EC306 Econometrics II: Time Series
2024 Vacation Assignment
Data: The data ile contains U.K. observations on the variables below. The irst observation is Jan 1955 in both
Table 1: Data set
Quarterly Monthlz
Nominal GDP sa CPI All Items: Annual Rate
Unit Labour Cost Index sa Unemployment rate sa
Unemployment rate sa Average weekly earnings sa
GDP Delator sa Index of Production cvmsa
quarterly and monthly data sets.
1. This is a question about forecasting GDP:
(a) Estimate a set of four arima models for Real GDP 1960q1-2000q4. Include at least one model which assumes GDP is a trend stationary process, and three plausible arima(p,d,q) models, including examples with the restrictions: i) q=0 and ii) p=0. (10 marks)
(b) Basing your discussion on the in-sample properties of your model, why did you select the models you chose in each class, and which of your four models appears to be the best? (5 marks)
(c) Using the estimates from part (a), evaluate the out of sample forecasting performance of each model on the sample 2000q1-2006q4, at horizons of one and four quarters. Comment on which is your best model at each horizon and how these results compare to those in (b) above. (15 marks)
(d) Re-estimate your preferred model from part (c) on the period 1960q1-2006q4. Forecast out of sample at horizons of one and four quarters for the period 2007q1 to 2013q4. Comment on the quality of the forecast compared to your expectations.(10 marks)
(e) How do you think the performance of your trend stationary model in part (a) would compare to the performance of your preferred model on the 2007-2013 sample? Give a brief explanation, or explain why the trend stationary model is your favourite. You do not have to produce forecast results. (10 marks)
2. This question is forecasting in a multivariate context:
(a) Estimate a VAR using additional data of your choice on the 1960q1-2006q4 sample. Compare the 1 and 4-step ahead out of sample forecast performance of your VAR on the 2007-2013 sample with your preferred forecast from question 1. Provide some intuition for your results. (10 marks)
(b) Consider the task of updating quarterly GDP forecasts as new information becomes available within the quarter. Demonstrate a method capable of doing this and present results of the exercise graphi- cally for next quarter GDP using a model estimated to 2006q4, and forecasting 2007q1-2013q4. (15 marks)
3. Consider a model of price inlation, wage inlation and unemployment, on the longest sample for which all are available in the data set.
(a) Estimate a model to account for the interactions between these variables, accounting for the time- series properties of the data on this sample. Discuss briely any economic interpretation/indings or intuition you ind from your estimated model. (15 marks)
(b) Comment on the role of unemployment in your model, and any economic implications of this. (10 marks)
Generic guidance and notes for non economists:
. The inlation measure provided is the year-on-year monthly change in the Consumer Price Index (CPI).
. sa stands for Seasonally Adjusted and cvmsa is a chained volume measure seasonally adjusted
. You should inspect data by plotting it; it is expected that you make basic transformations to ensure you work with variables suitable for linear models
. Graphs can be included in an answer e.g. if inspection of the graph helps you decide how to make modelling decisions. Unless otherwise stated in the question, graphical analysis alone is not su伍cient.
. When you report results of a statistical test, state the test regression, the parameter(s) of interest, the null hypothesis, the test statistic, the value of the test statistic you ind, the critical value and the decision you make. Bullet points are ine for such discussion
. Try to summarize the important parts of the regression output in a table, rather than screen-grabbing whole chunks from the raw stata output; this makes you think about the output and demonstrates understanding of what is important.
版权所有:编程辅导网 2021 All Rights Reserved 联系方式:QQ:99515681 微信:codinghelp 电子信箱:99515681@qq.com
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。