Examinando por Materia "GARCH"

Mostrando 1 - 7 de 7
  • No hay miniatura disponible
    Publicación
    Currency Prediction : Stochastic hybrid diferencial equations with LSTM
    (Universidad EAFIT, 2024) Arbeláez Betancur, Hoover Arley; Marín Sánchez, Fredy Hernán
  • No hay miniatura disponible
    Publicación
    Desarrollo y comparación de modelos ARIMA-GARCH y SARIMA-GARCH para la estimación del tipo de cambio USD/COP y propuesta de coberturas cambiarias con derivados forward para empresa importadora de autopartes en Colombia
    (Universidad EAFIT, 2025) Caballero Rosas, Daniel; Molina Sierra, Luis Felipe
    This research analyzed the estimation of the USD/COP exchange rate through the development and comparison of ARIMA-GARCH and SARIMA-GARCH models to design hedging strategies. Historical data from the Representative Market Rate (2019-2024) and optimization techniques in Python were used. Results indicated that SARIMA-GARCH provided higher predictive accuracy by capturing seasonal fluctuations and reducing errors compared to ARIMA-GARCH. Based on these forecasts, forward contract hedging strategies were proposed to mitigate exchange rate risk. However, market uncertainty and unexpected events may affect model accuracy, making recalibration every 60-90 days advisable. The combination of time series models with conditional heteroskedasticity proved essential in volatile markets, although its high computational demand can be a limitation. This study provides applicable tools for exchange rate risk management, optimizing financial decision making for importing companies.
  • No hay miniatura disponible
    Publicación
    Determinantes de la volatilidad del precio interno del café en el mercado colombiano entre los años 2015 y 2022
    (Universidad EAFIT, 2024) Chávez Pinzón , Laura María; Pardo Calderón, Lina María; Cruz Castañeda, Vivian; Ninguno
    Colombia is a country recognized worldwide for its agricultural products, especially for its coffee. Ninety-six percent of the coffee producers are small farmers that are exposed daily to the fluctuation of prices in the international market and, in many occasions, this dependence does not cover their production costs, encouraging the exit or little participation of the producers in the coffee industry. The main variables that influence the behavior of internal coffee prices are: the price of the C contract, the quality premium and the exchange rate. In this study a quantile regression of the returns of internal coffee prices as a function of the conditional volatility of the returns of the C Contract and the TRM, between the years 2015 and 2022, using ARMA-GARCH models, was carried out. The most impactful results show that a unit increase in the conditional volatility of the TRM returns at the 0.95 quantile generates an increase in domestic coffee returns by 0.97% and that a unit increase in the conditional volatility of C contracts generates an increase of 0.31% in the returns of domestic coffee prices. This study offers valuable insights that can contribute to improve the resilience and sustainability of the Colombian coffee industry. Furthermore, by identifying the factors that influence domestic coffee prices, this study lays the foundation for future research and policies aimed at mitigating volatility and promoting more equitable and stable conditions for coffee producers in the country.
  • No hay miniatura disponible
    Ítem
    GARCH-type volatility in the multiplicative quadrinomial tree method: An application to real options
    (Universidad Nacional Autonoma de Mexico, 2020-03-03) Pareja, J.; Marin-Sanchez, Freddy H.; Universidad EAFIT. Escuela de Ciencias; Modelado Matemático
    This article applies the multiplicative quadrinomial tree numerical method with non-constant volatility to assess a real option of abandonment, based on an estimate of the conditional volatility for WTI oil commodity prices and their respective equivalence in a GARCH-diffusion model. The methodology refers to the use of an estimate of type GARCH (1,1) and the numerical method through a quadrinomial tree. There are two main findings: 1) when employing the quadrinomial method, the value of the real option turned out to be greater than the value estimated through the traditional multiplicative binomial method, due to underestimation of the real value of volatility that occurs in a specific period according to the latter method; and 2) a methodological contribution that demonstrates plainly way the presence of non-constant conditional volatility as well as being able to value all types of options using stochastic volatility. © 2019 Universidad Nacional Autónoma de México, Facultad de Contaduría y Administración. This is an open access article under the CC BY-NC-SA (https://creativecommons.org/licenses/by-nc-sa/4.0/)
  • No hay miniatura disponible
    Publicación
    GARCH-type volatility in the multiplicative quadrinomial tree method: An application to real options
    (Universidad Nacional Autonoma de Mexico, 2020-03-03) Pareja, J.; Marin-Sanchez, Freddy H.; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)
    This article applies the multiplicative quadrinomial tree numerical method with non-constant volatility to assess a real option of abandonment, based on an estimate of the conditional volatility for WTI oil commodity prices and their respective equivalence in a GARCH-diffusion model. The methodology refers to the use of an estimate of type GARCH (1,1) and the numerical method through a quadrinomial tree. There are two main findings: 1) when employing the quadrinomial method, the value of the real option turned out to be greater than the value estimated through the traditional multiplicative binomial method, due to underestimation of the real value of volatility that occurs in a specific period according to the latter method; and 2) a methodological contribution that demonstrates plainly way the presence of non-constant conditional volatility as well as being able to value all types of options using stochastic volatility. © 2019 Universidad Nacional Autónoma de México, Facultad de Contaduría y Administración. This is an open access article under the CC BY-NC-SA (https://creativecommons.org/licenses/by-nc-sa/4.0/)
  • No hay miniatura disponible
    Ítem
    Quadrinomial Trees to Value Options in Stochastic Volatility Models
    (Institutional Investor Systems, 2019-01-01) Pareja-Vasseur, Julian A.; Marin-Sanchez, Freddy H.; Universidad EAFIT. Escuela de Ciencias; Modelado Matemático
    This article describes in detail the multiplicative quadrinomial tree numerical method with nonconstant volatility, based on a system of stochastic differential equations of the GARCH-diffusion type. The methodology allowed for the derivation of the first two moments of the proposed equations to estimate the respective recombination between discrete and continuous processes and, as a result, a numerical methodological proposal is formally presented to value, with relative ease, both real and financial options, when the volatility is stochastic. The main findings showed that in the proposed method, when volatility approaches zero, the multiplicative binomial traditional method is a particular case, and the results are comparable between these methodologies, as well as to the exact solution offered by the Black-Scholes model. Finally, the originality of the methodological proposal is that it allows for the emulation in a simple way of the presence of a nonconstant volatility in the price of the underlying asset, and it can be used to value all kinds of options both in the real world and in risk-neutral situations.
  • No hay miniatura disponible
    Publicación
    Quadrinomial Trees to Value Options in Stochastic Volatility Models
    (Institutional Investor Systems, 2019-01-01) Pareja-Vasseur, Julian A.; Marin-Sanchez, Freddy H.; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)
    This article describes in detail the multiplicative quadrinomial tree numerical method with nonconstant volatility, based on a system of stochastic differential equations of the GARCH-diffusion type. The methodology allowed for the derivation of the first two moments of the proposed equations to estimate the respective recombination between discrete and continuous processes and, as a result, a numerical methodological proposal is formally presented to value, with relative ease, both real and financial options, when the volatility is stochastic. The main findings showed that in the proposed method, when volatility approaches zero, the multiplicative binomial traditional method is a particular case, and the results are comparable between these methodologies, as well as to the exact solution offered by the Black-Scholes model. Finally, the originality of the methodological proposal is that it allows for the emulation in a simple way of the presence of a nonconstant volatility in the price of the underlying asset, and it can be used to value all kinds of options both in the real world and in risk-neutral situations.