Posted: October 18th, 2016

Forecast the AUD/USD exchange rate at 30 September 2016 using a time-series model.

Introduction
You will be asked to perform two major tasks in this individual assignment. The first task is to analyse trends in Australian dollar. The second task will provide you an opportunity to collect exchange rate and economic data, and use this data to forecast short-term exchange rates. In the context of your results, you will be required to discuss the determinants of exchange rates and issues associated with forecasting exchange rates.

IMPORTANT INFORMATION
 The assignment must be attempted and completed individually
 Use the university cover page when submitting reports
 Any assignment submitted late will be penalised at a rate of 10% per day
 Maximum word limit is 3,000 words in total to be written using Microsoft MS word. The table of contents and reference section are not included in the word count.
 To provide the supporting data for your report you will be required to record any calculation using Microsoft Excel. Use Excel formulas in your calculations so that the final answers can be justified and reconciled by reference to the formula in the cell. (Your Microsoft Excel spreadsheet needs to be prepared in a way that a colleague could open your spreadsheet and justify your calculations.) The Excel work file is to be sent to
 Upload the report through turnitin and provide a hard copy at the lecture on the due date. Please make sure your tutor’s name appears on the cover sheet.
 Make and state any assumptions if the given information is insufficient to justify your answer.

Assignment Questions
1) Broad trends in the Australian Dollar
Evaluate trends in the Australian dollar / United States dollar (AUD/USD) exchange rate and the trade-weighted index (TWI) over the period from July 2001 through June 2016 (i.e. 15 years). Discuss whether the trends are consistent with theory.

2) Exchange rate prediction
a) Building on the insight you gained from Part 1, formulate a regression model for quarterly changes in the AUD/USD. Use this model to forecast the exchange rate at 30 September 2016. The prediction should be made at the end of June (i.e. using data at the end of June). This means that all variables are assumed to be available up to 30th of June. Your dependent variable will be quarterly changes in the AUD/USD exchange rate. Potential explanatory variables include differences in interest rates, differences in unemployment rates, and differences in inflation rates. Explain any issues you faced in selecting the data.
b) Forecast the AUD/USD exchange rate at 30 September 2016 using a time-series model.
c) Discuss how your forecasts compare with actual exchange rates at 30 September 2016? Does it outperform a random walk? Which is the most accurate model? Reflect on your forecasting models by evaluating the performance of the two prediction models in the context of the theory of exchange rate determinants.