Correlation Analysis of Financial Contagion: What One Should Know Before Running a Test

Table 1: Hong Kong crisis - Conditional and Fisher tests

country	P	∕'	λ	Fisher -⅛ — 1 _____________er_________
Indonesia	0.31	0.60		*	9.7
Korea	0.16	0.07	+∞	_	38.7
Malaysia	0.20	0.43	64.5	_	24.8
Philippines	0.22	0.66	2.6	**	19.8
Singapore	0.36	0.76	1.5	**	6.5
Thailand	0.11	0.01	+∞	_	88.6
Argentina	0.26	0.21	+∞	_	13.5
Brazil	0.20	0.31	30,941.3	_	23.1
Mexico	0.29	0.45	49.3	_	10.8
Russia	0.19	0.53	13.8	*	26.9
“USA	0.15	0.26	254.0	_	42.1
Japan	0.28	0.33	7486.5	_	11.4
Germany	0.24	0.63	4.4	**	16.8
France	0.17	0.66	1.2	**	32.3
United Kingdom	0.17	0.63	2.3	**	33.0
Italy	0.00	0.63	0.00	**	732,762
Canada	0.27	0.37	389.8	_	12.8

Note: p and p are estimated correlation coefficients of two-day rolling
averages of returns in the tranquil and crisis periods; λ is the threshold
variance ratio as defined in the text (for S = 8.72). The fourth column
reports the results of the Fisher test: * (**) indicates that the hypothesis
P'P' ≤ P is rejected at the 5 (1) per cent significance level.

Table 2: Estimations of the variance ratio for Hong Kong

У = У		У	У
Cross section:
G 7	2.8	2.9	3.2
Full sample	2.4	2.6	2.6
World stock market index	3.6	3.0	4.5
Principal components:
First component	7.1
First two components	7.0
First five components	4.1

More intriguing information

1. The Effects of Reforming the Chinese Dual-Track Price System
2. The name is absent
3. The name is absent
4. The Variable-Rate Decision for Multiple Inputs with Multiple Management Zones
5. Effects of red light and loud noise on the rate at which monkeys sample the sensory environment
6. The name is absent
7. The name is absent
8. An alternative way to model merit good arguments
9. Education and Development: The Issues and the Evidence
10. Keystone sector methodology:network analysis comparative study