Forecasting U.S. Value & Growth Stocks Future Performance : Development & Implementation of a statistical model predicting this performance according to the economic and financial conditions

This thesis aims at forecasting the future performance of Growth and Value Stocks through the development and implementation of a statistical model capable of predicting these performances through macroeconomic data. In order to optimize the development of this model, some research was conducted on the existing literature related to this topic. The main point is that all the shares available on the markets can be either in the value category, they are therefore characterized by high valuation ratios (BV/P, EP, CF/P and D/P) because these stocks are often considered undervalued on the market or in the growth category, in this case, they are characterized by low valuation ratios because these stocks are often overvalued compared to their true value. In addition, trends in outperformance have been investigated. The results show that the U.S. markets experienced a Value Premium until the 2007 financial crisis but Growth Stocks tended to outperform after the financial crisis. Furthermore, economic cycles were also explained as having an impact on the performance of Value and Growth equity returns. There are two schools of thought but both theories could be observed and confirmed in the research results, which makes it difficult to determine which theory is the most accurate. Afterwards, a forecasting model was created and tested on 4 different indices: two representing the U.S. markets for large cap and mid-small cap, one for the European markets and one for the Japanese markets. The result was that the model was able to predict which stock style would outperform the next day with a confidence level close to 50%, which is not enough to be a relevant model.