Predicting Stock Market Trends: A Bivariate Regression Analysis of Trading Volume Impact

Abstract

Stock market prediction presents a perennial challenge for investors attempting to forecast the future value of traded stocks at the exchange. The allure of successful predictions, which can yield substantial profits, has driven this pursuit since the inception of stock markets. The 2008 financial crisis highlighted the complexity of market behaviour, emphasizing the need for a deeper understanding of stock market prediction. Regression analysis, a quantitative research method, quantifies relationships between dependent and independent variables. This study conducted a bivariate regression analysis to assess the predictive potential of traded volume on equity price movements within the Nigerian Exchange Limited (NGX or The Exchange), using Guaranty Trust Holding Company Plc (GTCO) historical data. The results revealed a statistically significant correlation (r=0.59) between volume (X) and price (Y). The regression equation (Y = 44.164 + 0.179X) demonstrated that 34.5% of price variance can be predicted from traded volume, indicating a reasonably strong relationship. These findings suggest a moderately robust association between trading volume and stock price movements. This implies that, to some extent, trading volume can guide predictions of stock prices. As investors seek to outperform the market, using linear Regression for stock price prediction remains an engaging area for further research, aiming to enhance return on investment. In conclusion, our linear regression model can generate more precise stock price predictions with a more extensive dataset. Keywords: Efficient market hypothesis, linear Regression, Price prediction, trading volume, Stock market.