Contact Information
Information Systems Department
The Hagan School of Business
Iona College
715 North Avenue
New Rochelle, NY 10801
rob.schumaker@gmail.com
Phone: (914) 633-2065

---

Director of the Information Systems Dept. Computer Lab (Hagan 306)

---

Founder & President,
The Schumaker Foundation

The Arizona Financial Text System
The Arizona Financial Text System (AZFinText) is a stock selection research project that utilizes the terms in financial news articles to predict future stock prices. Using the premise that certain terms can move stocks more than others, the goal of the AZFinText system is to leverage arbitrage opportunities that exist when investment experts over and under-react to certain news stories. By analyzing breaking financial news articles and focusing on specific parts of speech, portfolio selection and term weighting, the AZFinText system is a radically different way of looking at stock market prediction. In a comparison study against the top quantitative trading systems, AZFinText outperformed the majority.

AZFinText Papers
In Textual Analysis of Stock Market Prediction Using Financial News Articles, we laid the foundation of AZFinText by exploring several different models and linguistic representations. From this study, AZFinText was found to exhibit a 2.84% trading return during the five week study period.

The second paper, A Quantitative Stock Prediction System based on Financial News, the focus was on what articles work best at price prediction and how well the system performs against existing human trading experts and quantitative systems. Using the theory that news articles from peer industries may be more effective than articles in disparate sectors, AZFinText found that training based on sectors worked best by achieving an 8.50% trading return. This return also beat the human experts and 6 of the top 10 quantitative funds during the study period.

As a followup, Evaluating a News-Aware Quantitative Trader: The Effects of Momentum and Contrarian Stock Selection Strategies, the AZFinText system was further refined as a quantitative trader to take into account portfolio formation based on Momentum and Contrarian stock selection strategies. Using the idea that past winning stocks will continue to win and past losers will lose, AZFinText investigated the predictive value of these strategies and achieved a 20.79% trading return during the study period. This study was also notable because it uncovered evidence of trader over-reaction to news events which led to these abnormal returns.

In a different fork of research, we examined the role of author sentiment as a predictive tool. In Sentiment Analysis of Financial News Articles, we used the idea that authors of financial news articles can unwittingly shape stock price movement simply by the tone and polarity of thier writing. From this study, we found that sentiment is a valuable predictive element. Furthermore, AZFinText noted that Contrarian trading activity was occuring during the study period. Articles that were judged to have a positive writing style generally decreased in price while articles with a negative writing style increased in the short term.

A similar study went a step farther and looked into the specific verbs that can move prices the most. In An Analysis of Verbs in Financial News Articles and their Impact on Stock Price, AZFinText found that the five verbs with the highest positive impact on stock prices are planted, announcing, front, smaller and crude.

Notable Publicity for AZFinText
AZFinText has been featured in numerous media outlets in many different languages. Some of the more notable ones are the Wall Street Journal, Slashdot, MIT's Technology Review, Motley Fool, Crossing Wall Street, WBIR in Knoxville, TN and Motherboard TV.