TIME VARYING COINTEGRATION AND CAUSALITY: A COMPARATIVE CASE STUDY FOR RUSSIA, UKRAINE, HUNGARY AND ROMANIA USING VECM AND CCC – MGARCH MODELS
TIME VARYING COINTEGRATION AND CAUSALITY: A COMPARATIVE CASE STUDY FOR RUSSIA, UKRAINE, HUNGARY AND ROMANIA USING VECM AND CCC – MGARCH MODELS

Summary/Abstract: THIS PAPER AIMED TO INVESTIGATE COINTEGRATION AND CAUSALITY EFFECT WITH BASE TO RUSSIAN FINANCIAL MARKET OVER UKRAINE, HUNGARY AND ROMANIA, (2) WHETHER IS THERE ANY OPPORTUNITY FOR HEDGE THE INVESTMENT IN ONE MARKET BY INVESTING IN ANOTHER MARKET. FOR THIS PURPOSE WE EMPLOY VECTOR ERROR CORRELATION MODEL (VECM) AND CCC GARCH ON THE SAMPLE DATA FROM JANUARY 2013 TO AUGUST 2022 COLLECTED FROM BLOOMBERG. AIC CRITERIA HAS BEEN USED TO DETERMINE THE IDEAL LAG DURATION, WHICH IS DETERMINED TO BE 8 DAYS FOR PFTS - UKRAINE AND MOEX - RUSSIA, 9 DAYS FOR BET - ROMANIA AND MOEX - RUSSIA, AND 8 DAYS FOR BUX - HUNGARY AND MOEX - RUSSIA. THIS IS BECAUSE COINTEGRATION RESULTS ARE SENSITIVE TO LAG LENGTH. CAUSALITY MUST EXIST IF TWO PARAMETERS (EXTRINSIC AND INTRINSIC) COINTEGRATE IN AT LEAST ONE DIRECTION (GRANGER, 1986). ADDITIONALLY, IT WAS ASSERTED BY GHOSH (1993), LIEN AND LUO (1994), AND LIEN (1996) THAT IF THE TWO PRICE SERIES ARE DISCOVERED TO BE COINTEGRATED, THEN THERE ARE VIABLE ERROR CORRECTION REPRESENTATIONS OF THE PRICE SERIES THAT CONTAIN BOTH SHORT-TERM DYNAMICS AND LONG-TERM INFORMATION. THE STUDY USES THE VECTOR ERROR CORRECTION MODEL (VECM) BECAUSE SPOT AND FUTURES PRICES ARE COINTEGRATED IN ORDER TO EXAMINE THE LEAD-LAG RELATIONSHIP, THE LONG- AND SHORT-RUN SPEED ALTERATION TOWARDS EQUILIBRIUM, OR THE LONG-RUN STEADY STATE (CAUSALITY), BETWEEN PFTS INDEX (UKRAINE) AND MOEX INDEX – (RUSSIA) STOCK MARKET PRICES OF SAMPLE VARIABLES.

• Details
• Contents