Efficiency and Consistency Assessment of Value at Risk Methods for Selected Banks Data

Abstract

The study assesses Value at Risk (VaR) methods with respect to their efficiency and consistency in selected banks of the Nigeria Stock Market. The daily data on share prices of each bank was used from 2006 to 2018. The Value at Risk of each bank was estimated and the predictive performance of each method was assessed using the Failure Ratio and the Confidence Interval. The quality of each method was assessed based on the efficiency and consistency of the estimates. The VaR of each bank was estimated using Historical Simulation, Kernel Estimator, Empirical Estimator and Weighted Mean methods. The weighted mean method had the least estimates while Kernel estimator method had the highest estimates. The Failure Ratio and Confidence Interval show that Historical and Empirical methods had the least number of rejections at both confidence levels. The efficiency and consistency of the various methods shows the Historical Simulation and Weighted mean method had the minimum mean square errors (MSE). The Banks A, D and E gives an efficient and consistent result with Historical Simulation while B and C, is more efficient and consistent with weighted mean method.

Value-at-Risk Analysis in Risk Measurement and Formation of Optimal Portfolio in Banking Share

“…The value of this ratio is less than the critical value of Chi-Square (LRUC1 < 3.841), at the 95% confidence level, the historical simulation method is valid and is an accurate method for estimating the VaR value in the crisis market period. Musa et al (2020) showed that historical simulations were accurate and had minimum mean square error (MSE). Wicaksono et al (2014) and Machfiroh (2016) proved that the historical simulation method accurately estimated the VaR value.…”

“…Machfiroh (2016) showed that Value at Risk (VaR) with historical simulation methods to measure stock risk on the LQ-45 index is said to be accurate. Musa et al (2020) revealed that the Value at Risk (VaR) historical simulation method was the most suitable method in estimating the minimum capital at 3 banks out of 5 estimated banks. Susanti et al (2020) also applied VaR to banking single stocks and portfolios by proving historical simulation was the best and consistent method in assessing VaR of single stocks (BNI shares) and portfolios.…”