The Use of Python & Machine Learning to Optimise a Portfolio of U.S. Small-Cap Companies
Resumen
This paper deals with the disruption of data analytics and machine learning in the investment management industry. In particular the small cap industry, which will be deeply analysed and studied in order to find interesting investment opportunities. This mentioned opportunities will be found from the benchmarking index of the small cap universe , the S&P 600 Small cap index which will also be fully studied in the paper. Once all of the 600 components of the index are analysed the portfolio constructing process will allow the creation of a diversified portfolio which will be able to beat the market.
The prementioned process is a combination of the most important theories in the portfolio management history, which will be combined with a fundamental analysis to help the portfolio maximise returns. The process will start by using the k-mean algorithm to find 3 clusters of well diversified stocks. Once the groups are clear the stocks will be analysed using the Fama & French 3-Factor Model to deeply understand the nature of the success of the portfolio. Then Joel Greenblatt´s famous magic formula will be adapted in order to fit in the small cap investment universe. Finally , when the magic formula has selected the 30 stocks which will compose the portfolio, Markowitz´s efficient frontier model will identify the most efficient distribution of the 30 stocks in the portfolio, in order to find the portfolio which maximises Sharpe ratio thus the reward investors receive for the undertaken risk in the portfolio.
Once the portfolio has been constructed the results will be compared and contrasted with the index performance as a check of the well-functioning of the new magic formula This paper deals with the disruption of data analytics and machine learning in the investment management industry. In particular the small cap industry, which will be deeply analysed and studied in order to find interesting investment opportunities. This mentioned opportunities will be found from the benchmarking index of the small cap universe , the S&P 600 Small cap index which will also be fully studied in the paper. Once all of the 600 components of the index are analysed the portfolio constructing process will allow the creation of a diversified portfolio which will be able to beat the market.
The prementioned process is a combination of the most important theories in the portfolio management history, which will be combined with a fundamental analysis to help the portfolio maximise returns. The process will start by using the k-mean algorithm to find 3 clusters of well diversified stocks. Once the groups are clear the stocks will be analysed using the Fama & French 3-Factor Model to deeply understand the nature of the success of the portfolio. Then Joel Greenblatt´s famous magic formula will be adapted in order to fit in the small cap investment universe. Finally , when the magic formula has selected the 30 stocks which will compose the portfolio, Markowitz´s efficient frontier model will identify the most efficient distribution of the 30 stocks in the portfolio, in order to find the portfolio which maximises Sharpe ratio thus the reward investors receive for the undertaken risk in the portfolio.
Once the portfolio has been constructed the results will be compared and contrasted with the index performance as a check of the well-functioning of the new magic formula
Trabajo Fin de Grado
The Use of Python & Machine Learning to Optimise a Portfolio of U.S. Small-Cap CompaniesTitulación / Programa
Grado en Administración y Dirección de EmpresasMaterias/ categorías / ODS
KBNPalabras Clave
Key Words: Small Caps, Python, Data Analytics, Machine Learning, K-means, Clustering, Fundamental Analysis, Portfolio Management, Efficient Frontier, Sharpe Ratio, Fundamental Analysis, Joel Greenblatt, Fama & French and MarkowitzKey Words: Small Caps, Python, Data Analytics, Machine Learning, K-means, Clustering, Fundamental Analysis, Portfolio Management, Efficient Frontier, Sharpe Ratio, Fundamental Analysis, Joel Greenblatt, Fama & French and Markowitz