Map

ABOUT THE TALK

The basic strategy is termed Swing Trading (holding stocks a few weeks) as well as long term (hold stocks over a year for dividends) trading. The focus is on momentum, breakout and crossover algorithms that can provide a 5% profit in a few days. Stocks are found by using 10+ custom trading algorithms that scan the previous day’s results for interesting setups. This talk will be an update of a previous talk (November 18, 2015) and includes many new trading algorithms and results. Most of the algorithms are gleaned from standard trading books as well as talking to experts in the field. This strategy is actively used and the algorithms are constantly being revised.

One of the big problems with algorithms is how to set the parameters. In the past, these were set through much trial and error in a very time consuming process. One of the new advances this time around is the addition of Big Data’s Genetic Algorithms for automatic tuning. With automatic tuning, over fitting becomes an issue and needs to be addressed.

The thinking on when to sell a stock has also been updated and verified via testing. The key idea here is capture all the up movement of each trade. So, depending on how involved someone is in their trade, various techniques yield differing amount of profit.

ABOUT THE SPEAKER

Peter R. Paradis is a retired Engineer who has specialized in Computer Science his whole career. Peter has always been interested in building a computer system that just prints money made from automatically trading in the stock market. He has been working on this program for the last 20 years and with his retirement last year, he has pretty much been working on it full time now. His last job was in the Hedge Fund space where he learned about Machine Learning and how to apply that to picking stocks.

For a copy of the paper, “Printing Money II”, by Peter Paradis that describes all the trading algorithms and research, click on Paradis_Printing_Money_II

For a copy of the presentation click on Presentation