own credit adjustment on the bank’s return on equity ratio (RoE) and the overall amount of financial liabilities designated at fair value. In case the impact is found to be large, banks disclose more and better information. We interpret these findings on the basis of the importance of the RoE performance measure in the financial industry. RoE represents a particularly important ratio for shareholders as it is often used as an approximation to what extend the bank was able to generate value for its shareholders. In general, our findings support rather the view that disclosure quality tends to be influenced to a higher degree by the own credit adjustments impact on banks’ earnings than by the potential risk structure of the banks liabilities.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    42
    References
    1
    Citations
    NaN
    KQI
    []