Abstract:
Nowadays, stocks investment has been increasingly growing in the society. In
investing activities, there are risks that may be experienced by investors. However,
sometimes many investors do not realize how much risk they might suffer in the
future. One way that can be done to measure this risk is to calculate Value at Risk
(VaR) and Expected Shortfall (ES). This thesis will discuss the calculation of VaR
and ES values using two methods, including the Historical Simulation and Monte
Carlo Simulation method for digital bank stock portfolio. Furthermore, the VaR
value will be tested for accuracy using Kupiec Backtesting method with the
loglikelihood ratio approach. From the results of VaR and ES calculations using
Historical Simulation method sequentially obtained results of IDR 6,006,718 and
7,474,493 for 99% confidence level, IDR 4,135,857 and IDR 5,106,761 for 95%
confidence level, and IDR 3,219,885 and IDR 4,388,922 for 90% confidence level.
While the results of VaR and ES calculations using Monte Carlo Simulation
sequentially obtained results of IDR 10,797,904 and 15,272,779 for a 99%
confidence level, IDR 5,376,949 and IDR 9,159,777 for a 95% confidence level,
and IDR 3,417,553 and IDR 6,868,538 for a 90% confidence level. Based on these
results it is found that the results of VaR and ES are directly proportional to the
confidence level used. In this case, the backtesting test results that Monte Carlo
Simulation produces a more accurate VaR value compared to the Historical
Simulation.
