RESUMO
Código: 375
Área Temática: Finanças

 

Abrir
Arquivo

 
Empirical Evidence Of Long-range Correlations And Testing For Nonlinear Dependence In Stock Returns
 
Understanding stock markets price fluctuations is an important role in financial economics policy, corporate investment and financing strategies. The more efficient is the market, the more random is the sequence of price changes generated by the market. Therefore, the existence of long memory in stock market returns would affect the investment horizon of portfolio decisions. The principal purpose of this study is to research about long-range dependence in Latin American stock markets. Therefore, will be applied the BDS and K2K tests to verify the presence of some type of linear or nonlinear dependence, Hurst exponent and Detrended Fluctuation Analysis to verify the presence or not of the long-range dependence. This study provides a detailed examination of long-range dependence in weekly returns for the seven markets of Latin America from 1994 to 2005. The results provided from the BDS and K2K tests suggest the existence of nonlinear dependence in the returns series in all the analyzed countries. Also was evidenced random walk behavior for two markets and for the five other markets, the value the DFA test is smaller, which helps on the affirmative that the participation of long-range dependence in the formation of the prices weekly is minimum.