WeeklyReturnsData: Logarithmic Returns of Exchange Rates of 7 Currencies Against...

Description Usage Format Details Source References Examples

Description

This data set has logarithmic returns of exchange rates of 7 currencies against US dollar extracted from the International Monetary Fund's (IMF) database. These currencies are Australian Dollar (AUD), Canadian Dollar (CAD), Norwegian Kroner (NOK), Singapore Dollar (SGD), Swedish Kroner (SEK), Swiss Franc (CHF) and British Pound (GBP).

Usage

1
data("WeeklyReturnsData")

Format

An object of class ts with 605 observations on the following 7 variables.

AUD

The weekly logarithmic returns r_(AUD,t) of the exchange rates of AUD against US Dollar.

CAD

The weekly logarithmic returns r_(CAD,t) of the exchange rates of CAD against US Dollar.

NOK

The weekly logarithmic returns r_(NOK,t) of the exchange rates of NOK against US Dollar.

SGD

The weekly logarithmic returns r_(SGD,t) of the exchange rates of SGD against US Dollar.

SEK

The weekly logarithmic returns r_(SEK,t) of the exchange rates of SEK against US Dollar.

CHF

The weekly logarithmic returns r_(CHF,t) of the exchange rates of CHF against US Dollar.

GBP

The weekly logarithmic returns r_(GBP,t) of the exchange rates of GBP against US Dollar.

Details

The daily exhange rates of the currencies against US Dollar from March 22, 2000 to October 26, 2011 are extracted from the International Monetary Fund's (IMF) Exchange Rates database from https://www.imf.org/external/np/fin/ert/GUI/Pages/CountryDataBase.aspx. These rates are representative rates (currency units per US Dollar), which are reported daily to the IMF by the issuing central bank.

The weekly averages of these exchange rates are then calculated. The logarithmic returns of the average weekly exchange rates are calculated for the currency j as follows.

Let x_(j,t) be the exchange rates of j against US Dollar. Then

r_(j,t) = log x_(j, t)) - log(x_(j,t-1)),

where t = 1, …, 605 and j = AUD, CAD, NOK, SGD, SEK, CHF, GBP. The six missing values in r_(SEK,t) are changed to 0. The assumption used here is that there has not been any change from the previous week.

The weekly returns data is then changed to a multivariate time series object of class ts. The resulting ts object is then dataset WeeklyReturnsData.

An example analysis of the data is given in Miettinen et al. (2018). Same data has also been used in Hu and Tsay (2014).

Source

International Monetary Fund (2017), IMF Exchange Rates, https://www.imf.org/external/np/fin/ert/GUI/Pages/CountryDataBase.aspx

For IMF Copyrights and Usage, and special terms and conditions pertaining to the use of IMF data, see https://www.imf.org/external/terms.htm

References

Miettinen, M., Matilainen, M., Nordhausen, K. and Taskinen, S. (2020), Extracting Conditionally Heteroskedastic Components Using Independent Component Analysis, Journal of Time Series Analysis,41, 293–311.

Hu and Tsay (2014), Principal Volatility Component Analysis, Journal of Business & Economic Statistics, 32(2), 153–164.

Examples

1
2
3
4
5
6
7
8
plot(WeeklyReturnsData)

res <- gSOBI(WeeklyReturnsData)
res

coef(res)
plot(res)
head(bss.components(res))

tsBSS documentation built on July 10, 2021, 9:07 a.m.