Economics and Business Forecasting
Project description
Question
The file report2015.dta contains data on three (3) timeseries, namely
1. U.S. Imports of Goods from the United Kingdom, Customs Basis: Millions
of Dollars: NSA
2. U.S. Exports of Goods to China, Mainland, f.a.s. basis: Millions of Dol- lars: NSA
3. Harmonized Index of Consumer Prices: Passenger Transport by Air for United Kingdom: Index 2005=100: NSA
labelled appropriately. Your task is to produce five (5) year forecasts of each series using four (4) different modelling approaches for each one. In particular, each timeseries should be forecasted using:
1. classical timeseries decomposition techniques,
2. a simple linear regression model where you account for possible time trends,
3. a multiple regression model, and
4. an appropriate ARIMA/ARMAX model.
You should evaluate and compare forecasts from each modelling approach with each other and conclude which is the best for each series. Comparison and/or evaluation of forecasts from each approach should not be limited to statistical tests but also include your own judgement informed by economic theory and prevailing economic conditions. Note: you are neither required nor expected to obtain additional data for any of the above tasks.
———-
Added on 08.04.2015 15:54
Please produce a stata .do file, NOT a log file.
The main textbook for the module is:
J. Holton Wilson, Barry Keating and John Galt Solutions, Inc. Business Fore- casting (with ForecastX), 6th Edition, McGraw-Hill International Edition.
http://www.sendspace.com and will share a link with you