Page 1 of 15
Advances in Social Sciences Research Journal – Vol. 8, No. 4
Publication Date: April 25, 2021
DOI:10.14738/assrj.84.10143.
Almutairi, E. N., & Almutairi, H. N. (2021). Purchasing Power Parity (France and US). Advances in Social Sciences Research Journal,
8(4). 514-528.
Services for Science and Education – United Kingdom
Purchasing Power Parity (France and US)
Ebtihal N. Almutairi
School of Business, Nottingham University, UK
Hind N. Almutairi
School of Business, Bangor University, UK
ABSTRACT
Purchasing power parity is an important economic concept that presents a useful
way of analyzing various currencies by comparing a similar basket of goods.
Essentially, the concept of purchasing power parity (PPP) implies that the
purchasing power parity will be same for two countries if their currencies fetch the
same basket of goods, thereby allowing a measure to assess and compare different
currencies. In order to test this concept and presence of purchasing power parity
between countries, this essay aims to extend a statistical exercise for assessing the
PPP for France and USA for both the long and short run. The data used is a monthly
time series for the price indexes for France, US, and the nominal and real exchange
rates.
Keywords: Purchasing power parity; economic; Augmented Dickey Fuller; Fuller test;
Vector autoregressive model.
INTRODUCTION
Existing literature presents a varied examination of purchasing power as monetary concept
that carries practical relevance. The link with income and purchasing power parity for
analyzing consumption patterns has been documented for the global poor (1). In terms of
assessing the validity of the concept, researchers have deployed conventional and
unconventional statistical methods to explore the long run validity of PPP in developing
countries using a time series approach (2). The interlinkage between PPP and the exchange
rates has also been the focus of a strand of relevant research. Interestingly, researchers find
that if PPP hold true continuously, nominal exchange rates will not have an impact on trade
flows (3). However, if PPP holds true for the long run, the real exchange rates might be effected
by monetary fluctuations (3). In addition to this, recent researchers also examine the PPP
theory for large and developing economies like the US and Japan. Pilbeam and Litsios (2018)
explore the PPP theory for the US Dollar and Japanese Yen to discover that PPP theory explains
changes in the nominal exchange rates for the long run. While this literature review
encapsulates the relevance and contributions of recent research of the validity and monetary
applicability of the PPP theory, there is a need to validate the theory in different contexts. In
order to extend the analysis, this report will examine the validity of the PPP theory using time
series data on France and US.
Page 2 of 15
515
Almutairi, E. N., & Almutairi, H. N. (2021). Purchasing Power Parity (France and US). Advances in Social Sciences Research Journal, 8(4). 514-527.
URL: http://dx.doi.org/10.14738/assrj.84.10143
STATISTICAL ANALYSIS OF TIME SERIES US-FRANCE
Prior to beginning with the time series analysis, the date format is set and the ttset command is
used to set the time series in STATA. Natural log series are created, as shown in the do file. In
order to carry out further tests and analysis, it is crucial to determine the stationarity or non- stationarity of the data. The Dickey-Fuller or Augmented Dickey-Fuller unit root test is applied
to check the stationarity of the series. However, it is imperative to identify the minimum Akaike
Information Criterion (AIC) for best fit, optimum variation and appropriate lag length.
Akaike Information Criterion (AIC) (in level form) to identify optimum lag length
(i) AIC value, in level form, shows that lag 1 is optimum for France’s price index
(ii) AIC value, in level form, shows that lag 1 is optimum for price index for US
(iii) AIC value, in level form, shows that lag 1 is optimum for Nominal Exchange rate (s)
(iv) AIC value, in level form, shows that lag 5 is optimum for price index of Real exchange
rate
Augmented Dicky Fuller Test (ADF) in level form
After the lag transformations of all the variables and based on the AIC criterion in levels, the
Augmented Dicky Fuller Test is performed. The results of the ADF test are displayed in Table 1
below. Since all t calculated values are less that the critical t tabulated value, we fail to reject
the null hypothesis of non-stationarity or random walk. This proves that all these variables are
non-stationary and have a random walk with a possibility of drift in the level form analysis.
Table 1: ADF Unit Root Test Results for times series in levels
Variables (in log form) t-calculated t-tabulated
P_france -1.438 -3.439
P_Us -2.037 -3.439
s -2.778 -3.439
qt -2.851 -3.440
Akaike Information Criterion (AIC) (in first difference form) to identify optimum lag
length
(i) AIC in first difference, suggests lag 1 to be optimum for Price index for France
(ii) AIC in first difference, suggests lag 1 to be optimum for Price index for US
(iii) AIC in first difference, suggests lag 1 to be optimum for Price index for Nominal
Exchange rate
(iv) AIC in first difference, suggests lag 4 to be optimum for Price index for Price Index of
Real Exchange rate.
Augmented Dicky Fuller Test (ADF) in first difference form
The ADF test is performed on the P_france, P_Us, s and qt time series using the selected
lags. The results of the unit root test are presented in the table below. All t calculated values are
greater than t-tabulated values hence we reject the null hypothesis of non-stationarity and