Beryl Gemstone

COURSEWORK ASSIGNMENT FOR G26 – FEBRUARY 2019

The following THREE files contain data for your coursework assignment.  

Each of you will have been assigned to a dataset group.  Dataset 1 (Green), Dataset 2 (Blue) and Dataset 3 (Yellow).

The file Country and Currency List_2018.xls provides the names and currencies of the countries in each of the datasets. The currency codes as well as country and currency name are provided in this file.

The file CPI DATA 2018.xls contains monthly data on consumer price indices (CPI) for each of the five countries in the three different groups from January 2000 until December 2016 (ignore the precise date of the month).  Refer to the currency codes in the files up to identify the respective currencies in this Excel file.  The three datasets have also been colour coded. These are the series that are referred to as [pic 1]below.  The base year for each series is January 2000 = 100.

This file also contains CPI data for the United States (dollars) – referred to as US - for the same time span and with the same base year.  This is the series that is referred to as [pic 2]below and is to be interpreted as the numeraire country – i.e. the foreign country with respect to which deviations in prices are being studied.

The file Spot Rate in terms of USD_2018.xls contains the spot exchange rate of the currency of each of the five countries in the three different groups with respect to the United States dollar.

The real exchange rate for any country i with respect to a foreign country (the US in this assignment) is defined as:

[pic 3]

where [pic 4]denotes the spot exchange of the currency of country i with respect to the numeraire currency.

Focus only on the five countries assigned to you in your dataset, and undertake exercises i) – vii) below for each of your five countries.

1. For U.S. as the numeraire currency, express the real exchange rate relationship in logarithms (denoting lower case letters to denoted logs, so [pic 5]etc. ) Begin by graphing the series and test the unit root hypothesis for [pic 6] over the full sample of observations using the Augmented Dickey Fuller (ADF) test.  [18%]

1. Repeat the exercise in (i) using the Phillips-Perron (PP) test for a unit root.  Both for the ADF and PP tests you should make clear your choices concerning lag length selection and of the deterministic terms in your regression. [18%]

1. How are your findings altered if you use a test with stationarity as the null hypothesis – i.e. if you used, for example, the KPSS test? [18%]

1. Test the integration/stationarity property of the series [pic 7]and assess if purchasing power parity (PPP) holds over the full sample period studied (using ADF, PP and KPSS tests) -   i.e. [pic 8] as defined above is I(0) .  Now repeat the tests for [pic 9]in the period following the financial crash (i.e. starting your data from January 2009).  Assess the impact of the financial crash. [18%]

1. How could you formulate a test for PPP in (iv) as a cointegration test, instead of a test on[pic 10]? For your formulation, use a simple cointegration test to test if PPP holds both over the full sample as well as for the period starting January 2009.  How do your results compare with your findings in iv)? You may assume here that each of the series used in the cointegration test is I(1). [18%]

1. Finally discuss briefly how your results above might be affected if the spot rate series for each country (including the numeraire countries) is expressed as an index – so that the spot rate for 1 January 2000 for each country is assigned a value of 100 and all subsequent values of this spot rate for each country are calculated relative to the value of the spot rate for 1 January 2000.  [10%]

[You should not attempt any calculations, only think about how the answers may or may not change.]