STOP LOSS ORDER PERFORMANCE ANALYSIS BASED ON VALUE AT RISK USING EGARCH (1,1) MODEL IN EUR/USD FOREX TRADING

• Main Author: MIRAJ DANAR (NIM: 29112059) Pembimbing: Dr. Ir. Budhi Arta Surya, MSc., HALLEY
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/18577
• Institution: Institut Teknologi Bandung

Forex trading especially EUR/USD trading not only may bring promising profit to newcomers traders but also loss. Newcomer traders may be only focus in generating profit and sometimes loss is being overlooked by them. One of trading strategis in forex trading used by traders is moving averages cross over to generate trading signals. Scalping is one of trading styles in forex market. Scalping style depends on the volume of transaction in forex trading. Scalping really depends on the accuracy of trading strategies in generating trading signal. Sometimes false signal occurs and to reduce the loss in trading, stop loss order is the best alternatives to reduce the loss. This research emphasizes in determining stop loss order based on Value-at-Risk (VaR) in scalping trading style using moving averages cross over. 4 types of moving averages ; simple moving averages (SMA), linear weighted moving averages (LWMA), smoothed moving average (SMMA), and exponential moving average (EMA); are used to test how good VaR can be used as stop loss order. From trading simulation using 17785 EUR/USD data from December 2, 2013 until February, 28 2014, all of moving averages trading strategies may cause 38.6% average loss from beginning balance. <br /> <br /> <br /> VaR depends on the variance. 2 kinds of variance, conditional and unconditional, will be calculated and compared. Conditional variance calculated based on EGARCH (1,1) model based on 17785 EUR/USD data from December 2, 2013 until February, 28 2014 in 5 minutes time frame. Unconditional variance calculated based on the position of trading. From in-sample trading simulation, by using stop loss order based on VaR using EGARCH (1,1), the end balance will only have 15.08% average loss compare to 38.6% average loss without using any stop loss order. Using unconditional VaR as stop loss order will generate 31.84% average loss. <br /> <br /> <br /> Backtest conducted to test the performance of EGARCH (1,1) model as the basis in VaR calculation. Backtest conducted using 1000 period of EUR/USD data from March, 3 2014 until March, 6 2014 in 5 minutes time frame. From backtest trading simulation, stop loss order using VaR based on EGARCH (1,1) can reduce the loss to 1.515% from beginning balance from 4.785% loss when no stop loss order used averaged for all moving averages trading strategy. Unconditional VaR as stop loss order can only reduce the loss to 1.73% from beginning balance. <br /> <br /> <br /> It can be concluded that VaR based on EGARCH (1,1) can be used as stop loss order to reduce the loss occured during trading. To implement this model, it is recommended that this concept need to be combined with “Expert Advisor” to reduce the time needed in calculation.