Advances in Risk Management

232 CORRELATION BREAKDOWNS IN ASSET MANAGEMENT

Table 12.3 Regression equation of correlation EUR–JPY changes explained
by volatility differences

Variable Coefficient Std. error t-statistic Prob.

DEURV 3.136068 0.655131 4.786932 0.0000
DJPYV 3.246925 0.580543 5.592912 0.0000

R-squared 0.056743 Mean dependent var 0.221726

AdjustedR-squared 0.055047 S.D. dependent var 0.588052

S.E. of regression 0.571638 Akaike info criterion 1.722956
Sum squared residuals 181.6841 Schwarz criterion 1.738456

Log likelihood −478.7048 F-statistic 33.44724
Durbin–Watson stat. 1.932469 Prob(F-statistic) 0.000000

Notes: The dependent variable; differences of exponential correlations between the equity euro mar-
ket and the equity Japanese market. Explanatory variables are DEURV: differences of exponential
volatility of the equity euro market; and DJPYV: differences of exponential volatility of the equity
Japanese market. Number of observations=558 daily changes.

10

8

6

4

2

0

100 200 300 400 500

 2

RESID

Spread between actual and estimated

dependent variable

Number of observations  558 daily changes

No. of observations

Figure 12.6 Plot of residuals of correlation EUR–JPY changes explained by

volatility differences

value of the sum of squared residuals which is roughly 30 percent lower
(from 43.136 to 30.321).R-squared is 95 percent higher in the shorter time
series (0.099577 instead of 0.004663). The complete time series has been
estimated twice, the first time using the two changes in volatilities. The
volatility of the Japanese market is statistically non-significant (student