Week #5 Update


This week in our Stock Algorithm Project we met with the sponsor to discuss our current results and what we should expect to see with them. We also discussed a bit of what to do going forward.

So back from our prior meeting we discussed what we wanted to do for the week ahead to get everyone on the team up to speed and at the same level. We met and discussed what we should all have ready to go for this week. When we met this week we were able to show our status of what we had worked on through the week and then discuss any issues we were having. Currently it looks like everyone on our team is up to speed with the algorithm from Quantconnect and the provided algorithm from the sponsor. We have all tested and shown the results we’ve obtained from running the algorithms with our assigned ticker symbols.

In this weeks meeting we’ve set a few tasks going forward into next week. One was to try and get the results to print out a report to a document instead of the terminal. The other was to run the algorithm again but using the same testing dates for our assigned tickers instead of different test dates based on each tickers up trend and down trends. This will allow us to get a better evaluation of ensuring the algorithm is working as expected. What I mean by that is we each have two tickers assigned to us. For mine one is for the S&P500 and the other one is the inverse of the S&P500. So what we should see when we run the same test dates is if the S&P500 is in an uptrend then we should be making a profit there and the inverse ticker for the S&P500 should be shown in a downtrend then and we should not be investing in that inverse ticker at that time. The other task we will begin working on is to start thinking of is building additional trading algorithms to try and beat the sponsors algorithm as well as trying to optimize the sponsors algorithm. Essentially this project is a research project on how to optimize the sponsors algorithm as well as test out our own additional algorithms to beat the sponsors algorithm.

Print Friendly, PDF & Email

Leave a Reply

Your email address will not be published. Required fields are marked *