|
on Econometric Time Series |
| By: | Gilles Dufrenot (GREQAM - Groupement de Recherche en Économie Quantitative d'Aix-Marseille - Université de la Méditerranée - Aix-Marseille II - Université Paul Cézanne - Aix-Marseille III - Ecole des Hautes Etudes en Sciences Sociales - CNRS : UMR6579); Dominique Guegan (CES - Centre d'économie de la Sorbonne - CNRS : UMR8174 - Université Panthéon-Sorbonne - Paris I); Anne Peguin-Feissolle (GREQAM - Groupement de Recherche en Économie Quantitative d'Aix-Marseille - Université de la Méditerranée - Aix-Marseille II - Université Paul Cézanne - Aix-Marseille III - Ecole des Hautes Etudes en Sciences Sociales - CNRS : UMR6579) |
| Abstract: | This paper presents a 2-regime SETAR model with different long-memory processes in both regimes. We briefly present the memory properties of this model and propose an estimation method. Such a process is applied to the absolute and squared returns of five stock indices. A comparison with simple FARIMA models is made using some forecastibility criteria. Our empirical results suggest that our model offers an interesting alternative competing framework to describe the persistent dynamics in modeling the returns. |
| Keywords: | SETAR - Long-memory - Stock indices - Forecasting |
| Date: | 2008–03–06 |
| URL: | http://d.repec.org/n?u=RePEc:hal:papers:halshs-00185369_v1&r=ets |
| By: | Alexander Subbotin (CES - Centre d'économie de la Sorbonne - CNRS : UMR8174 - Université Panthéon-Sorbonne - Paris I, Higher School of Economics - State University) |
| Abstract: | We decompose volatility of a stock market index both in time and scale using wavelet filters and design a probabilistic indicator for valatilities, analogous to the Richter scale in geophysics. The peak-over-threshold method is used to fit the generalized Pareto probability distribution for the extreme values in the realized variances of wavelet coefficients. The indicator is computed for the daily Dow Jones Industrial Averages index data from 1986 to 2007 and for the intraday CAC 40 data from 1995 to 2006. The results are used for comparison and structural multi-resolution analysis of extreme events on the stock market and for the detection of financial crises. |
| Keywords: | Stock market, volatility, wavelets, multi-resolution analysis, financial crisis. |
| Date: | 2008–03 |
| URL: | http://d.repec.org/n?u=RePEc:hal:papers:halshs-00261514_v1&r=ets |
| By: | Neil Shephard; Torben G. Andersen |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:sbs:wpsefe:2008fe23&r=ets |
| By: | Massimiliano Caporin (Università di Padova); Michael McAleer (University of Western Australia) |
| Abstract: | DAMGARCH extends the VARMA-GARCH model of Ling and McAleer (2003) by introducing multiple thresholds and time-dependent structure in the asymmetry of the conditional variances. DAMGARCH models the shocks affecting the conditional variances on the basis of an underlying multivariate distribution. It is possible to model explicitly asset-specific shocks and common innovations by partitioning the multivariate density support. This paper presents the model structure, describes the implementation issues, and provides the conditions for the existence of a unique stationary solution, and for consistency and asymptotic normality of the quasi-maximum likelihood estimators. The paper also provides analytical expressions for the news impact surface implied by DAMGARCH and an empirical example. |
| Keywords: | multivariate asymmetry, conditional variance, stationarity conditions, asymptotic theory, multivariate news impact curve |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:pad:wpaper:0064&r=ets |
| By: | Turk, Mehmet; Ozun, Alper |
| Abstract: | Volatility in financial markets should be correctly estimated for an efficient risk management. In emerging markets, due to relatively low trade volume, economic and political instability, and regulatory changes, higher volatility persists in financial asset prices as compared to those in advanced markets. In highly volatile markets, unexpected shifts in financial asset prices can be predicted by using flexible models enabling data filtering. In this research article, we use logarithmic normal stochastic volatility with Kalman filter and two regime switching stochastic volatility with Hamilton filter to estimate volatilities of exchange rates. In a comparative way, we examine the success of the two models in volatility estimation using time series from the Turkish markets. By employing daily USD/TRY exchange rates from 01/01/2004 to 25/07/2007, we empirically examine if the models are successful in predicting exchange rates in short-term and long-term. The article has originality in being first research article, as much as the authors know, which examines stochastic volatility models in a comparative perspective using data from Turkish exchange rate markets. |
| Keywords: | Regime Switching models; stochastic volatility; Hamilton filters; Kalman filters; exchange rate; Turkish lira |
| JEL: | G14 C14 F31 |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:7670&r=ets |
| By: | Turk, Mehmet; Ozun, Alper |
| Abstract: | High and sudden volatility in the financial markets might cause unexpected losses. Increasing volatility in the prices of financial securities follows regime shifts in the markets. In general, there exist two regimes in the financial markets, namely, “stable” and “volatile” regimes. Therefore, estimating regime shifts in financial time series is crucial for the efficient risk management. From that perspective, the regime switching probabilities in emerging stock markets are examined with one of the regime switching models called Two State MSH(2)-AR, Autoregressive Markov Switching Heteroscedasticity Model. In the empirical analysis, we use daily time series data between 09/01/2004 and 13/09/2007 from i) emerging markets including Turkey, Russia, Ukraine, Brazil and Lebanon; ii) an advanced market, namely Dow Jones Industrial Average, iii) a world stock index, MSCI (Morgan Stanley Composite Index). Using data from different markets gives us to chance of evaluating the model’s performance with different time series. In addition, finding different regimes in the indexes within the same time period means that the investor have chance to diversify their portfolios. |
| Keywords: | Emerging markets; Regime switches; Markov chains; Volatility ; stock exchanges |
| JEL: | E32 G15 F21 F36 |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:7673&r=ets |
| By: | Adrian R. Pagan (School of Economics, The University of New South Wales); M. Hashem Pesaran (Faculty of Economics, University of Cambridge) |
| Abstract: | This paper considers the implications of the permanent/transitory decomposition of shocks for identification of structural models in the general case where the model might contain more than one permanent structural shock. It provides a simple and intuitive generalization of the influential work of Blanchard and Quah (1989), and shows that structural equations with known permanent shocks can not contain error correction terms, thereby freeing up the latter to be used as instruments in estimating their parameters. The approach is illustrated by a re-examination of the identification schemes used by Wickens and Motto (2001), Shapiro and Watson (1988), King, Plosser, Stock, Watson (1991), Gali (1992, 1999) and Fisher (2006). |
| Keywords: | Permanent shocks; structural identification; error correction models; IS-LM models |
| JEL: | C30 C32 E10 |
| Date: | 2008–03 |
| URL: | http://d.repec.org/n?u=RePEc:swe:wpaper:2008-04&r=ets |