Research on Stock Market Prediction based on Inverse Probability of Treatment Weighting Method

Abstract

Although randomized controlled trials (RCTs) are considered the best method to evaluate the impact of interventions, their application in the financial industry is not feasible due to ethical and practical issues. Propensity Score Matching (PSM) minimizes the problem of bias in observational studies. By providing a way to circumvent these limitations. This study uses logistic regression and inverse probability of treatment weighting (IPTW) to evaluate the relationship between market capitalization and important financial indicators, such as revenue and income. Using data from 12 companies, this study identified the key drivers of market capitalization, including revenue, gross profit, and number of employees. The calculation of the average treatment effects (ATE) for Google and Apple showed a good correlation between these financial factors and market capitalization results. The results show that propensity score matching and inverse probability of treatment weighting can provide valuable investment strategies in the stock market by accounting for confounding variables, which are pretty much available for everyone. Though the market is fluctuating in the short run. The average treatment effects estimate the expected return in the long run and might give better market returns and higher Sharpe Ratios for investors.