Previous

Part - V

Quantification consists of applying value-at-risk (VaR *) and earning-at-risk (EaR) approach for measuring and controlling quantifiable bank-wide risks (across risk types and business units). However, quantification does not specifically address the following: -

• inadequate corporate governance processes;
• reputational risk;
• regulatory risk and
• long-term strategic risk.

Bottllenecks to establlish efficient bank-wide risk management framework

a. Data Adequacy
There are several significant VaR models and techniques available for the banks. However, all these models require lot of qualitative and quantitative historical data inputs/information related to credit and facility scoring, probability of customer’s defaults (PD **), articulate information of different collaterals to evaluate recovery rates (RR ***).

Significant conceptual and practical issues pertaining to resources, technology, data, processes and culture have to be taken into account when implementing this process. Because risks are highly interdependent and cannot be segmented and managed solely by independent business units. Bank should adopt an evolutionary bank-wide risk management, which would add significantly more value than fragmented risk management process. The objective of bank-wide risk management should continuously enhance the integration of credit, market, operational and strategic risks.

b. Lack of efficiency
As stated above most of the employees’ skill set in risk management department are restricted to verification of documents and related accounting calculations as part of pre and post fact analysis of the credit. The simple apprehension of our study is most of the banks particularly nationalized banks does not have appropriate data and efficient people to comprehend and go forward for advance measurements as per Basel II. Also, risk management seemed as part of corporate banking. Is this parlance appropriate?

Next

* VaR is the maximum loss estimated over a given holding period at particular confidence level.
VaR = capital exposure x probability of occurrence x loss in the event of occurrence;

** PD = the likelihood that a loss occurrence takes place. Ex. The probability of customer default can be estimated from historical data by tracking migration of borrowers from their previous rating to a default rating.
*** LGD = 1 - recovery rate (i.e., RR).