SVAR ,VECM and ARIMA model for forecasting exchange rate SVAR model has a better performance

This thesis presents a comparison for modeling and forecasting Chinese futures market of copper and aluminum with single time series and multivariate time series under linear restrictions. For single time series, data transformation for stationary purpose has been tested and performed before ARIMA model was built. For multivariate time series, co-integration rank test has been performed and included before VECM model was built. Based on selected models, the forecasting shows multivariate time series analysis has a better result than single time series, which indicates utilizing the relationships among the series can improve the accuracy of time series forecasting.

This paper provides methods to select pairs potentially profitable within the frame of statistical arbitrage. We employ a cointegration approach on pairwise combinations of five large energy companies listed on the New York Stock Exchange for the period 27th September 2012 to 22nd April 2014. We find one cointegrated pair, for which we further investigate both short and long run dynamics. A vector-error correction model is constructed, supporting a long run relationship between the two stocks, which is also supported by the mean-reverting characteristic of a stationary linear combination of the stocks. Impulse response functions and variance decomposition are also studied to further describe the interrelation of the stocks, supporting a unidirectional causality between the stocks.

This thesis focuses on five of the founding nations of the Association of Southeast Asian Nations (ASEAN). The countries are Malaysia, Singapore, Thailand, Indonesia and the Philippines (ASEAN5). Asset pricing for the ASEAN5 equity markets is the main focus of this thesis, although we also develop vector error correction models (VECM) for GDP, trade and local equity market returns for the ASEAN5. While this allows further analysis of the robustness of asset pricing models, it also facilitates study of the fundamental links that exist within these economies. The traditional CAPM and the four factor-model that include market, size, value and momentum effects (Fama and French, 1993; Carhart, 1997) are employed in testing the variation in size/book-to-market equity (size-BTME) and industry portfolio returns for these markets for the period from January 1990 to March 2006. Three macro factors as well as world excess returns are then added to the basic four-factor asset pricing model. These macro factors include unexpected GDP, unexpected total trade and unexpected equity market returns, which are derived from VECM or VAR estimates for ASEAN5 GDP, total trade and equity market returns. This model is referred to as the macro-factor model. The results suggest that the explanatory power of the four-factor model consistently exceeds those of the one-factor CAPM in explaining size-BTME and industry portfolio returns. Further, the macro-factor model analysis suggests that collectively, this model does not substantially improve the explanatory power of the basic four-factor model, suggesting that the variation in portfolio returns is mostly captured by the four-factor model. There is some cross-country variation in these results. Regardless, these macro factors − taken as a group or individually − are statistically significant, particularly for Thailand and Malaysia. In addition, the cointegration test results document evidence of long-run linkages for the equity markets within the ASEAN5. This is also true for GDP within the ASEAN5. In both cases, closer links prevail in the post-crisis period. This is not the case for trade where there is little consistent evidence of close links between the countries. Mixed results are found for different ASEAN5 trade measures where the linkages for total trade, import and exports vary substantially according to the selected period of study, whether full period, pre-crisis or post-crisis period.

No description available.

>Magister Scientiae - MSc / Purpose:This research investigates the existence of long-term equilibrium relationships among the stock market indices of Brazil, Russia, India, China and South Africa (BRICS). It further investigates cointegrated stock pairs for possible implementation of statistical arbitrage trading techniques.Design:We utilize standard multivariate time series analysis procedures to inspect unit roots to assess stationarity of the series. Thereafter, cointegration is tested by the Johansen and Juselius (1990) procedure and the variables are interpreted by a Vector Error Correction Model (VECM). Statistical arbitrage is investigated through the pairs trading technique.Findings:The five stock indices are found to be cointegrated. Analysis shows that the cointegration rank among the variables is significantly influenced by structural breaks. Two pairs of stock variables are also found to be cointegrated. This guaranteed the mean reversion property necessary for the successful execution of the pairs trading technique. Determining the optimal spread threshold also proved to be highly significant with respect to the success of this trading technique.Value:This research seeks to expand on the literature covering long-run co-movements of the volatile emerging market indices. Based on the cointegration relation shared by the BRICS, the research also seeks to encourage risk taking when investing. We achieve this by showing the potential rewards that can be realized through employing appropriate statistical arbitrage trading techniques in these markets.

This thesis shows that the relationship between economic growth and expansions of telephone main lines and electricity generating capacity is two-way, when looking at the period of 1955 - 1995 and half of the world's countries. In other words, expansions of these two kinds of infrastructure seem to both initiate and be induced by economic growth, highlighting the problem of much previous research that does not account for a bi-directional relationship. Furthermore, this research suggests that the effect that these two kinds of infrastructure have on economic growth was during this period great enough to be of policy interest, and it is likely that it can explain part of the vast differences seen between countries in GDP per capita today. While the impact that these two kinds of infrastructure had during this specific time will surely not be the same in the future, the results still point to a potentially important role for infrastructure expansions in determining economic growth, even though the kinds of infrastructure that have most impact will vary with time and technological progress.

Made available in DSpace on 2016-03-15T19:30:48Z (GMT). No. of bitstreams: 1 Alexandre Ripamonti.pdf: 1715355 bytes, checksum: bafe49730ceb2c3f261ef6a51ffb5f5c (MD5) Previous issue date: 2011-08-22 / Fundo Mackenzie de Pesquisa / Rational valuation formula and time varying cointegration are the main thesis´ concepts, under the Muth´s (MUTH, 1961) rational expectations and theory of price movements as underlying theory, and also testing the null of time invariant error correction mechanisms and another one of inequality of fundamental value and share´s price. The data were obtained from Brazilian listed companies for 1986 to 2010 and also from 1871 to 2010 US stock market. The Johansen´s maximum likelihood and trace models, combined to Chebyshev time polynomials, as proposed by Bierens and Martins (2010) were used in order to test the null. The finds have shown first null rejection and no rejection for the second null. These finds are consistent to Bierens e Martins (BIERENS e MARTINS, 2010) and non-consistent with Muth (MUTH, 1961) / A presente tese aborda os conceitos de fórmula de valoração racional e cointegração variante no tempo para, sob o referencial da teoria das expectativas racionais e de movimentação de preços de Muth (MUTH, 1961), supor a variabilidade das taxas de retorno de ativos no mercado brasileiro, no período de 1986 a 2010, testando as hipóteses nulas de mecanismos de correção de erros dos vetores de cointegração constantes no tempo e de desigualdade entre valor fundamental e preço da ação. Foram coletados dados de preços e dividendos de ações componentes da carteira teórica do IBOVESPA de janeiro de 1986 a outubro de 2010. Além disso, também aplicamos os modelos propostos aos dados norte-americanos de preço e dividendos de 1871 a 2010, disponibilizados por Shiller. Os dados foram analisados através das técnicas de séries temporais e os coeficientes estimados através da técnica de máxima verossimilhança, especificamente com os modelos de cointegração de Johansen combinados com os polinômios temporais de Chebyshev, como proposto por Bierens e Martins (2010). Os resultados indicam a rejeição da hipótese nula de constância dos vetores de cointegração e, ainda, a não rejeição da hipótese nula de desigualdade entre valor fundamental e preço da ação para todas as séries temporais analisadas. Tais resultados são consistentes com os obtidos por Bierens e Martins (BIERENS e MARTINS, 2010) e não consistentes com a teoria das expectativas racionais de Muth (MUTH, 1961).

This thesis examines the validity of the cost of carry model for pricing FTSE100 futures contracts and the relationship between FTSE100 spot and futures markets during two sub-periods characterised by different market trading systems employed by the LSE and LIFFE. The empirical work is carried out using three approaches to econometric modeling: a basic VECM for spot and futures prices, a VECM extended with a DCCTGARCH framework to account for the conditional variance-covariance structure for spot and futures prices and a threshold VECM to capture regime-dependent spot-futures price dynamics. Overall, both the basic VECM and the DCC-TGARCH analysis suggest that there are deviations from the cost of carry relationship in the first sub-sample when transactions costs in both markets are relatively high but that the cost of carry relationship tends to be valid in the second sub-sample when transactions costs are lower. This is further confirmed by the evidence of higher conditional correlations between the two markets in the second sub-sample as compared with the first, using the DCC-TGARCH analysis. This implies that the no-arbitrage cost of carry relationship between spot and futures markets is more effectively maintained by index arbitrageurs in the second period when market conditions are closer to perfect market assumptions, and hence the cost of carry model could be more reasonably used as a benchmark for pricing stock index futures. The threshold VECM analysis depicts regime-dependent price dynamics between FTSE100 spot and futures markets and leads to some interesting and important findings: arbitrage may not be practicable under some market conditions, either because it is difficult to find counterparties for the arbitrage transactions, or because there is significant risk associated with arbitrage; as a result, the cost of carry model may not always be suitable for pricing stock index futures. Furthermore, the threshold values yielded from estimating the threshold VECM reflect the average transaction costs for most arbitrageurs that are more reliable and fair than subjective estimations.

No description available.