Implementing Markov Switching Regression Using Best Subset Approach For BSI Stock Price Prediction Analysis

Abstract

Stocks are evidence of ownership of the capital or funds of a company or institution and are represented by a document that includes the par value, the company name, and the rights and obligations described for each owner. Since so many factors affect the rise and fall of stock prices, investors should pay attention to the factors that influence the rise and fall of stock prices to avoid incurring losses or profits when buying and selling stocks. The rise and fall of stock prices can be analyzed with Markov switching regression by trying all possible placements of factors to get the best subset. Public holdings will continue to increase due to nation-building and Sharia Bank Indonesia (BRIS) stock price appreciation. This study aims to determine the impact of increases and decreases in the closing price of BSI stock. The modeling used in this study is Markov switching regression using the best subset approach. The data used in this study are secondary in the form of daily data for the closing price of Bank Syariah Indonesia shares, Inflation, BI Rate, Selling Exchange Rate, Money Supply, and Gross Domestic Product (GDP). Data are obtained from the official BPS website. The results of this study show that Markov switching regression modeling can identify the feasibility of regimes as "bull" and "bear" periods. State 2 indicates an uptrend or "bullish," and state 1 indicates a downtrend or "bearish." The best subset approach obtains the best model with the lowest SSE value. The study concluded that the statistical modeling results of  BSI stock's closing prices during "bull" and "bear" periods provide significant predictors: BI Rate, Selling Exchange Rate, and Money Supply.