![IconGems-Print](images/IconGems-Print.png)
670 – Statistical Computing in Data Science
Trends and Volatility of Stock Prices Following Occurrence of Specific Technical Patterns
James Shine
U.S. Army Topographic Engineering Center
James Gentle
George Mason University
Charles Perry
USDA
A well-known heuristic in financial time series is that a "head and shoulders" pattern (a small peak, a larger peak, and a second small peak) after a long upward trend is a predictor of a future downward trend. We look to assess this and other heuristics against actual financial time series data. We define specific technical patterns and then detect these patterns on real stock data over multiple stocks, time intervals and step sizes. We then compare trend and volatility measures of the stock after the pattern occurs. Our results and their comparison with other studies will be presented and discussed.