This thesis deals with INGARCH models for a count time series. Main emphasis is placed on a linear INARCH model. Its properties are derived. Several methods of estimation are introduced - maximum likelihood method, least squares method and its modifications - and later compared in a simulation study. Main properties and maximum likelihood estimation for INGARCH(1,1) model are stated. Higher order linear INGARCH models and nonlinear INGARCH models are discussed briefly. An application of the presented models on time series of car accidents is given.

Darbe „Akcijų kainų kintamumo analizė“ nagrinėjami ir lyginami Baltijos (Lietuvos, Latvijos, Estijos) bei Lotynų Amerikos (Meksikos, Venesuelos) šalių duomenys. Atliekama pasirinktų akcijų kainų grąžų analizė. Jai naudojami trijų metų kiekvienos dienos duomenys (akcijų kainos). Pirmoje darbo dalyje supažindinama su bendra prognozavimo metodų teorija, aprašomi skirtingi, dažnai literatūroje ir praktikoje sutinkami modeliai. Antrojoje dalyje aprašyti prognozavimo metodai taikomi realiems duomenims, t.y. pasirinktoms akcijoms. Prognozuojama akcijų kainų grąža, kuri po to yra palyginama su realia reikšme, apskaičiuojamos prognozavimo metodų paklaidos. Pagrindinis darbo tikslas – atlikti lyginamąją prognozavimo modelių analizę su pasirinktomis akcijomis ir atrinkti tuos metodus, kurie duoda geriausius rezultatus. Darbo tikslui įgyvendinti naudojama SAS statistinio paketo ekonometrikos ir laiko eilučių analizės posistemė SAS/ETS (Time Series Forecasting System). / Most of empirical surveys in macro and financial economics are based on time series analysis. In this work, data of Baltic and Latin America countries is being analyzed and compared. Analysis of stock price returns is presented using daily long term (three years) period data. In the first part of this work general forecasting theory is presented, also different methods, frequently met in the literature and practice, are described. In the second part, forecasting models are being applied for real data. We present results of forecasting stock returns comparing them with real values. Also a precision of forecasts is being calculated, which let us to decide about propriety of each model. Consequently, the aim of this work is to forecast returns of stock price by various time series models and to choose the best one. The analysis was made using SAS statistical package and its econometrics and Time Series Analysis System (SAS/ETS).