Predictor alternativo del precio de los activos financieros. El caso de la plata y el oro
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Oikos Polis
Abstract
El objetivo de este trabajo es comprobar la eficacia del modelo ARIMA optimizado con fuerza bruta operacional para pronosticar en el mercado financiero el precio de dos commodities: oro y plata, como una técnica alternativa que permita rentabilizar inversiones y mejorar la toma de decisiones de los actores financieros. Se utilizó información de los precios de cierre semanales del oro y de la plata durante el período 2015 - 2018, observando las variaciones de precios y comparando los datos reales con las pronosticadas a través del modelo. Se utilizaron para ambos modelos ocho variables, generando un millón de iteraciones aleatorias con fuerza bruta, ya que esta técnica no restringe la obtención de algún resultado, como lo hace la optimización por simplex y/o solver. Con la técnica de fuerza bruta se pudo establecer una capacidad de predicción en ambos activos superior al 60%.
The objective of this work is to test the effectiveness of the ARIMA model optimized with operational gross force to forecast the price of two commodities: gold and silver in the financial market, as an alternative technique to make investments profitable and improve the decision-making of financial actors. We used information from the weekly closing prices of gold and silver during the period 2015 - 2018, observing the price variations and comparing the real data with the ones predicted through the model. Eight variables were used for both models, generating one million random iterations with gross force, as this technique does not restrict the obtaining of any result, as does simplex and/or solver optimization. With the brute force technique it was possible to establish a predictive capacity in both assets of more than 60%.
The objective of this work is to test the effectiveness of the ARIMA model optimized with operational gross force to forecast the price of two commodities: gold and silver in the financial market, as an alternative technique to make investments profitable and improve the decision-making of financial actors. We used information from the weekly closing prices of gold and silver during the period 2015 - 2018, observing the price variations and comparing the real data with the ones predicted through the model. Eight variables were used for both models, generating one million random iterations with gross force, as this technique does not restrict the obtaining of any result, as does simplex and/or solver optimization. With the brute force technique it was possible to establish a predictive capacity in both assets of more than 60%.
Description
Vol. 6, No. 2