Modelling and forecasting the global financial crisis: Initial findings using heterosckedastic log-periodic modelsстатья
Информация о цитировании статьи получена из
Scopus
Статья опубликована в журнале из списка Web of Science и/или Scopus
Дата последнего поиска статьи во внешних источниках: 28 мая 2015 г.
Аннотация:The financial crisis of 2007-2009 has begun in July 2007 when a loss of confidence by investors in the value of securitized mortgages in the United States resulted in a liquidity crisis. World stock markets peaked in October 2007 and then entered a period of high volatility which culminated with the market crashes in September and October 2008. Since March 2009, the world stock markets have rebounded, but strong uncertainties still remain. In order to get more insights into the current world markets operation, we consider log-periodic models of price movements, which has been largely used in the past to forecast financial crashes and "anti-bubbles". Both the original and an extended model which accounts for heteroskedasticity and autocorrelation are fitted to the American S&P500 index. The empirical analysis reveal three interesting points: i) the log-periodic models outperform standard financial models when long-term out-of-sample forecasting is of concern. ii) the log-periodic-AR(1)- GARCH(1,1) model has residuals with better statistical properties than the original model and iii) the current market rebound should peak at the beginning of 2010.