R/Finance 2015 Abstract31 Jan 2015
Follow the Leader - the application of time-lag series analysis to discover leaders in S&P 500.
Keywords: time series correlation, leader discovery
The discovery of lead / lag relationships in the time series data can improve strategy returns, reduce risk exposure, and simplify re-balancing process. The inference process, first examines the rolling window similarity between each pair of time series over various lag lengths to determine the existence and strength of co-movements. The lead / lag relationship is used next to build a directed graph to compute leadership score and extract leaders.
We will use R to implement this methodology and present a case study on stocks in S&P 500 index to validate predictive power and benefits of incorporation of lead / lag relationship discovery in your investment process.
Detecting Leaders from Correlated Time Series D. Wu, Y. Ke, J. Yu, P. Yu, L. Chen
Time-lag Method for Detecting Following and Leadership Behavior of Pedestrians from Mobile Sensing Data, 2013 M. Kjargaard, H. Blunck, M. Wustenberg, Kaj Gronbak, M. Wirz, D. Roggen, G. Troster
What do you think?