Predicciones a partir del reconocimiento de patrones de precios para estrategia de trading

Multiple predictive methods can be used as tools to try to anticipate the price movement of some asset and use it as trading strategy. This has been done previously with regressions, with models of neural networks, with Arima models, among other methodologies. This work explores the utility of finding repeating price patterns to make price prediction. The model was evaluated, not only statistically but also financially, for the Euro-Dollar market. Using a Python code capable of searching the current price pattern in the price history. This algorithm gives the order to buy or sell if the pattern has a certain percentage of repeating patterns that lead to a rise or fall. Finally, the model was evaluated and contrasted with three other predictive methods to determine its usefulness, accomplishing a satisfactory result that shows a possible potential of the tool developed especially due to the level of predictive accuracy of the 75% obtained with the tested data.

• Autores: Sánchez Torres, Pablo
• Tipo de recurso: Trabajo de grado de pregrado
• Fecha de publicación: 2020
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: spa
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/51397
• Acceso en línea: http://hdl.handle.net/1992/51397
• Palabra clave: Tipos de cambio; Euro; Dólar; Economía
• Rights: openAccess
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/