Inference Robustness of ARIMA Models under Non-normality - Special Application to Stock Price Data

609 12 629 2 1237 document 629 RM-76-075.pdf application/pdf ddbbbc2146c2deec3e9cc3fbaf2ed04d MD5 480103 2016-01-15 01:43:36 https://pure.iiasa.ac.at/id/eprint/609/1/RM-76-075.pdf 609 1 1 text en public

RM-76-075.pdf

20 29055 1 40787 document 29055 indexcodes.txt text/x-c++ 5fe5311e4ea727f114f8d55c16d0b912 MD5 4821 2016-01-15 22:31:37 https://pure.iiasa.ac.at/id/eprint/609/2/indexcodes.txt 609 2 2 text/x-c++ other Generate index codes conversion from text to indexcodes en public

indexcodes.txt

lightbox.jpg

preview.jpg

medium.jpg

small.jpg

RM-76-075.pdf

http://eprints.org/relation/isVersionOf https://pure.iiasa.ac.at/id/document/629 http://eprints.org/relation/isVolatileVersionOf https://pure.iiasa.ac.at/id/document/629 http://eprints.org/relation/isCoversheetVersionOf https://pure.iiasa.ac.at/id/document/629 archive 351 disk0/00/00/06/09 2016-01-15 01:43:36 2021-08-27 17:08:24 2016-01-15 01:43:36 monograph show 2 Ledolter J. AL0225 Inference Robustness of ARIMA Models under Non-normality - Special Application to Stock Price Data pub iis_met iis_mod prog_sds Wold's decomposition theorem states that every weakly stationary stochastic process can be decomposed into orthogonal shocks. For practical reasons, however, it is desirable to employ models which use parameters parsimoniously. Box and Jenkins show how parsimony can be achieved by representing the linear process in terms of a small number of autoregressive and moving average terns (ARIMA-models). The Gaussian hypothesis assumes that the shocks follow a normal distribution with fixed mean and variance. In this case the process is characterized by first and second order moments. The normality assumption seems to be reasonable for many kinds of series. However, it was pointed out by Kendall, Mandelbrot, and Fama that particularly for stock price data the distribution of the shocks appears leptokurtic. In this paper we investigate the sensitivity of ARIMA models to non-normality of the distribution of the shocks. We suppose that the distribution function of the shocks is a member of the symmetric exponential power family, which includes the normal as well as leptokurtic and platikurtic distributions. A Bayesian approach is adopted and the inference robustness of ARIMA models with respect to (i) the estimation of parameters, and (ii) the forecasts of future observations is discussed. 1976-10 published RM-76-075 RM-76-075 10 2103 public research_memorandum IIASA, Laxenburg, Austria 23 FALSE info:eu-repo/semantics/book