The aim of the study is to verify a hypothesis, whether a company’s financial statement items can be modelled using econometric techniques incorporating accounting and macroeconomic variables. For the modelling and forecasting are selected items, necessary to calculate a company’s free cash flow (FCF) of four Lithuanian companies: telecommunication provider TEO LT, cheese manufacturer Rokiškio sūris, producer of household refrigerators Snaigė and distributor and supplier of electric energy VST. From their financial statements are taken such items as operating profit, current assets and current liabilities, long - term assets and long - term liabilities, and modeled as endogenous variables. Two types of exogenous variables are used: accounting variables (revenues and various types of expenditures) and macroeconomic variables (interest rates, disposable income or net earnings, growth of gross domestic product, country’s export, foreign direct investment and inflation). Initial econometric analysis of the variables includes verification of seasonality and stationarity according to the time series graphs and unit - root tests as well as correlation and causality analysis using cross - correlation matrices and Granger causality tests. For the modelling are selected two types of econometric methods: structural simultaneous - equations models (SEM), estimating them using two - stage least squares technique, and vector autoregression (VAR) models. After estimation of the models... [to full text] / Darbo tikslas yra patikrinti hipotezę, ar įmonės finansinės atskaitomybės straipsniai gali būti modeliuojami naudojant ekonometrinius metodus įtraukiant apskaitos ir makroekonominius kintamuosius. Modeliavimui ir prognozavimui yra pasirinkti įmonės laisvam pinigų srautui (angl. free cash flow, FCF) apskaičiuoti reikalingi straipsniai ir keturios Lietuvos įmonės: telekomunikacijų paslaugų teikėja „TEO LT“, sūrių gamybos įmonė „Rokiškio sūris“, buitinių šaldytuvų gamintoja „Snaigė“ bei elektros energijos skirstytoja ir tiekėja VST. Iš šių bendrovių finansinių atskaitomybių yra paimti tokie straipsniai, kaip veiklos pelnas, trumpalaikis turtas ir trumpalaikiai įsipareigojimai, ilgalaikis turtas ir ilgalaikiai įsipareigojimai. Šie rodikliai yra modeliuojami kaip endogeniniai kintamieji. Modeliuojant naudojami egzogeniniai kintamieji yra dviejų tipų: apskaitos kintamieji (pardavimai ir įvairios sąnaudos) bei makroekonominiai kintamieji (palūkanų normos, disponuojamos pajamos, neto darbo užmokestis, bendrojo vidaus produkto augimas, šalies eksportas, tiesioginės užsienio investicijos ir infliacija). Pradinė ekonometrinė kintamųjų analizė apima sezoniškumo ir stacionarumo tikrinimą pagal laiko eilučių grafikus ir vienetinės šaknies testus bei koreliacijų ir priežastingumo analizę, naudojant kryžmines koreliacijas ir Granger priežastingumo testus. Modeliavimui yra pasirinkti du ekonometriniai metodai: struktūrinių vienalaikių lygčių modeliai (angl. structural simultaneous – equation... [toliau žr. visą tekstą]

The accurate (exact) prediction of tax revenue is a very important task for state budget planning. Both underestimation and overestimation of the planned revenue could cause problems in case the revenue is used to finance government functions. In the past few years planning of profit tax revenue was not very exact: the sum of the profit tax collected was considerably smaller or considerably larger than the planned profit tax revenue. The difference between the actual and planned revenue was about 12–56 % every year. There are several related factors which aggravate profit tax revenue modelling. It is doubtful if the indicator of profit tax revenue is stationary. The assumption of the stationarity of indicators is usually made when applying econometric models to the indicators. This problem is caused by a frequent change of the Profit tax law. In addition, transitional processes, invoked by privatization, integration to EU, and etc were typical of the Lithuanian economy in the past few years. Therefore very general equations used to describe the profit tax revenue of macroeconometric models of many countries are not relevant to model and predict the profit tax revenue in Lithuania. In order to predict budget revenues accurately, their modelling methodology needs to be created so as to be effective in the situation where regression relations are complicated while disposable series of observations are rather short (the quarterly profit indicators have been known since 1998 in... [to full text]

The accurate (exact) prediction of tax revenue is a very important task for state budget planning. Both underestimation and overestimation of the planned revenue could cause problems in case the revenue is used to finance government functions. In the past few years planning of profit tax revenue was not very exact: the sum of the profit tax collected was considerably smaller or considerably larger than the planned profit tax revenue. The difference between the actual and planned revenue was about 12–56 % every year. There are several related factors which aggravate profit tax revenue modelling. It is doubtful if the indicator of profit tax revenue is stationary. The assumption of the stationarity of indicators is usually made when applying econometric models to the indicators. This problem is caused by a frequent change of the Profit tax law. In addition, transitional processes, invoked by privatization, integration to EU, and etc were typical of the Lithuanian economy in the past few years. Therefore very general equations used to describe the profit tax revenue of macroeconometric models of many countries are not relevant to model and predict the profit tax revenue in Lithuania. In order to predict budget revenues accurately, their modelling methodology needs to be created so as to be effective in the situation where regression relations are complicated while disposable series of observations are rather short (the quarterly profit indicators have been known since 1998 in... [to full text]