The analytic form of Fisher Information Matrix (IM) written for DCC-MGARCH(1,1) was suggested. After that it was applied for simplifying the general algorithm: the statistical hypothesis about constant correlation matrix usage was put forward and statistical verification was made. IM was employed for Russian share market: to do investigations the five equilibrium portfolios was compounded from four different shares in each case. Computations made showed that there are three types T1-T3 of trading days on the market and day type changing from T1 to T2 and vice versa is happening over the time moments T3. Moreover, the clusterization effect of multivariate volatility, that investigated by scientists from all around the world in the univariate case, was discovered and described.

Найдена аналитическая форма записи информационной матрицы Фишера для эконометрического алгоритма DCC-MGARCH(1,1). Выдвинута статистическая гипотеза о постоянстве матриц корреляций во времени, и проводится ее статистическая проверка. Информационная матрица применяется для эконометрического исследования фондового рынка России. Обнаружен эффект кластеризации обобщенной волатильности, подтвержденный в одномерном случае другими исследователями.