Asset Allocation Strategy with Monte-Carlo Simulation for Forecasting Stock Price by ARIMA Model

Abstract

Asset allocation was an important topic in the financial market and Monte-Carlo simulation always played a key role. However, the traditional Monte-Carlo simulation allocated assets by historical data and could not react to short-term market volatility. Since introduced in 1976, The Auto-Regressive Integrated Moving Average model or ARIMA model of time-series analysis showed its ability to provide forecast to time series, including the stock price. Therefore, in this paper, ARIMA and Monte-Carlo simulation was combined. By applying the Monte-Carlo simulation to stock price prediction provided by ARIMA model time series analysis, a more flexible weekly trading strategy was created. ARIMA model predicted stock prices for five selected popular tech stocks, AAPL, AMZN, TSLA, TWTR and MSFT in 5 days for the first trading week of Sept of 2021 and allocated the asset combination by Monte Carlo Simulation. The combination was compared to the allocation provided by the traditional Monte Carlo simulation. The model was proven to be profitable and has higher profitability and accuracy than using Monte Carlo simulation alone.