• ISSN: 2010-023X
    • Frequency: Bimonthly
    • DOI: 10.18178/IJTEF
    • Editor-in-Chief: Prof.Tung-Zong (Donald) Chang
    • Executive Editor: Ms. Cherry L. Chen
    • Abstracting/ Indexing:  ProQuest, Crossref, Electronic Journals Library , EBSCO, and Ulrich's Periodicals Directory
    • E-mail: ijtef@ejournal.net
IJTEF 2018 Vol.9(4): 182-185 ISSN: 2010-023X
DOI: 10.18178/ijtef.2018.9.4.612

A Machine Learning Approach to Spinoff Investment Optimization

Mehdi Chouiten and Romain Ekert
Abstract—A spinoff is an event that consists of the creation of a new company based on an existing division of a mother company. Often, this event is due to a new strategic vision of the mother company or to unsatisfying financial performance of the spun off division. In our study, we try to capitalize on several features to measure their impact on spinoffs success or failure and thus, build a predictive model that allows us to select best spinoffs to invest in. Our method aims to predict the stock price performance over different time horizons: 6, 12,18, and 24 months. Allowing profitable exits to investors (either stock traders or option traders). Using mainly Bloomberg platform data, we compared several machine learning algorithms (SVM, Random Forest, Gradient Boosting) and different methodologic approaches (Binary classification, time-series classification, multi-class clustering) to build an efficient, yet improvable, model to reach our goal.

Index Terms—Spinoff, investment, machine learning, modeling, option trading, time-series classification, stock prediction.

Mehdi Chouiten is with Datategy SAS, France (e-mail: mehdi.chouiten@datategy.net).


Cite: Mehdi Chouiten and Romain Ekert, "A Machine Learning Approach to Spinoff Investment Optimization," International Journal of Trade, Economics and Finance vol.9, no.4, pp. 182-185, 2018.

Copyright © 2008-2018. International Journal of Trade, Economics and Finance. All rights reserved.
E-mail: ijtef@ejournal.net