Fundamental analysis Macroeconomic indicators

Retail Sales Figures and Their Influence on Exchange Rates

Elena Berseneva 12 january 2022 212 3

Growth of retail sales figures reflects a good state of customer demand and positive dynamics of economic development of the country. In this research, we will examine the relationship between retail sales changes and the related changes in exchange rates of national currencies.


First, we let us provide a brief definition:


Retail Sales are one of the macroeconomic indicators of the state of the national economy that reflects the retail sales figures of goods and services to the final buyers for personal use. The indicator reflects sales by a personal contact, trading networks, e-mail, the Internet, cell phones, etc.


We should note that significant changes in channels and subjects of retail trade occur over time due to the advancements in modern technologies.

Hypothesis
To conclusion

Growth of retail sales figures causes strengthening of exchange rates of the national currency.

To conclusion

The following liaison mechanism is expected:

The growth of sales figures indicates prospects of investments for the economy of a given country, and, therefore, increases the demand for its currency that leads to strengthening of its rate.


So, we will test this hypothesis with different methods relying on actual data of ten greatest economies of the world for the last 10 years.

Data used

For the analysis, we will use monthly reports of changes of retail sales of Australia, Brazil, the USA, the United Kingdom, Italy, Germany, Canada, China, South Korea, and Japan. The period under consideration is 10 years (2009-2019).

In the calculation, 120 indicators of monthly statistics of every country are used.


These countries were selected for the analysis because the majority of transactions in Forex trading are carried out with currency pairs that incorporate the currencies of listed countries.


To estimate the changes of currency rates, we will use historical values of daily quotes for opening prices for a ten-year period. In total, the dynamics for 50 currency pairs will be analyzed.


The total size of the sample is 132 125 values of daily quotes of currency pairs.


Data sources.


Country
Currency
Data Source
Source website
The USA
The US dollar(USD)
U.S. Census Bureau
https://www.commerce.gov/bureaus-and-offices/census
China
Chinese yuan(CNY)
National Bureau of Statistics of China
http://www.stats.gov.cn/
Japan
Japanese yen(JPY)
Minister of Economy, Trade and Industry
http://www.meti.go.jp/
Germany
Euro(EUR)
TheFederalStatisticalOffice
https://www.destatis.de/
The United Kingdom
Pound sterling(GBP)
Statistics: Calendar of Releases
https://www.gov.uk/
Italy
Euro(EUR)
Italian National Institute of Statistics
http://www.istat.it/
Brazil
Brazilian Real (BRL)
Brazilian Institute of Geography and Statistics
http://www.ibge.gov.br/
Canada
Canadian dollar(CAD)
Statistics Canada
https://www.statcan.gc.ca/
South Korea
South Korean won (KRW)
Statistics Korea
http://kostat.go.kr/
Australia
Australian dollar (AUD)
Australian Bureau of Statistics
https://www.abs.gov.au/



Definition of Volatility.


Before we move on to the assessment of the direct impact of publication of retail sales data on the direction of currency rate changes, we will consider the influence of new publications of retail sales on volatility of the Forex market.


For this reason, we suggest comparing the values of daily volatility of currency pairs on the day of data publication with the values of average daily volatility for the preceding week and for three weeks. For the analysis, we will use 5 569 news releases by the data of retail trade for 10 years in 10 countries.



Comparison of a daily volatility, %


Country
On a day of publication
Average for the preceding week
Average for three preceding weeks
Australia
1.216
1.127
1.115
Brazil
1.601
1.537
1.521
The United Kingdom
1.037
1.002
0.994
Germany
1.061
0.983
0.984
Italy
0.994
0.971
0.993
Canada
1.071
0.992
1.014
China
0.858
0.803
0.790
South Korea
0.965
0.926
0.957
The USA
0.991
0.951
0.930
Japan
1.185
1.133
1.171
Average value
1.098
1.042
1.047


From the table values, we can that on the day of data publication the volatility is 5% higher compared to average volatility of the preceding week period.



We will now move straight into the next stage of testing the hypothesis that growth of a country’s retail sales results in strengthening of its currency.


This research can be carried out with two independent methods:

  • Through assessing the correlation;
  • Through assessing the average probability receiving a positive return.


The calculation of correlation coefficients is carried out on the following timeframes of observation:

  • one week after publication of the daily data on retail trade;
  • three weeks after publication of the daily data on retail trade;



Correlation, depending on the period of observation.


Country
One week
Three weeks
Australia
0.023
0.008
Brazil
-0.245
-0.235
The United Kingdom
-0.014
-0.040
Germany
-0.051
-0.079
Italy
0.023
-0.008
Canada
0.112
0.122
China
0.064
0.012
South Korea
0.035
0.055
The USA
0.047
0.015
Japan
0.052
-0.031
Average value
0.005
-0.018


As can be noticed from the obtained results, there is a lack of correlation dependence in all countries (except for a weak relation with Brazil), while both positive and negative relations are present in the table.


The next stage involves calculation of similar correlation coefficients for five conditional groups of the rate of retail trade’s change, namely:

  • Large decrease of sales figures;
  • Moderate decrease;
  • Minor decrease;
  • Moderate increase;
  • Large increase.


Let us take the same number of values for all groups. Since the considered sample includes 5 569 values, every group comprises 1114 (1113) values.



Correlation, depending on the rate of sales figures.


Number of values in the interval
Intervals of changes of sales figures, %
Conditional assessment
Correlation
(week)
Correlation(three weeks)
1 113
от -9,7 до -4,0
Large decrease
0,011
0,011
1 114
от -4,0 до 0,0
Moderate decrease
-0,010
-0,076
1 114
от 0,0 до 0,5
Minor increase
-0,012
-0,012
1 114
от 0,5 до 1,4
Moderate increase
-0,057
-0,057
1 114
от 1,4 до 22,1
Large increase
0,038
0,048


As indicated in the diagram, there are no any clear correlations.



As has already been mentioned, the second method of correlation determination between changes of retail sales and changes of currency rates involves calculating the average probability of profit-making.


Calculations will be carried out with a test using the historical data of currency pair quotes.



Description of the testing algorithm:

If a published indicator of retail sales is again higher than the preceding one, we place an open-to-buy position of its relevant national currency with respect to other currencies. If the indicator of retail sales is lower than the preceding one, it is the opposite, we open a short position for the country’s currency. For example, if U.K. retail sales demonstrate an increase, we open buys for GBPUSD, GBPCAD, GBPAUD, GBPJPY and open a sale of EURGBP.


Positions are opened with the opening price of a daily candlestick on the next day following data publication. Positions are closed similarly at the opening price of a daily candlestick upon the expiry of specified time periods for position hold.


In total, the result of 6000 positions will be analyzed.


In this method, the magnitude of difference of retail sales values is not considered, nor is the volatility of currency pairs caused by data publication. Instead, we assess only the number of positions closed with a profit or a loss.


The obtained coefficients demonstrate the probability (in %) of closing the position with a profit, depending on the time its hold time. The average value on all currency pairs indicates that the numbers of profitable and loss-making transactions are almost the same. Nevertheless, a slight prevalence of non-profit transactions is indicated.



Share of profitable transactions, depending on a period of observation, %.


Country
One week
Three weeks
Australia
50
49
Brazil
49
45
The United Kingdom
47
48
Germany
50
50
Italy
52
52
Canada
48
45
China
49
51
South Korea
51
50
The USA
47
48
Japan
51
44
Average value
49
48



Probability of profit-making.


Country
For a week
For three weeks
U.S. Retail Sales
47
48
China Retail Sales
49
51
Japan Retail Sales
51
44
Germany Retail Sales
50
50
U.K. Core Retail Sales
47
48
Italy Retail Sales
52
52
Brazil Retail Sales
49
45
Canada Core Retail Sales
48
45
South Korea Retail Sales
51
50
Australia Retail Sales
50
49



Assessment of volatility in a day of publication.


Volatility
Value, %
Interpretation
U.S. Retail Sales
4,22
Weak increase in volatility
China Retail Sales
6,78
Weak increase in volatility
Japan Retail Sales
4,59
Weak increase in volatility
Germany Retail Sales
8,01
Weak increase in volatility
U.K. Core Retail Sales
3,54
Weak increase in volatility
Italy Retail Sales
2,40
Lack of influence on volatility
Brazil Retail Sales
4,18
Weak increase in volatility
Canada Core Retail Sales
7,90
Weak increase in volatility
South Korea Retail Sales
4,18
Weak increase in volatility
Australia Retail Sales
7,93
Weak increase in volatility
Conclusion

Correlation and probability of profit-making:

The considered indicator of retail sales has no correlation with changes of national currency rates.


Volatility:

On the day of data publication on retail sales, daily volatility of all considered currency rates demonstrates a weak increase (except Italy, where a lack of increase in volatility of the Euro is indicated on the day of publication).

Nevertheless, as mentioned before, an increase in volatility does not lead to a significant change of currency rates during one-week- or three-week periods.


The influence of publications of “Retail Sales” data on the direction of currency rates has not been identified.

XLSX (0.02 MB)Retail Sales.xlsx

Comments

3

Write the comment

Commenting rules
Only authorized users can leave comments. Sign in, please.
Only users with verified email can leave comments. To verify your email, click on the link in the message that has been sent to your email address . Send an email for activation again.

Subscribe to our newsletter and stay up to date with all the news!