Related Experiment Videos
Time-dependent cross-correlations between different stock returns: a directed network of influence.
L Kullmann1, J Kertész, K Kaski
1Department of Theoretical Physics, Budapest University of Technology and Economics, Budafoki út 8, H-1111, Budapest, Hungary.
Physical Review. E, Statistical, Nonlinear, and Soft Matter Physics
|September 21, 2002
Summary
We analyzed stock return correlations over time using tick-by-tick data. Our findings reveal a directed influence between companies, consistent with market efficiency.
Area of Science:
- Quantitative Finance
- Network Science
- Market Microstructure
Background:
- Understanding inter-company relationships is crucial for financial market analysis.
- Previous studies often overlook the temporal dynamics of stock return correlations.
Purpose of the Study:
- To investigate time-dependent cross-correlations in stock returns.
- To identify directional influence between companies using high-frequency trading data.
- To assess the implications of these findings for market efficiency.
Main Methods:
- Analysis of tick-by-tick stock return data.
- Calculation of time-dependent cross-correlations with varying time shifts.
- Network construction to represent inter-company influence.
Main Results:
- A weak but statistically significant effect of maximum correlation at non-zero time shifts was observed.
- Characteristic time scales for these correlations are on the order of minutes.
- Evidence suggests a directed network of influence between companies.
Conclusions:
- The observed directional influence is subtle and short-lived.
- Findings support the hypothesis of an efficient market, where such effects are weak.
- Company interactions can be modeled as a directed influence network.