Introduction,
On the first Friday after the month ends, two important news release by Bureau of Labor Statistics. One is unemployment rate and the other one (the more or perhaps the most important one) is non-farm payrolls (NFP). NFP accounts for a majority of overall economic activity and show changes in the number of employed people during the previous month, excluding the farming industry. Unemployment rate is an important signal of overall economic health and it is Percentage of the total work force that is unemployed. In this article, I try to show the influence of these indicator on USD index return based on portfolio analysis.

Portfolio analysis,
portfolio analysis means that NFP & unemployment rate data are sorted into 10 portfolios from lowest to highest separately and USD index returns of each portfolio are examined for evidence of anomaly;. Portfolio analysis intuitively, clearly reflects the picture of how returns vary with the characteristic variable; however, with this method it is difficult to conduct multivariate tests and difficult to test the functional form. In the picture below, basket is NFP or unemployment rate and eggs are USD index return.

In this research, the average daily return of USD index is calculated from the time when news release until next release and the period is from April 2nd, 2010 to February 3rd, 2017 and for statistic calculation Microsoft Excel and SPSS software were used. Analysis for NFP Table below present that USD return decreases from 0.07% for low NFP (-155.33%) to -0.05% for high NFP (270.48%).

Based on table, we can see there is negative relation between NFP and USD index return. This issue is clearer in chart below (consider linear).

Table below shows data is normal by One-Sample Kolmogorov-Smirnov Test:

When data is normal for understanding significant of difference between portfolio 1 and 10 we use Student's t-test and table below shows that difference between portfolio 1 and 10 is significant at 94% level. And there is a negative relation between NFP and USD index return.

Furthermore, the Pearson correlation coefficient between NFP portfolio and USD index return portfolio is -65% which shows there is moderately strong negative relation. The correlation of -65% would be significantly greater based on 10 data (portfolio) instead of (for example) 100 data. Analysis for unemployment rate As it could be seen in table below, both portfolio 1 and 10 for USD return have 0.01% return and there is not any differences and there is no relation so further statistic tests are not required.

Also, the Pearson correlation coefficient between unemployment rate portfolio and USD index return portfolio is 2.8% and there is very weak positive relation (or no relation).

Conclusion,
As mentioned above, the unemployment rate is not important but NFP is significantly effective. As a strategy whenever there is a sharp decrease or increase in NFP we can decide buy or sell. Sharpe positive growth means the USD index will decrease until next news release and sharp negative growth means there will be an increase on USD. Be careful it is not for all positive or negative it is mostly for sharp and significant move. In table below I mixed NFP and unemployment rate so we can see when NFP growth is negative and unemployment rate is high (positive) the portfolio has a good USD index return. Finally, further analysis for NFP is required, for example considering forecast is important.

Translate to English Show original