| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 4.78 MB | Adobe PDF |
Autores
Resumo(s)
Esta dissertação documenta o estudo e caracterização do mercado cambial e as ferramentas
de análise disponíveis para a sua negociação, com o objetivo de desenvolver uma estratégia
de negociação automatizada através de uma plataforma de negociação eletrónica.
Para tal, e após a caracterização deste mercado financeiro, são abordados os conceitos
fundamentais da análise técnica e fundamental, de modo a avaliar quais os indicadores e
padrões gráficos mais utilizados neste tipo de sistemas de negociação e quais as melhores
formas de os aplicar. Dentro dos indicadores foram estudados os quatro tipos de indicadores:
tendência, momento, volume e volatilidade e no caso dos padrões foram identificados os
tipos de Velas Japonesas mais utilizadas e os padrões agrupados mais comuns.
A definição da plataforma de negociação entre as várias opções do mercado e a seleção do
tipo de corretora a utilizar como intermediário do mercado são pontos relevantes neste
trabalho, cujas escolhas podem condicionar a implementação de estratégias de negociação e
até resultar em custos para o investidor/utilizador. Uma vez selecionada a plataforma de
negociação a utilizar – Metatrader 4 – serão desconstruídos e explicados com maior detalhe
os vários módulos e funcionalidades que a constituem e as boas práticas a aplicar.
O desenvolvimento da componente prática e a implementação do algoritmo foi sendo
progressiva, com a introdução à linguagem MQL4, a elaboração de funções para as diversas
funcionalidades do algoritmo, a definição do sistema de negociação e a seleção dos
indicadores a implementar de modo a que algoritmo automatizado efetue a substituição da
análise gráfica e manual realizada por um investidor. O desenvolvimento deste algoritmo,
denominado na plataforma por Expert Advisor (EA), foi sendo evolutivo, com a realização
de vários ensaios para se proceder à escolha de qual seria a melhor forma de aplicar um
indicador ou padrão gráfico no par de câmbio EUR/USD. Para os ensaios foi definido um
conjunto de períodos de otimização e teste, sendo a performance do algoritmo avaliada em
função do lucro líquido e do menor abaixamento de capital (drawdown) de cada ensaio. Após análise crítica, foram selecionadas as melhores configurações do algoritmo de modo a
maximizar os resultados positivos, ou lucros, mas de forma estável e estruturada ao longo
do tempo. O resultado final foi o desenvolvimento de um algoritmo com performance
bastante positiva nos ensaios realizados, onde o lucro acumulado e o abaixamento máximo
apresentaram valores interessantes, tendo em conta a maturidade do algoritmo criado.
This paper documents the study and characterization of the foreign exchange market, FOREX, and the analysis tools available for trading, in order to develop an automated trading strategy through an electronic trading platform. To that end, after a characterization of this financial market, an approach is performed covering the fundamental concepts of technical and fundamental analysis to assess which indicators and chart patterns are most commonly used in this type of trading systems and how best to apply them. Within the indicators four types of indicators were evaluated: trend, momentum, volume, and volatility, and in the case of the graphic patterns have been identified the most commonly used in Japanese Candlestick. The definition of the trading platform between the various options in the market and selecting the type of broker to use as a market intermediary are relevant points in this work, whose choices can influence the implementation of trading strategies and even result in costs to the investor / user. Once selected the trading platform to use - Metatrader 4 - will be deconstructed and explained in greater detail the existing modules and features and the good practices to apply. The development of the practical component and the implementation of the algorithm was progressive, with the introduction of the MQL4 language, the development of functions for the various features of the algorithm, the definition of the trading system and the selection of indicators to be implemented to the automated algorithm make the replacement of graphic and manual analysis by an investor. The development of this algorithm, called in the platform by Expert Advisor (EA), was evolutionary, with the completion of various tests to make the choice of how best to implement an indicator or chart pattern in the EUR/USD pair. For the assays was defined a set of optimization and testing periods, where the algorithm performance was evaluated on the basis of the total net profit and maximal drawdown of each test. After analysis, the best settings for the algorithm to maximize profit results were selected, but in a stable and structured manner over time. The end result was the development of an algorithm with very positive performance in tests, where the accumulated net profit and maximum drawdown showed interesting values, taking into account the maturity of the created algorithm.
This paper documents the study and characterization of the foreign exchange market, FOREX, and the analysis tools available for trading, in order to develop an automated trading strategy through an electronic trading platform. To that end, after a characterization of this financial market, an approach is performed covering the fundamental concepts of technical and fundamental analysis to assess which indicators and chart patterns are most commonly used in this type of trading systems and how best to apply them. Within the indicators four types of indicators were evaluated: trend, momentum, volume, and volatility, and in the case of the graphic patterns have been identified the most commonly used in Japanese Candlestick. The definition of the trading platform between the various options in the market and selecting the type of broker to use as a market intermediary are relevant points in this work, whose choices can influence the implementation of trading strategies and even result in costs to the investor / user. Once selected the trading platform to use - Metatrader 4 - will be deconstructed and explained in greater detail the existing modules and features and the good practices to apply. The development of the practical component and the implementation of the algorithm was progressive, with the introduction of the MQL4 language, the development of functions for the various features of the algorithm, the definition of the trading system and the selection of indicators to be implemented to the automated algorithm make the replacement of graphic and manual analysis by an investor. The development of this algorithm, called in the platform by Expert Advisor (EA), was evolutionary, with the completion of various tests to make the choice of how best to implement an indicator or chart pattern in the EUR/USD pair. For the assays was defined a set of optimization and testing periods, where the algorithm performance was evaluated on the basis of the total net profit and maximal drawdown of each test. After analysis, the best settings for the algorithm to maximize profit results were selected, but in a stable and structured manner over time. The end result was the development of an algorithm with very positive performance in tests, where the accumulated net profit and maximum drawdown showed interesting values, taking into account the maturity of the created algorithm.
Descrição
Palavras-chave
Mercado Cambial Indicadores Técnicos Plataformas de Negociação Eletrónica Sistemas de Negociação Expert Advisor FOREX – Foreign Market Exchange Technical Indicators Electronic Trading Platforms Trading Systems Expert Advisor