بررسی و مقایسه شاخص های ARMS و BSI در اندازه گیری احساسات سرمایه گذاران با هدف پیش بینی روند قیمت سهام (مقاله پژوهشی دانشگاه آزاد)
درجه علمی: علمی-پژوهشی (دانشگاه آزاد)
آرشیو
چکیده
اهداف: در این پژوهش، شاخص های احساسات سرمایه گذاران در بورس اوراق بهادار تهران، اندازه گیری و میزان موفقیت آنها در پیش بینی آینده بازار سهام ارزیابی شده است. در این راستا و بر اساس محاسبات انجام شده، روند قیمتی شرکت ها و جریان کلی بازار در دوره های آتی نیز پیش بینی و شاخص مناسب معرفی شده است.روش: برای اندازه گیری احساسات سرمایه گذاران از روابط مربوط به دو شاخص ARMS و BSI استفاده و احساسات برای دوره های یک، سه و شش ماهه در بازه زمانی 1398 تا 1402 محاسبه شده است. برای ارزیابی میزان موفقیت این شاخص ها نیز از پس آزمون استفاده شده است.یافته ها: یافته ها نشان داد درصد موفقیت شاخص های اندازه گیری احساسات ARMS و BSI در پیش بینی روند قیمتی سهام در بازه شش ماهه، بیشتر از بازه های سه ماهه و ماهانه است. شاخص BSI نشان می دهد سرمایه گذاران حقوقی نسبت به سرمایه گذاران حقیقی پیش بینی مناسب تری از بازار ارائه می کنند. نتایج کلی پس آزمون شاخص احساسات سرمایه گذاران نیز نشان می دهد شاخص ARMS نسبت به BSI پیش بینی مناسب تری از روند قیمتی سهام ارائه می کند؛ بنابراین شاخص اندازه گیری احساسات ARMS در دوره های شش ماهه نماینده اندازه گیری احساسات سرمایه گذاران درخصوص سهام پذیرفته شده در بورس اوراق بهادار تهران معرفی می شود.نوآوری: نوآوری این پژوهش نسبت به پژوهش های گذشته، بررسی پس آزمون شاخص اندازه گیری احساسات سرمایه گذاران در بازار سهام، ارائه شاخص بهینه و پیشنهاد دوره زمانی مناسب در راستای افزایش قدرت پیش بینی آن است.Exploring and Comparing ARMS and BSI Indices for Measuring Investor Sentiments to Predict Stock Price Trend
This study focused on measuring investors' sentiments in the Tehran Stock Exchange (TSE) and evaluating their effectiveness in predicting future market trends. The study utilized two indices, ARMS and BSI, to gauge investors' sentiments and their performance in predicting stock market behavior was assessed over different (monthly, three-month, and six-month) time periods. Back-testing was employed to determine the accuracy of these indices. The findings revealed that the success rate of the indices in predicting stock prices was higher over a six-month period compared to shorter durations. Moreover, the BSI indicated that institutional investors tended to have a more accurate prediction capability than individual investors. Furthermore, the back-testing results demonstrated that the ARMS index outperformed the BSI index in forecasting stock prices. Therefore, the ARMS index, particularly over a six-month timeframe, served as a reliable indicator for predicting investors' sentiments. This research contributes by exploring the application of back-testing on investors' sentiment indices in the stock market, providing an optimized index and recommending an appropriate time period to enhance its predictive power.Keywords: Investors’ Sentiments, ARMS Index, BSI Index, Back-Testing, Tehran Stock Exchange (TSE). IntroductionInvestor sentiment holds significant importance in financial markets as recognized in the financial literature (Sun et al., 2016). The presence of sentiment and the potential risks and costs associated with arbitraging against sentiment-driven investors (Brunnermeier & Pedersen, 2005) may lead rational investors to deviate from standard pricing principles. Behavioral finance, in contrast, provides a fresh perspective for studying the influence of investor sentiment on stock market returns (Baker et al., 2007; Hribar & McInnis, 2012). The reality is that investor sentiment can impact asset prices, allowing swing traders to generate profits and sustain their presence in the market over the long term (Palomino, 2016).In Iran, several studies have examined investor sentiment and its correlation with financial and economic variables. Noteworthy studies include those conducted by Nadiri and Panahian (2023), Jalilvand and Rostami Nowroozabad (2018), Gholami Jamkarani et al. (2019), Muridipour et al. (2020), Gol-Arzi and Piri (2022), Dadger et al. (2023), and Jalilvand et al. (2016). Furthermore, numerous studies have demonstrated the predictive value of investor sentiment in forecasting stock returns (Brown & Cliff, 2004; Finter et al., 2012; Kim et al., 2019; Lee et al., 2002). In this context, the present study aimed to introduce, for the first time, a more suitable index for measuring sentiment in the Tehran Stock Exchange (TSE), utilizing the ARMS and BSI indices. Furthermore, through back-testing, the study evaluated the appropriate timeframe for calculating this index. Materials and MethodsThe research was conducted from 2019 to 2023, encompassing a 5-year period. The sample comprised 380 companies listed on the TSE, of which 279 companies were selected for data analysis and measuring sentiment indicators. For sample selection in subsequent years, all companies admitted to the TSE before 2019 were considered, unless they were delisted from the exchange or had ceased operations for an extended period by the end of the research period in June 2023.The research methodology involved calculating investor sentiment indices, specifically the ARMS and BSI indices, on a monthly basis, as well as using simple and moving average calculations over quarterly and 6-month periods. Subsequently, a backtest was conducted to estimate the success rate of the indices for different timeframes: monthly, quarterly, and 6 months. A success rate exceeding 50% was interpreted as a measure of success, while a rate below 50% indicated failure. In other words, when the sentiment indicators correctly predicted the future price direction of the stocks (based on both the number and market value of the desired shares) in the upcoming period with an accuracy of over 50%, it was considered successful; otherwise, it was deemed a failure. Finally, based on the performance of the most effective index, a market forecast for the upcoming period was provided. FindingsThe findings indicated that the ARMS sentiment measurement index exhibited a higher success rate in predicting stock price trends on the TSE over a 6-month period compared to 3-month and monthly intervals. This suggested that the ARMS index provided a more accurate forecast of the stock prices of listed companies over a longer timeframe. Similar observations were made for the BSI sentiment measurement index. Furthermore, the BSI index highlighted that institutional investors offered more reliable predictions of market and stock price trends than individual investors.Table 1 presents the results of back-testing the investors' sentiment indices for stocks listed on the TSE. The table demonstrates that the ARMS sentiment measurement index outperforms the BSI in predicting stock exchange price trends. Additionally, the success rate of institutional investors' BSI index surpasses that of individual investors' BSI index. Based on these findings, the paper recommends adopting the ARMS sentiment measurement index as a representative measure of investors' sentiment for stocks listed on the TSE, particularly over a 6-month period. Discussion & ConclusionTo date, no study has addressed the issue of identifying the appropriate index, either ARMS or BSI, for measuring investor sentiment and predicting stock price behavior. Thus, the objective of this study was to propose a suitable index for measuring investor sentiment and market behavior based on the past approach and behavioral finance theory. Our findings revealed that the ARMS index exhibited a higher level of accuracy in predicting stock price trends over a 6-month period. This finding supported the hypothesis that investor sentiment has a more pronounced impact on stock prices in an inefficient or weakly efficient market, such as the Tehran Stock Exchange (TSE). The implications of this study are twofold. Firstly, shareholders and portfolio managers can utilize the results to construct optimal stock portfolios by considering the future price behavior of stocks listed on the TSE. Secondly, regulatory bodies, such as the Securities and Exchange Organization and the supervision departments of the exchange, can employ these findings, particularly the recommended index for measuring investors' sentiments over a 6-month period in conjunction with other control tools to monitor suspicious behavior by various market participants.