The task at hand is to considering the data here, come up with a set of rules to maximize the trading performance.
Tasks The file at https://ortex-static-files.s3.eu-west-2.amazonaws.com/SignalsQuantProgrammingExercise.csv is the output of an alpha signal generator. The file contains 6 columns.
| Date | Date when the signal triggered, yyyy-mm-dd |
| Ticker | The ticker identifying the stock |
| Signal_return | The percentage profit at the recommended holding day |
| Rec_holding_days | The number of trading days recommended to the position. |
| Buy | True if the signal suggests to buy, false if sell short. |
| Significance | The signal generators expected significance of the signal |
Considering all but signal_return, we want you to come up with a set of rules or a ML model to maximize the trading performance.
The code should work out a way to select the best trades and how much of your capital should be invested in any trade.
Considerations: You should only consider buy trades.
Trades should be closed ‘rec_holding_days’ trade days after the ‘date’.
Only trades that are closed on or before the 31st of December 2020 should be considered.
For the purpose of this exercise, Monday-Friday can be considered trading days.
For the purpose of visualizing and weighting, the starting capital is $100 000.
Output: A well-formatted, labelled chart showing performance over time, including cash in bank and cash invested. A percentage increase over time. A runnable, trained model that can be tested against an unseen dataset.
BUILD