Tracing Causality and Co-movement between Pakistani and the Leading Foreign Stock Markets: A Graph Theoretic Approach

Abstract

This paper developed and modify Peter and Clark (PC) causality algorithm to revisit the causal linkages between Pakistan and the leading foreign stock markets. Initially, the PC algorithm was conceived to determine causality in cross sectional data. Later on, (Swanson & Granger, 1997) for the first time used VAR residuals in PC algorithm to determine the causal ordering in time series. However, the weak point attached to VAR residuals are that it carries only contemporaneous causal information and remove all the past information. This study modify the PC algorithm based on recursive residuals proposed by (Rehman & Malik, 2014) and explore the causality among exchange rate, interest rate and stock market prices. The overall empirical results of modified PC algorithm indicate that causality is running from exchange rate, interest rate and stock market of India and Bangladesh to Pakistani stock market. The results observed from GARCH-GJR model show spill over effect from the leading foreign stock markets toward Pakistan stock market excluding Sir Lanka. The results of the study will guide the investors to be vigilant in decision making in diversified portfolio investment and hedging. Keywords: Financial Markets, PC algorithms, Causality, Graph theoretic Approach, GARCH, GJR.

Effectiveness of monetary policy: Application of modified Peter and Clark (PC) algorithms under Graph-Theoretic Approach

“…To explain the term unshielded collider, we have Figure no. 2 which does not show any direct relationship between node A and B which mean that these two nodes are unconditionally uncorrelated, but conditionally correlated on node D (Fazal et al, 2020and Hoover, 2005, 2020.…”