PROJECT: Development of a fully automated trading strategy
· By the end of the semester, complete the development and testing of a fully automated trading strategy using the Ninja Trader platform and submit a detailed report of your findings.
· As part of class participation, students will be expected to share/present some of their work to the class. The presentations will begin before the report is due, maybe as early as the middle of the semester, focusing on your interim (early) findings.
· Your grade will not depend on producing a highly successful trading model (algorithm). Your grade will be based on the depth and quality of analysis of your final written report. (Caution: You will not get an “A” in your report unless you have supported your work by extensively referencing the assigned course readings/books).
Student Final Project: Report Requirements
· Each student is required to develop an algorithmic trading strategy and prepare a report on the key steps in the process.
o Your trading strategy should be supported by extensively referencing the two assigned books in this course plus other outside sources when relevant. The assigned books are:
§ Building Winning Algorithmic Trading Systems by Kevin Davey
§ High Probability Trading: Take the Steps to Become a Successful Trader by Marcel Link
o The report should be single spaced and vary in length between 10 and 15 pages. More pages may be added if needed, but try to be as succinct as possible. Appendices may be used and referenced in the main body of your report to illustrate or provide additional data tables or charts.
· The following milestones are involved in the development of a trading strategy, and should be included in the final report:
a. Establish financial objectives
· What are the desired profits, risks, and time frame?
· Specify how you will measure your objectives.
b. Define the market and select a portfolio of stocks (five or greater), describing the:
· Market Sector (like healthcare, technology, industrial goods, etc.)
· Industry (like automobiles, drugs, restaurants, etc.)
· Firm characteristics (like market capitalization, etc.)
· The following screening program (financial visualizations) may help you do this:
c. Develop a trading idea, specifying your entry and exit conditions and actions.
· Provide a reasonable rationale for your initial trading strategy.
· Explain any technical indicators that you use, including the formulas and precise entry and exit conditions.
Provide research support (referencing sources) on the
effectiveness of the indicators that you have selected.
d. Perform limited testing, and revise your trading idea as needed
· Initially, test your model using data for 1 or 2 years.
· Explain the results of the limited testing you have done and how you have revised the initial algorithmic trading model before you found it acceptable enough to warrant more in-depth testing. (Purpose is to share, briefly, the initial work you have done to develop a reasonable model that looked successful enough to do more extensive testing. Here you might summarize the types of models you attempted to use that were not successful.)
e. Conduct in-depth testing and explain the results
· Expand the depth and breadth of your analysis of the initial model by looking at (testing) additional years, stocks, parameter values of inputs and variables; and a more extensive study of the performance measures used to evaluate the success of your trading algorithm.
· Include the following:
· It is important to reserve a couple of years that you do not use/include in your testing until the very end, after you have completed your final model.
f. Describe, in detail, your “final” strategy that you have found to be most acceptable.
· Copy and include images of the “Sets” (showing conditions and action commands). In each of your “sets”, make sure to explain the conditions and actions.
· Export your strategy (designed within the strategy builder) and include this file when you post your final project report to the drop-box in myCourses.
g. Do a final confirmation test with data previously not used, but do not change your strategy after seeing the results of this test.
· Once you are satisfied with your model, do a last test using one or two years of data that you have not used in your prior analysis. You may even test different stocks that you have not used to develop your trading model.
· Do not revise your strategy at this point, even if the results are not good. If the results of this last test are not good, you should just discuss the possible reasons for this in your conclusion, and make suggestions for further research.
h. Conclude by discussing the strengths and pitfalls of the algorithmic trading strategy.
· Summarize what you have done and what you have found.
· Is the algorithmic trading strategy one that you recommend? Why or why not? Discuss it strengths and weaknesses.
· Make suggestions for the future enhancement/development of this model.
i. Post your report to myCourses drop-box.
· Also, export your strategy (designed within the strategy builder) and include this file when you post your final project report to the drop-box in myCourses.