Прогнозирование рисков и доходности для портфеля акций биотехнологических компаний

The objective of this paper is to estimate volatility and forecast the daily returns of stock using the autoregressive integrated moving average (ARIMA) model and the generalized autoregressive conditional heteroscedasticity (GARCH) model for the time series of the NASDAQ Biotechnology Stock Index (NBI) using the daily close price from the period of January 2013 to December 2017, which is an accounting for 1,258 observations. This study is divided into four sections. The first section is to study of movement of NBI index returns. The second part is to specify the ARIMA model for explanation the data and select the model that has the lowest AIC and BIC value. The third part is to identify the GARCH model fitting for the data also selecting by the lowest AIC and BIC value. The last is to forecast stock price and estimate the volatility. It is found the appropriate model for estimation volatility and forecasting the close price of the NASDAQ Biotechnology stock Index (NBI) is ARIMA(5,1,4), which derived from the comparison of the lowest RMSE and the lowest MAPE equaling to 30.92450 and 0.697796 respectively. Therefore, the daily price from the forecasting the price of NBI and the variance estimation using this ARIMA(5,1,4) model between the period of January 1, 2018 to January 12, 2018 are $3,367.09, $3,369.34, $3.366.81, $3,361.53, $3,355.61, $3,361.64, $3,360.58, $3,360.57, and $3,352.98. The daily volatility of this forecasting model are 45.74, 45.84, 45.89, 45.94, 45.94, 46.05, 46.19, 46.66, 46.54, and 46.54. It is also found at the confident level of 1% the VaR form the estimation using the ARIMA(5,1,4) model between the period of January 1, 2018 to January 12, 2018 are -$132.24, $132.23, -$132.21, -$132.19, -$132.17, -$132.15, -$132.13, -$132.12, -$132.10, and $132.08. Keyword: ARIMA model, GARCH model, Biotechnology stock, NBI, Forecasting model, Time series