Loading...
Publication
A hipótese de eficiência nos mercados financeiros ASEAN-5
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1002.17 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O Coronavírus Covid-19 foi um surto que apareceu pela primeira vez em dezembro de 2019 na cidade de Wuhan, província de Hubei, China. Foi declarado como uma pandemia pela Organização Mundial de Saúde (OMS) em 12 de março de 2020. Face a estes acontecimentos esta investigação pretende testar a eficiência, na sua forma fraca, nos mercados de ações da Indonésia, Malásia, Tailândia, Singapura, Filipinas, China, Japão, Coreia do Sul, e os EUA, no período entre 01 de julho de 2019 a 28 de outubro de 2020. Para realizar este estudo foram empregues diferentes abordagens de forma a responder às seguintes questões de investigação: i) A pandemia global de 2020 fez diminuir a eficiência dos mercados de ações em análise?. ii) Se os coeficientes de correlação cruzada entre mercados forem fortes a hipótese de diversificação de carteiras poderá ser colocada em causa?. Os resultados sugerem que a hipótese de random walk é rejeitada em todos os índices bolsistas, quer em Ranking (homocedástico), quer em Sinais (heterocedástico). Os resultados apoiam, portanto, a conclusão de que a hipótese de random walk não é suportada pelos mercados financeiros analisados, neste período de pandemia global. Os valores dos rácios de variâncias são inferiores à unidade, designadamente nos meses de fevereiro e com um maior significado em março de 2020, o que implica que as rendibilidades estão autocorrelacionadas no tempo e, existe reversão à média, não tendo sido identificadas diferenças entre os mercados financeiros. Os expoentes 𝛼𝐷𝐹𝐴 mostram a presença de memórias longas, permitindo evidenciar que o pressuposto da hipótese de eficiência de mercado poderá ser colocado em causa, uma vez que a previsão do movimento de mercado pode ser melhorada se considerados os movimentos defasados dos restantes mercados, possibilitando a ocorrência de operações de arbitragem. Estes achados corroboram o teste de Ranking e Sinais de Wright, e validam a primeira questão de investigação. Os valores dos 𝑟h𝑜𝐷𝐶𝐶𝐴 mostram 28 coeficientes de correlação cruzada sem tendência fortes (0.666 → ≌ 0.999), 7 coeficientes de correlação médios (≌ 0.333 → ≌ 0.666), e 1 coeficiente de correlação fraco (≌ 0.000 → ≌ 0.333). Estes resultados evidenciam que a hipótese de diversificação de carteiras poderá ser colocada em causa, validam assim a segunda questão de investigação.
Coronavirus, Covid-19 was an outbreak that first appeared in December 2019 in the city of Wuhan, Hubei province, China. It was declared a pandemic by the World Health Organization (WHO) on March 12, 2020. Facing these events this investigation aims to test the efficiency, in its weak form, of the stock markets of Indonesia, Malaysia, Thailand, Singapore, Philippines, China, Japan, South Korea, and the USA, in the period between July 1, 2019, and October 28, 2020. To conduct this study, different approaches were employed to answer the following research questions: i) Did the 2020 global pandemic decrease the efficiency of the stock markets under analysis? ii) If the cross-correlation coefficients between markets are strong could the hypothesis of portfolio diversification be questioned? The results obtained showed that the hypothesis of random walk is rejected in all stock market indices, either in Ranking (homoscedastic) or in Signs (heteroscedastic). The results support, therefore, the conclusion that the hypothesis of random walk is not supported by the financial markets, in this period of a global pandemic. The values of the variance ratios are lower than the unit, namely in February and with a greater significance in March 2020, which implies that the returns are autocorrelated in time and there is a reversion to the average, not identifying differences between the financial markets. 𝛼𝐷𝐹𝐴 exponents show the presence of long memories, showing that the assumption of the market efficiency hypothesis can be called into question, since the forecast of market movement can be improved if we consider the lagged movements of the other markets, enabling the occurrence of arbitration operations. These findings corroborate the Wright's Ranking and Signals test and validate the first investigation question. The rhoDCCA values show 28 strong correlation coefficients without a strong trend (0.666 → ≌ 0.999), 7 average correlation coefficients (≌ 0.333 → ≌ 0.666), and 1 weak correlation coefficient (≌ 0.000 → ≌ 0.333). These findings portray the questioning of the hypothesis of portfolio diversification, thus, working to validate the second research question.
Coronavirus, Covid-19 was an outbreak that first appeared in December 2019 in the city of Wuhan, Hubei province, China. It was declared a pandemic by the World Health Organization (WHO) on March 12, 2020. Facing these events this investigation aims to test the efficiency, in its weak form, of the stock markets of Indonesia, Malaysia, Thailand, Singapore, Philippines, China, Japan, South Korea, and the USA, in the period between July 1, 2019, and October 28, 2020. To conduct this study, different approaches were employed to answer the following research questions: i) Did the 2020 global pandemic decrease the efficiency of the stock markets under analysis? ii) If the cross-correlation coefficients between markets are strong could the hypothesis of portfolio diversification be questioned? The results obtained showed that the hypothesis of random walk is rejected in all stock market indices, either in Ranking (homoscedastic) or in Signs (heteroscedastic). The results support, therefore, the conclusion that the hypothesis of random walk is not supported by the financial markets, in this period of a global pandemic. The values of the variance ratios are lower than the unit, namely in February and with a greater significance in March 2020, which implies that the returns are autocorrelated in time and there is a reversion to the average, not identifying differences between the financial markets. 𝛼𝐷𝐹𝐴 exponents show the presence of long memories, showing that the assumption of the market efficiency hypothesis can be called into question, since the forecast of market movement can be improved if we consider the lagged movements of the other markets, enabling the occurrence of arbitration operations. These findings corroborate the Wright's Ranking and Signals test and validate the first investigation question. The rhoDCCA values show 28 strong correlation coefficients without a strong trend (0.666 → ≌ 0.999), 7 average correlation coefficients (≌ 0.333 → ≌ 0.666), and 1 weak correlation coefficient (≌ 0.000 → ≌ 0.333). These findings portray the questioning of the hypothesis of portfolio diversification, thus, working to validate the second research question.
Description
Keywords
Eficiência de Mercado Random Walk ASEAN-5 COVID-19 Market Efficiency