Unsupervised machine learning using Python for a pair-trading strategy.
Budget: $30 – $250 USD
In short, a trader who follows a pair-trading strategy will simultaneously take a long and short position in two similar stocks. Hence, it is necessary to scan a group of stocks to classify which stocks similar to others. The pair of stocks is chosen according to their high positive correlation and their contratory prices compared to their historical data, e.g. one is overvalued and another is undervalued.
So there is a scenario to scan 12 stocks based on two criteria namely Return on Equity (ROE) and beta. ROE measures a firm’s financial performance by dividing net income by shareholders’ equity, while represents non-diversified risk of a stock in relation to the overall market. You will need to classify the 12 stocks into two groups. In each group, indicate a pair of stocks with the highest correlation.
Other additional files will be sent.
So there is a scenario to scan 12 stocks based on two criteria namely Return on Equity (ROE) and beta. ROE measures a firm’s financial performance by dividing net income by shareholders’ equity, while represents non-diversified risk of a stock in relation to the overall market. You will need to classify the 12 stocks into two groups. In each group, indicate a pair of stocks with the highest correlation.
Other additional files will be sent.