Research Paper using ForecastX and forecasting techniques

BUSI 405

Research Paper Instructions

1. Define the chosen variable

· New one-family houses sold (NHS)

2. Collect the data for the variable

· Data source

· www.economagic.com or other sources

· Monthly data and not seasonally adjusted ( NSA )

· Time period – January 1975 to the present

3. Enter the data in Excel according to ForecastX format

· 1/1/1975 for Jan. 1975 and 2/1/1975 for Feb. 1975

4. Plot the graph for whole series

· ForecastX -- Preview

5. ACF for the whole series

· ForecastX – Analyze

· Identify the trend

6. ACF for the first differenced series

· ForecastX – Analyze – Differencing – Non-Seasonal >1

· Identify seasonality

7. Decide the sample period for model selection

· Historical period: 1/1/1975-12/1/2010

· Holdout period: 1/1/2011-12/1/2011

8. Select models according to the data pattern

· Table 2.1, p. 58

· Time-series models: Modified naïve model, Winters’ exponential smoothing, Time-series Decomposition, and ARIMA

· Regression mode:

· NHS = f ( IR , DPI , dummy variables)—need data for IR (30-year mortgage rate), DPI(disposable personal income) and dummy variables (for seasonality)

9. Perform estimation and forecasting for the data in the historical period, 1/2001 – 12/2010

· ForecastX – Forecast Method – choose models

· ForecastX – Statistics -- RMSE

· For regression model, standard report shows the forecasts for independent variables

10. Compare MAPEs and RMSEs of different models for the historical and holdout periods

11. Perform ex-ante forecast – 11/2011-12/2012

· Use the whole series, 1/2001 – 10/2011

· Either choose the best model or combine two models