Abstract—Globally there is a spur in demand of natural gas
and business entities are interested to realize natural gas price
forecast. The forecast is likely to suffice to different objectives
to producers, suppliers, traders and bankers engaged in
natural gas exploration, production, transportation and
trading as well as end users engaged in utilization of natural
gas. For the supplier the objective is to meet the demand with
profit and for the trader it is for doing business. Researchers
have invoked different methodologies to forecast natural gas
price. It was intended to develop a model to forecast natural
gas price in terms of specific parameters which may have
relationship with natural gas price by applying Time Series as
well as nonparametric regression invoking Alternating
Conditional Estimation (ACE). Discernible predictor variables
that may elucidate statistically important amount of the
inconsistencies in the response variable have been recognized
and the correlation between variables has been established.
Statistical tests and techniques viz. unit root and ADF tests,
and ARIMA have been applied. Present work is an attempt to
demonstrate these models pertaining to natural gas pricing