In the last few decades, vector autoregressive (VAR) models have gained tremendous popularity as an all-purpose tool in econometrics and other disciplines. Some of their most prominent uses are for forecasting, causality tests, tests of economic theories, hypothesis-seeking, data characterisation, innovation accounting, policy analysis, and cointegration analysis. Their popularity appears to be attributable to their flexibility relative to other models rather than to their virtues per se. In addition, analysts often use VAR models as benchmark models. VAR modeling has not gone uncriticised, though. A list of relevant arguments against VAR modelling can be found in Section 2.3 of this thesis. There is one additional problem which is rarely mentioned though, namely the often heavily biased estimates in VAR models. Although methods to reduce this bias have been available for quite some time, it has probably not been done before, at least not in any systematic way. The present thesis attempts to systematically examine the performance of bias-reduced VAR estimates, using two existing and one newly derived approximation to the bias. The thesis is orginanised as follows. After a short introductory chapter, a brief history of VAR modelling can be found in Chapter 2 together with a review of different representations and a compilation of criticisms against VAR models. Chapter 3 reports the results of very extensive Monte Carlo experiments serving dual purposes: Firstly, the simulations will reveal whether or not bias really poses a serious problem, because if it turns out that biases appear only by exception or are mainly insignificant, there would be little need to reduce the bias. Secondly, the same data as in Chapter 3 will be used in Chapter 4 to evaluate the bias approximations, allowing for direct comparison between bias-reduced and original estimates. Though Monte Carlo methods have been (rightfully) criticised for being too specific to allow for any generalisation, there seems to be no good alternative to analyse small-sample properties of complicated estimators such as these. Chapter 4 is in a sense the core of the thesis, containing evaluations of three bias approximations. The performance of the bias approximations is evaluated chiefly using single regression equations and 3D surfaces. The only truly new research result in this thesis can also be found in Chapter 4; a second-order approximation to the bias of the parameter matrix in a VAR(p) model. Its performance is compared with the performance of two existing first-order approximations, and all three are used to construct bias-reduced estimators, which are then evaluated. Chapter 5 holds an application of US money supply and inflation in order to find out whether the results in Chapter 4 can have any real impacts. Unfortunately though, bias reduction appears not to make any difference in this particular case. Chapter 6 concludes. / Diss. Stockholm : Handelshögsk.

The purpose of this thesis is to identify a general model to forecast GDP growth for the Baltic States, Estonia, Latvia and Lithuania. If the model provides reliable results for these states, then the model should be able to forecast GDP growth for other countries of interest. Forecasts are made by using a reduced vector autoregressive (VAR) model. The VAR models make use of past values of Gross Domestic Product-Inflation-Unemployment as explanatory variables. The performed forecasts have provided good results for horizons up to t+8. The forecasts for 2009 (t+12) are in line with those of several other actors. It is reasonable to assume that some of the forecasts for t+16 have reliable results. The Lithuanian forecast show a fall in GDP with 12.51 per cent in 2009 and a GDP growth of 4.23 per cent in 2010. The forecast for Estonia show that the GDP will decrease with 1.49 per cent in 2009 and 12.72 per cent in 2010. Finally the forecast for Latvia show a fall in GDP of 3.1 per cent in 2009 and 18 per cent in 2010. From the findings it is possible to conclude that the model provided reliable estimates of future levels of GDP for the Baltic States and the benchmark countries. This indicates that the model should be applicable on other countries of interest.

In this study, time series analysis and factor analysis of multivariate analysis were applied to surface water quality monitoring data sets. In order to gain a better understanding on water pollution situations and sources of different river basin, we have done some statistical analysis according to the water quality monitoring data from the database of Environmental Protection Administration Executive Yuan in the years from 2004 to 2008. The statistical analysis is based on the aforementioned contents to find the common latent factors of the water pollution and try to explain the main sources of water pollution of river basins. Specifically the water quality monitoring data of the Tamsui River and the Kao-Ping River basin are analyzed. In seriously polluted areas, from the time series results they reveal that the water quality monitoring data has significant serial correlations. Therefore we firstly use vector autoregression (VAR) model to eliminate the dynamic dependencies of data, and later perform the factor analysis for the residuals. With this analysis procedure, the resulting models and factor variables from the factor analysis, sufficiently explained water quality variables with the potential relationship between environmental pollution in the region.

The Great Moderation means the reduction in the volatility of aggregate economic activity and here we use GDP growth rate to stand for economic activity. In this paper, we apply a Markov switching model to estimate the timing of the Great Moderation in France. Subsequently, by using a Time-varying structural vector autoregression model to determine which are the main variables that cause the reduction of French GDP growth rate and to see the relationship of these variables we choose.

This thesis compares the simple Autoregressive (AR) model against the k- Nearest Neighbor (k-NN) model to make a point forecast of five energy commodity prices. Those commodities are natural gas, heating oil, gasoline, ethanol, and crude oil. The data for the commodities are monthly and, for each commodity, two-thirds of the data are used for an in-sample forecast, and the remaining one-third of the data are used to perform an out-of-sample forecast. Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) are used to compare the two forecasts. The results showed that one method is superior by one measure but inferior by another. Although the differences of the two models are minimal, it is up to a decision maker as to which model to choose. The Diebold-Mariano (DM) test was performed to test the relative accuracy of the models. For all five commodities, the results failed to reject the null hypothesis indicating that both models are equally accurate.

Abstract Japan is the second largest economic power in the world. It has a great deal of FDI outflows but few FDI inflows. Therefore, Japan is in the serious situation of ¡§FDI balance of payments deficit.¡¨ In terms of inward FDI stocks as a percentage of GDP and gross fixed capital formation, Japan is the lowest place of G-7. The purpose of this research is focusing on discussing the shortage of FDI inflows and causing factors which lower the desires of investments in Japan by using the simplest way which is based on the actual situation and the limit of the information in Japan. This paper takes the quarterly data of Japan from 1978 to 2005 and four variables (wage index, real exchange rate, trade and FDI inflows). In this research, the unit root test is used to check if the data have the stationarity or not, and then it uses vector autoregression model (VAR) to proceed impulse response function and forecast error variance decomposition. According to the result of these two approaches, we can figure out the influences of four variables for each other, and then find out the causing factors which lead Japan to have less FDI inflows. The calculation shows that the reason which leads Japanese wages to increase gradually results not only from real exchange rate, trade and FDI inflows, but also from Japanese labor system (lifetime employment system and payment according to working seniority) and the labor quantities. The causality runs from real exchange rate to trade is greater than vice versa. Trade has a positive impact from the real exchange rate which means that the depreciation can accelerate trade. However, the main factor of hindering FDI inflows is Japanese high wages rather than real exchange rate or trade. Therefore, in order to get rid of the depression which was caused by the bubble economy in 1990s, Japanese government not only opens up the restrictions in policy but also takes the control of the prime costs into the most important consideration.

The purpose of this thesis is to identify the factors playing a key role in the determination of the Turkish stock market liquidity in aggregate terms in a time series context and discuss the joint dynamics of the market-wide liquidity with its selected determinants and the trade volume. The main determinants tested are the level of return, the return volatility and the monetary stance of the Central Bank of the Republic of Turkey. The expected positive relationship between the liquidity and the return is confirmed, while the negative effect of the volatility on liquidity appears one-week later. The behavior of various liquidity variables are also examined around the macroeconomic data announcement dates, during the 2008 financial crisis, and after the tick size change in the Istanbul Stock Exchange (ISE). The time series dynamics between the trade volume, return, volatility and the liquidity are put forward within the Vector Autoregression analysis framework. The GARCH modeling of the return series, which is an input to the liquidity model estimations, is a byproduct of this thesis. It is observed that the return series exhibits volatility clustering, persistence, leverage effects and mean reversion. In addition, while the level of the ISE market return decreased, the volatility of the return increased during the 2008 crisis. Accordingly, EGARCH model assuming normally distributed error terms and allowing a shift in the variance during the crisis period is chosen as the best model.

none

This paper reexamines the long-run real interest rate parity of the OECD countries by using the unit root test proposed by Ng and Perron (2001) and by the application of simulation to establish the small sample distribution under the null and the alternative hypothesis. By using the small sample distribution of the unit root statistics, we can make sure that first, size distortions are not the reasons contributing to the rejection of the fact that the alternative hypothesis is unit root. Second, the inference that the low power is not necessary causes the not rejecting the alternative hypothesis is correct. If still can not decide which distributions might cause the real interest difference series by comparing the unit root statistics and the relative location of the small sample distribution, we test that whether the series are asymmetric in those countries which we can not decide what kind of distributions they are by the threshold autoregression model proposed by Caner and Hansen (2001). Finally, the empirical results indicate that the RIPH holds in Australia¡BBelgium¡BCanada¡BFinland¡BFrance¡BGermany¡BJapan and Sweden whenever data frequency under linear time series model. Under quarterly data of Italy and United Kingdom and monthly data of Denmark, it turns out that the data have the traits of nonlinear time series model. Besides, the evidence of supporting the long-run real interest rate parity can not be reached and the phenomena that partial unit root exist in United Kingdom and Denmark.

Thesis (Ph. D.)--University of Missouri-Columbia, 1999. / Typescript. Vita. Includes bibliographical references (leaves 75-77). Also available on the Internet.