earticle

영어

The purpose of this study is to investigate the risk management of finacial institutions by using VaR(Value at Risk) method. It was analyzed and compared with differences between the real profit-loss amounts calculated by using data of 1997 one year and VaR estimates investigated by using data of 3 months(72days) derived from Jan, 1998 to March 1998. VaR estimates was derived from three VaR methods, as the time weighted average method, MonteCarlo simulation method and historical simulation method. The empirical results are as follow: First, we found that in case the real values exceed expected loss amounts at the 95% confidence level, the lowest error rate came out when λ=0.99 of time weighted average method. These results may imply that in case weighted values allow at the nearest data more than past data, the expected excess days decrease in the time weighted average method. Second, empirical result proved that MonteCarlo simulation method with the lowest error rate had most smallest mean of bias in the significance test between the real observation values and VaR values. These results may imply that time weighted average method is suitable in order to manage the situation being placed in extreme financial distress situation and MonteCarlo simulation method is suitable in order to minimize the opportunity cost according to retention of reserve money caused the loss. Specifically, we found what the financial institutions dealing with money goods and derivatives responding sensitive to price volatility risk and credit risk should be managed the integrated risk management. We may suggest financial institutions could use effectively the market risk values estimated by VaR method for risk management.