Narges Salehnia, Mohamad Ali Falahi, Ahmad Seifi, Mohammad Hossein Mahdavi Adeli,
Volume 4, Issue 14 (12-2013)
Abstract
Developing models for accurate natural gas spot price forecasting is critical because these forecasts are useful in determining a range of regulatory decisions covering both supply and demand of natural gas or for market participants. A price forecasting modeler needs to use trial and error to build mathematical models (such as ANN) for different input combinations. This is very time consuming since the modeler needs to calibrate and test different model structures with all the likely input combinations. In addition, there is no guidance about how many data points should be used in the calibration and what accuracy the best model is able to achieve. In this study, the Gamma Test has been used for the first time as a mathematically nonparametric nonlinear smooth modeling tool to choose the best input combination before calibrating and testing models. Then, several nonlinear models have been developed efficiently with the aid of the Gamma test, including regression models Local Linear Regression (LLR), Dynamic Local Linear Regression (DLLR) and Artificial Neural Networks (ANN) models. We used daily, weekly and monthly spot prices in Henry Hub from Jan 7, 1997 to Mar 20, 2012 for modeling and forecasting. Comparing the results of regression models show that DLLR model yields higher correlation coefficient and lower MSError than LLR and will make steadily better predictions. The calibrated ANN models specify the shorter the period of forecasting, the more accurate results will be. Therefore, the forecasting model of daily spot prices with ANN can interpret a fine view. Moreover, the ANN models have superior performance compared with LLR and DLLR. Although ANN models present a close up view and a high accuracy of natural gas spot price trend forecasting in different timescales, its ability in forecasting price shocks of the market is not notable.
Narges Salehnia, Mohammad Ali Fallahi, Ahmad Seifi, Mohammad Hossein Mahdavi Adeli,
Volume 6, Issue 20 (7-2015)
Abstract
This paper aims at estimating Geometric Brownian Motion (GBM) Model, based on two central parameters in this model (volatility and drift), and forecasting Henry Hub natural gas daily spot prices (07/01/1997-20/03/2012). Researches reveal that two mentioned parameters estimation can be satisfied with different approaches and in various time scales. Therefore, two approaches of backward looking and forward looking have been used in different time scales and sub-periods. Results show that the volatility and drift values are highly dependent on the time scale and backward results are lower than the forward ones. Moreover, along with increasing the number of random runs of the model although the fluctuating range decreases, the predicted line slope is very close to the actual line. Ultimately, the performance evaluation criteria yields that forward method, clearly in 2009, has the best performance. The sub-periods of 2001-2004 in backward and forward methods have the next best performances, respectively. These sub-periods can be used as a basis for calculating the central parameters of the model. In addition, the results suggest that relying on data used in the most recent period is not sufficiently accurate. Also, it is observed that sub-periods or time scales with higher volatility show better performance evaluation criteria, therefore they can be applied in price forecasting with GBM model.
Ebrahim Ghaed, Mahdi Khodaparast Mashhadi, Mohammad Taher Ahmadi Shadmehri, Narges Salehnia,
Volume 15, Issue 56 (8-2024)
Abstract
The main purpose of this study is to predicting the effects of fiscal policies on greenhouse emissions in Iran from 1991 to 2021. To achieve this, bayesian model averaging (BMA) and Bayesian vector autoregression (BVAR) approaches were employed. The results indicate that out of 14 fiscal policy variables, the top five models with the highest posterior probabilities were identified using the aforementioned methods. The most effective models included variables such as financial asset acquisitions, oil revenues, corporate taxes, wealth taxes, current expenditures, and other revenues. Subsequently, the impact of these variables on CO2 emissions was analyzed over 10 periods using the BVAR method. The impulse response function results revealed that shocks to the financial asset acquisitions, oil revenues, wealth taxes, current expenditures, and other revenues had positive effects on CO2 emissions, with the most significant impact stemming from shocks to financial asset acquisitions. Conversely, only shocks to the corporate taxes demonstrated a negative effect. Additionally, the variance decomposition of CO2 emission forecast errors indicated that the oil revenues and wealth taxes played the most significant roles in explaining forecast errors, with their contributions increasing during intermediate periods.
