This thesis concerns the asymptotic behaviour of the sequence of 1-step recursive Chow statistics and various tests derived therefrom. The 1-step statistics are produced as diagnostic output in standard econometrics software, and are expected to reflect model misspecification. Such misspecification testing is important in validating the assumptions of a model and so ensuring that subsequent inference is correct. Original contributions to the theory of misspecification testing include (i) a result on the pointwise convergence of the 1-step statistics; (ii) a result on the extreme-value convergence of the maximum of the statistics; and (iii) a result on the weak convergence of an empirical process formed by the statistics. In Chapter 2, we describe the almost sure pointwise convergence of the 1-step statistic for a broad class of time series models and processes, including unit root and explosive processes. We develop an asymptotic equivalence result, and use this to establish the asymptotic distribution of the maximum of a sequence of 1-step statistics with normal errors. This allows joint consideration of the sequence of 1-step tests via its maximum: the sup-Chow test. In Chapter 3, we use simulation to investigate the power properties of this test and compare it with benchmark tests of structural stability. We find that the sup-Chow test may have advantages when the nature of instability is unknown. In Chapter 4, we consider how the test may be adapted to situations in which the errors cannot be assumed normal. We evaluate several promising approaches, but also note a trade-off between robustness and power. In Chapter 5 we analyse an empirical process formed from the 1-step statistics, and prove a weak convergence result. Under the assumption of normal errors, the limiting distribution reduces to that of a Brownian bridge. The asymptotic approximation appears to works well even in small samples.

We are interested in the stochastic properties, individual and joint, of mis- specification testing when the data are generated by an autoregressive process. Good mis-specification tests are invariant to the dynamic properties of the pro- cess summarized by its characteristic roots, and to irrelevant misspecifications. Invariance in parameter space obviates inference prior to mis-specification test- ing. This is important as the latter is used to validate the former. Mutual independence of the tests allows calibration of the overall significance level. Es- tablishing such results requires work on individual tests and on their stochastic interactions. In Chapter 2, we derive the asymptotic distribution of two types of CUSUM of squares test, one implemented with standardized one-step-ahead OLS pre- diction errors and another implemented with OLS residuals. The latter is found to be valid in all but singular explosive cases, but the former only in purely non-explosive, or regular explosive cases with all roots in the explosive region of the parameter space; in Chapter 3, we show that a nuisance term arises in the mixed case. In Chapter 4, we derive numerically a finite sam- ple correction to render the tests implementable into software, and Chapter 1 contains two examples of applications. In Chapter 5, we consider inference on the parameters associated with the stationary part of the process, together with tests for a unit root, lag length, variance constancy, and normality of the regression innovations. In character- izing the joint distribution of these tests, we rely on asymptotic theory, and show independence in the limit. A simulation experiment suggests that finite sample correlations between some of the tests are statistically significant but small. Asymptotically, then, control of the overall significance level of the test procedure is feasible, and there is no reason to discount inference for the use of these mis-specification tests in model selection.

With recent financial crisis the importance of liquidity not only as indicator of financial health of banks heightened. Thus this thesis aims the focus to relationship between real economy and bank liquidity creation, and provides empirical evidence of significant relationship between bank liquidity creation and GDP or inflation. Moreover, it shows that implementation of bank liquidity creation indicator into Taylor rule, in order to address for financial stability and health, is suitable alternative for financial stress index.

This dissertation includes two chapters in the field of international finances about foreign exchange rate predictability and testing purchase power parity. In each chapter, we build the theory, methodology, and the empirical results to present the paper¡¦s construction. The first chapter, we studies whether the pure price inflation rate which is extracted from stock return can help us to test the relative of purchasing power parity in where Asian countries include Malaysia, Korea, Taiwan, Thailand, Hong Kong, and Singapore against the United States. The paper of Chowdhry et al. (2005) argue that relative PPP may not hold for the official price inflation rates which is constructed from consumer price indices, since relative price changes and other frictions cause price to be sticky. Thus, they use the Fisher equation and Fama-French three factors elaborately to build up a model on the nominal return of real risk-free asset to extract the pure price inflation rates. Their argument is supported in the case of Japan, Germany, the United States, and the United Kingdom. We are interested in the case of some Asian countries. So, this chapter, we extend the model and methodology of Chowdhry et al. (2005) to test the relative PPP for Asian countries. If our empirical evidence is firmly supported, it will be a strongly reconfirmed the elaborated idea of Chowdhry et al. (2005). In our study, the PPP rule is not supported for Asian countries since joint null hypothesis of a=0 and b=1 are rejected at all horizons except Taiwan at monthly horizon. The testing results by constrained seeming unrelated regression (SUR) and system equation in pooled data are similar to the tests of country-by-country. Therefore, we apply the methods of panel unite root from Im et al. (2003), Maddala and Wu (1999), and Pesaran (2007) to test the PPP doctrine, and it is strongly supported PPP for Asian countries. The second chapter, we extract the estimated data of pure price inflation by Chowdhry et al. (2005), and use the data to build up a nonlinear STR (smooth transition autoregressive) model by Granger and Teräsvirta (1993), then compare the performance of linear or nonlinear model of exchange rate predictability with random walk model in the United States, the United Kingdom, Japan, and Germany. This study has presented evidences that the extracted inflation rates offer a good predictability on the prediction of exchange rate for the United Kingdom and Germany. Those extracted data in which are calculated from the industry portfolio returns of stock market. The issue of series correlation in regression error does matter the estimated coefficients £]k, thus we estimate the simulation of Gaussian bootstrap distribution for testing variables with Newey West standard deviation in regression estimate. The empirical evidences show that the PPP doctrine affects the predictability performance of exchange rate change by the extracted inflation rates.

Vector autoregression (VAR) models are widely used in an attempt to identify and measure the effect of monetary policy shocks on an economy and to forecast economic times series. However, the sparse information sets used in the VAR approach have been subject to criticism and in recent decades, the use of factor models as a means of dimension reduction has been a subject of greater focus. The method of summarizing information contained in a large set of macroeconomic time series by principal components, and use these as regressors in VAR models, has been pointed out as a potential solution to the problems of limited information and estimation of too many parameters. This paper combines the standard VAR methodology with dynamic factor analysis on Swedish data for two purposes, to assess the effects of monetary policy shocks and to examine the forecasting properties. Latent factors estimated by the principal components method are in this study found to contribute to a more coherent picture in line with economic theory, when examining monetary policy shocks to the Swedish economy. The factor-augmented models can on the other hand not be shown to increase the forecasting accuracy to a great extent compared to standard models.

This paper puts forward a Bayesian Global Vector Autoregressive Model with Common Stochastic Volatility (B-GVAR-CSV). We assume that Country specific volatility is driven by a single latent stochastic process, which simplifies the analysis and implies significant computational gains. Apart from computational advantages, this is also justified on the ground that the volatility of most macroeconomic quantities considered in our application tends to follow a similar pattern. Furthermore, Minnesota priors are used to introduce shrinkage to cure the curse of dimensionality. Finally, this model is then used to produce predictive densities for a set of macroeconomic aggregates. The dataset employed consists of quarterly data spanning from 1995:Q1 to 2012:Q4 and includes 45 economies plus the Euro Area. Our results indicate that stochastic volatility specifications influences accuracy along two dimensions: First, it helps to increase the overall predictive fit of our model. This result can be seen for some variables under scrutiny, most notably for real GDP and short-term interest rates. Second, it helps to make the model more resilient with respect to outliers and economic crises. This implies that when evaluated over time, the log predictive scores tend to show significantly less variation as compared to homoscedastic models. (author's abstract) / Series: Department of Economics Working Paper Series

Exponential smoothing algorithms are very attractive for the practical world such as in industry. When considering bivariate exponential smoothing methods, in addition to the properties of univariate methods, additional properties give insight to relationships between the two components of a process, and also to the overall structure of the model. It is important to study these properties, but even with the merits the bivariate exponential smoothing algorithms have, exponential smoothing algorithms are nonstatistical/nonstochastic and to study the properties within exponential smoothing may be worthless. As an alternative approach, the (bivariate) ARIMA and the structural models which are classes of statistical models, are shown to generalize the exponential smoothing algorithms. We study these properties within these classes as they will have implications on exponential smoothing algorithms. Forecast properties are studied using the state space model and the Kalman filter. Comparison of ARIMA and structural model completes the study. / Mathematical Sciences / M. Sc. (Statistics)

Made available in DSpace on 2008-05-13T13:47:40Z (GMT). No. of bitstreams: 1 1850.pdf: 600111 bytes, checksum: 512f00eac39c4e25d72fa4b37ecc7e97 (MD5) Previous issue date: 2005-04-12 / Using Brazilian financial data for some shares traded in the Brazilian Stock Market (BOVESPA) we test the expectation hypothesis of present value models discounted by a constant factor. This model relates the price of a stock to its expected dividends. To perform econometric testing we use mainly the jointly restriction through Wald Test in a Vector Autoregression framework, as well as alternative testing procedures. The empirical results partially support the present value model discounted by a constant factor to predict prices for stock through its expected dividends. / Através de dados financeiros de ações negociadas na Bolsa de Valores de São Paulo, testa-se a validade do modelo de valor presente (MVP) com retornos esperados constantes ao longo do tempo (Campbell & Schiller, 1987). Esse modelo relaciona o preço de uma ação ao seu esperado fluxo de dividendos trazido a valor presente a uma taxa de desconto constante ao longo do tempo. Por trás desse modelo está a hipótese de expectativas racionais, bem como a hipótese de previsibilidade de preço futuro do ativo, através da inserção dos dividendos esperados no período seguinte. Nesse trabalho é realizada uma análise multivariada num arcabouço de séries temporais, utilizando a técnica de Auto-Regressões Vetoriais. Os resultados empíricos apresentados, embora inconclusivos, permitem apenas admitir que não é possível rejeitar completamente a hipótese de expectativas racionais para os ativos brasileiros.

Exponential smoothing algorithms are very attractive for the practical world such as in industry. When considering bivariate exponential smoothing methods, in addition to the properties of univariate methods, additional properties give insight to relationships between the two components of a process, and also to the overall structure of the model. It is important to study these properties, but even with the merits the bivariate exponential smoothing algorithms have, exponential smoothing algorithms are nonstatistical/nonstochastic and to study the properties within exponential smoothing may be worthless. As an alternative approach, the (bivariate) ARIMA and the structural models which are classes of statistical models, are shown to generalize the exponential smoothing algorithms. We study these properties within these classes as they will have implications on exponential smoothing algorithms. Forecast properties are studied using the state space model and the Kalman filter. Comparison of ARIMA and structural model completes the study. / Mathematical Sciences / M. Sc. (Statistics)

This dissertation examines the causal relationship between inward foreign direct investment (FDI) and economic growth (GDP) in SADC countries. The study investigates, within a panel data context, whether causation is short-term, long-term or both; and explores whether the causal relationship between the two variables differs according to income level. The study covered a panel of 15 SADC countries over the period 1980-2012. In order to assess whether the causal relationship between FDI inflows and economic growth is dependent on the level of income, the study divided the SADC countries into two groups, namely, the low-income and the middleincome countries. The study used the recently developed panel data analysis methods to examine this causal relationship. It adopted a three stage approach, which consists of panel unit root, panel cointegration and Granger causality to examine the dynamic causal relationship between the two variables. Panel unit root results show that both variables in the two SADC country groups were integrated of order one. Panel cointegration tests showed that the variables for low-income country group were not cointegrated, while the variables for the middle-income countries were cointegrated. Since the low-income country group panels were not cointegrated, Grangercausality tests were conducted within a VAR framework, while causality tests for the middleincome country group were conducted within an ECM framework. Panel Granger causality results for the low-income countries showed no evidence of causality in either direction. However, for the middle-income countries’ panel, there was evidence of a unidirectional causal flow from GDP to FDI in both the long- and short- run. The study concludes that the FDI-led growth hypothesis does not apply to SADC countries. The results imply that the recent high economic growth rates recorded in the SADC region, especially middle-income countries, have been attracting FDI. In other words, it is economic growth that drives FDI inflows into the SADC region, and not vice versa. These findings have profound policy implications for the SADC region at large and individual countries. / Economics / MCOM (Economics)