Thesis Creación y optimización de una estrategia de Trading para el mercado spot de forex
Loading...
Date
2014
Authors
Journal Title
Journal ISSN
Volume Title
Program
Ingeniería Civil Industrial
Departament
Campus
Abstract
En este estudio, se desea cubrir la totalidad, o la mayoría de los componentes necesarios para crear y automatizar una estrategia de trading y contestar todas las posibles preguntas que pueda llegar a tener el trader durante el proceso y durante las operaciones.
Así, no cabe espacio para decisiones basadas en juicios personales y peor aún, basadas en emociones. Plani_car cada aspecto del trading es clave para lograr operar en el largo plazo de manera rentable, consistente, disciplinada y con_ada.
Una vez presentados estos tópicos, se escoge un estilo de trading adecuado a un especulador novato y se crea una estrategia de trading basada en este estilo, esta se programa en el lenguaje MQL4, provisto por la plataforma MetaTrader4 y se realizan pruebas en data histórica para veri_car si es conveniente utilizar pruebas en data histórica como una herramienta para mejorar la rentabilidad de la estrategia.
Lamentablemente se obtiene una baja correlación entre rentabilidades de optimizaciones y walkforward tests, con lo cual se rea_rma la calidad dinámica del mercado spot de forex.
Se concluye que los esfuerzos intelectuales y creativos se deben destinar a crear estrategias de trading, y luego de empleadas estas en trading real, se puede comenzar un proceso de automatización, sabiendo a priori que la estrategia es rentable. Bajo este proceso lógico, se descarta dedicar muchos esfuerzos en pruebas de data histórica.
Por otro lado, las pruebas en data histórica sirven para ahorrar tiempo y descartar estrategias inservibles. Se debe tomar en consideración que tampoco se debe caer en un exceso, ya que de esta forma se incurre en el curve _tting, obteniéndose excelentes resultados que probablemente no se repetirán a futuro.
In this study, I wanted to cover all or most of the essentials needed to create and automate a trading strategy and answer all possible questions in this process and while trading. So, there is no room for decisions based on personal judgment and worse yet, based on emotions. Plan every aspect of trading is key to successfully operate in the long-term being pro_table, consistent, disciplined and con_dent. Once submitted these topics, a style of trading is chosen, suitable for a rookie trader and a trading strategy is designed based on this style, after this, the strategy is programmed on MQL4 language provided by the MetaTrader4 platform and tests are performed on historical data to verify whether to use historical data testing as a tool to improve the pro_tability of the strategy. Unfortunately a low correlation between walkforward tests returns and optimizations returns, whereby the dynamic quality of forex spot market is rea_rmed. I conclude that the intellectual and creative e_orts should be devoted to creating trading strategies, and then, after use these strategy in real trading, you can begin a process of automation, knowing beforehand that the strategy is pro_table. Under this logical process devote much e_ort in testing historical data is discarded. On the other hand, evidence based on historical data saves time by discarding useless strategies quickly. It should take into account that one should not fall into excess, because this ways falls into curve _tting, yielding excellent results that are unlikely to be repeated in the future.
In this study, I wanted to cover all or most of the essentials needed to create and automate a trading strategy and answer all possible questions in this process and while trading. So, there is no room for decisions based on personal judgment and worse yet, based on emotions. Plan every aspect of trading is key to successfully operate in the long-term being pro_table, consistent, disciplined and con_dent. Once submitted these topics, a style of trading is chosen, suitable for a rookie trader and a trading strategy is designed based on this style, after this, the strategy is programmed on MQL4 language provided by the MetaTrader4 platform and tests are performed on historical data to verify whether to use historical data testing as a tool to improve the pro_tability of the strategy. Unfortunately a low correlation between walkforward tests returns and optimizations returns, whereby the dynamic quality of forex spot market is rea_rmed. I conclude that the intellectual and creative e_orts should be devoted to creating trading strategies, and then, after use these strategy in real trading, you can begin a process of automation, knowing beforehand that the strategy is pro_table. Under this logical process devote much e_ort in testing historical data is discarded. On the other hand, evidence based on historical data saves time by discarding useless strategies quickly. It should take into account that one should not fall into excess, because this ways falls into curve _tting, yielding excellent results that are unlikely to be repeated in the future.
Description
Catalogado desde la version PDF de la tesis.
Keywords
Mercado de divisas, Planificación estratégica, Modelos predictivos, Toma de decisiones, Inversión financiera
