Loss given default (LGD) assessment is crucial for banks since it enables them to derive a more conservative estimate of loan recovery rate after factoring time value of money and recovery cost. Limited data on loan losses are publicly available in emerging market economies like India as bank loans are not tradable instruments. This study attempts to cover this gap and a distinctive effort has been made to provide insights on the LGD statistics, its temporal pattern and key factors that contribute to loan losses in banks in emerging countries such as India. The study derives sector level, regional level, security wise, and bank group wise estimates of LGD statistics. It empirically observes that banks have better recoveries if loans are settled through better and timely negotiation after arriving at a compromise. Moreover, higher collateral margins and availability of better security leads to better recoveries and lower LGD. Bank loan recoveries are higher (hence low LGD) in case of fund based facilities, higher collateral margins and for assets backed by more liquid collaterals. Further, regional factors also significantly influence LGD. The findings of this paper would enable banks, as well as policy makers and auditors, to understand the importance of LGD and derive benchmark estimates for calculation of expected credit loss as well as credit risk regulatory capital. The facility level multivariate regression based LGD rating model developed in this article may be used by the banks to predict future losses for evaluating recovery risk of a standard loan facility. The study findings suggest that banks need better and prompt recovery strategies to reduce the burden of non-performing assets and future loss provisions. The LGD trends over various time periods have been assessed to establish their linkage with probability of default (PD). This can be used in stress testing bank capital as well as loan loss provisions.
