Berlin 2005 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

AKSOE: Physik sozio-ökonomischer Systeme

AKSOE 14: Financial Markets and Risk Management II

AKSOE 14.4: Vortrag

Mittwoch, 9. März 2005, 15:30–16:00, TU P-N203

Stock markets cross-levels correlations: Short-range stock return linear regression models for developing stock markets — •Michael Romanovsky, Vladislav Kharlamov, Andrew Shurup, and Igor Zhukov — General physics Institute, Vavilov str., 38, 119991 Moscow Russia

An account of levels of international stock markets is convenient for analysis of one-level and cross-level correlations of securities on stock markets since one may expect a reaction of stock rates on cumulative indexes changes, especially in developing economics. Cross-level correlations between stocks of developing markets (beginning Russian and Brazil markets) and indices of all highest levels are investigated. Linear multi-regression models are generated. It includes basic model variables: various MSCI indexes and several commodity price returns, and the function: the investigated stock price return. The result was done for the Russian Trade System index RTSI (Russian electronic stock exchange) using 2 variables of MSCI indexes WRLD/ENERGY and WRLD/INFORMATION TECH as well as 2 variables of commodities OIL and ALUMINIUM. Using the base of correlation matrices traced for all 2001, the value R2adj for the first 10 days of 2002 is 0.248. The same analysis was conducted for Brazil national stock index BOVESPA. There were used two variables (MSCI indexes) WRLD/CONSUMER STAPLES and WRLD/INFORMATION TECH only. The prediction coefficient was smaller than for RTSI: R2adj = 0,168. The proposed method could be used for short-range forecasts of developing stock markets dynamics.