PERBANDINGAN VOLATILITAS EWMA, GARCH DAN MONTE CARLO TERHADAP NILAI TUKAR MATA UANG ASING BANK B JB

SAFRIN MARULITUA SIMARMATA, Prof. Dr. Jogiyanto Hartono, MBA.

2014 | Tesis | S2 Manajemen

Penelitian ini membahas perbandingan keakuratan model volatilitas khususnya Exponentiall Weighted Moving Average (EWMA), Generalized Autoregressive Conditional Heteroscedastic (GARCH) beserta klasifikasinya dan Monte Carlo Simulation (MCS) untuk pengukuran risiko pasar dalam rangka menghitung VaR portofolio terhadap nilai tukar mata uang AUD/IDR, EUR/IDR dan USD/IDR Bank Bjb. Nilai posisi devisa neto Bank Bjb per 31 desember 2010 untuk ketiga nilai mata uang sebesar 98,47% dari total posisi devisa neto. Pengujian data imbal hasil ketiga mata uang melalui tahapan stasioneritas, normalitas, heteroskedastisitas dan perhitungan volatilitas ketiga model, yang selanjutnya dilakukan backtesting pengujian validitas model. Model yang lebih akurat dan valid akan dipergunakan dalam perhitungan VaR portofolio. Sejak backtesting Kupiec memiliki kelemahan hanya menitikberatkan pada jumlah eksepsi (kegagalan), Basel traffic light memiliki kelemahan tidak dapat memprediksi kesalahan model VaR yang baik ataupun buruk, dan tes Christoffersen tidak mampu menangkap ketergantungan semua eksepsi, maka penelitian ini menerapkan Kupiec mixed sebagai strategi backtesting yang mampu menangkap ketergantungan antara satu eksepsi ke eksepsi-eksepsi sebelumnya dan juga memperhitungkan jumlah eksepsi yang terjadi. Sementara dalam perhitungan VaR portofolio risiko pasar, penelitian ini menerapkan filtered historical simulation yang lebih realistis dan superior dalam mengestimasi VaR yang mampu menghasilkan residual independent and identically distributed sebagai solusi penyempurnaan dari historical simulation. Berdasarkan hasil penelitian menunjukkan bahwa ketiga nilai tukar AUD, EUR dan USD dalam perhitungan imbal hasil berdistribusi tidak normal, sehingga harus dikoreksi tingkat nilai kepercayaan dengan perhitungan cornish fisher. Hasil analisis selanjutnya, meskipun mata uang EUR memiliki sifat homokedastik, yang berarti perhitungan volatilitas cukup menggunakan standar deviasi, tetapi mata uang EUR dapat dimodifikasi ke dalam pemodelan EWMA, GARCH dan MCS. Di dalam model GARCH sendiri, mata uang EUR memiliki hasil yang signifikan. Hasil pengujian dalam validitas model dengan menggunakan backtesting Kupiec mixed menunjukkan bahwa model volatilitas GARCH dan klasifikasinya dengan tingkat kepercayaan 95% terbukti valid dan akurat untuk ketiga nilai tukar mata uang AUD, EUR dan USD, sementara EWMA valid untuk mata uang AUD dan EUR, dan MCS valid untuk mata uang USD. Hasil estimasi VaR portofolio risiko pasar menggunakan metode filtered historical simulation sebesar Rp797.083.763, memberikan informasi cadangan atau beban modal minimum yang harus disediakan oleh Bank Bjb, selain harus memperhitungkan juga risiko kredit dan risiko operasional.

This study addresses the comparison of accuracy volatility models especially Exponentiall Weighted Moving Average (EWMA), Generalized Autoregressive Conditional Heteroscedastic (GARCH) with its classification and Monte Carlo Simulation (MCS) for measuring market risk in order to calculate VaR of portofolio to exchange rate of AUD/IDR, EUR/IDR and USD/IDR of Bank Bjb. Net open position of Bank Bjb December 31, 2010 for three currency as amount of 98,47% from total net open position. Testing the data return for three currency through stasionerity, normality, heteroscedasticity stage and calculate volatility for three models, which further do backtesting of validity model. Volatility model that is more accurate will be used in the calculation of Value-at-Risk (VaR) of market risk portfolio. Since Kupiec backtest has weakness just focus on the number of exception (failure), Basel traffic light has its disadvantages can not predict good or bad a VaR model, and Christoffersen test are not able to capture the dependence of all exception, this study applies a mixed backtesting strategies which are able to capture the dependence between one exception to the previous exceptions and also take into account the number of exception that occured. While in the calculation of portfolio market risk VaR, this study applies filtered historical simulation that more realistic and superior in estimating VaR to able in produce independent and identically distributed residual as a refinement of the historical simulation solutions. Based on the results of the study shows that all three exchange rate AUD, EUR and USD in the calculation of data return distribution is not normal, so that level of confidence it must be corrected with cornish fisher. Further analysis of the result although EUR currency has homoscedastic nature, which means calculation of volatility only using standard deviation, but the currency EUR can be modified into the EWMA, GARCH modeling and MCS. In the GARCH model itself, the currency EUR has significant results. The test results in the validity of the model using backtesting Kupiec mixed shows that GARCH volatility models and its classification with confidence level of 95% proved that three currencies exchange rate AUD, EUR and USD has valid and accurate model, while EWMA valid for AUD and EUR currencies and MCS valid for USD currency. Results of the portfolio market risk VaR estimation using filtered historical simulation method as amount of Rp797.083.763, gives information reserve capital or minimum capital charge to be provided by Bank BJB, in addition must also take into account the credit risk and operational risk.

Kata Kunci : value-at-risk, exponentiall weighted moving average, generalized autoregressive conditional heteroscedastic, monte carlo simulation, backtesting kupiec mixed, filtered historical simulation