About us

World's most important data is inaccurate

Author: Dziennik Zwiazkowy
Address: http://dziennikzwiazkowy.com/pacc/najwazniejsze-dane-swiata-nieprecyzyjne/
Published: 18.06.2016

On the first Friday of each month, the financial world anticipates the data from the American labor market. This month's reading was definitely below expectations. This caused some serious consequences in the market. Has the American economy really stopped creating new workplaces in the middle of the year? A comment from Marcin Lipka, Cinkciarz.pl analyst.

Investors focus on the data from the Bureau of Labor Statistics (BLS) every month. This data gives the market much information about the employment in the USA. For example, they show the level of the unemployment rate, the changes in hourly wages and the amount of workplaces that are created by the economy in month on month relations (payrolls).

The situation in the labor market has a crucial meaning for the Federal Reserve. This causes the American monetary policy to significantly depend on the changes of employment in the United States. Moreover, the level of the American interest rates impacts many other instruments as well, such as bonds, shares and currencies. Changes in the USA are immediately translated to the situation in the global markets. This causes the entire world to observe the BLS data.

Most important data is inaccurate

Much economic data is based on surveys. This concerns the payrolls as well. The BLS wrote that the estimations of an increase in the amount of workplaces are based on surveys that are conducted among 146k enterprises. Of course, these are not all companies from the USA. Moreover, each sample group is encumbered with an error.

The Labor Department estimates that the 90% trust range for payrolls is approximately 115k. This goes to show that if the BLS publishes a reading that is at the level of 150k, there is a 90% likelihood that the actual value is between 35k and 265k.

The above matter can be interpreted in another way as well. There is a 10% risk that the actual reading is outside of that range. As a result, approximately each year we receive a reading that is either very high or very low. Moreover, such readings are not at all the results of the economic changes. They are just a result of statistical errors.

There is quite a simple method to make sure whether a significant difference between the reading and the trend is an element of statistics, or an actual deterioration in the labor market. The estimations of changes in the amount of workplaces in the American economy, are published by Automatic Data Processing (ADP) as well.

It is interesting that the sample group used by the ADP is significantly larger than the one used by the BLS (411k vs 146k). This clearly decreases the risk of an error. It is also confirmed by a simple analysis that is based on comparing the two readings, and extracts a few month's median. We will quickly notice that the BLS data is much more often different than the average value of the reading. On the other hand, the median of six or twelve readings is basically the same in case of both surveys.

Mass reaction

The BLS data from May showed that the American economy created only 25k workplaces in the private sector, while the forecasts were at the 150k level. Such a significant difference caused an immediate discussion concerning the delay in raising interest rates by the Federal Reserve. This wore-off the dollar quite quickly, as well as decreased profitability of the American debt.

On the other hand, the ADP data showed that the amount of workplaces increased by 173k in May. The difference between these two readings is very large. When considering the other data from the labor market (a low value of jobless claims, as well as the amount of job offers), we were probably dealing with the above mentioned problem. So, why were investors selling the dollar and reducing their expectations regarding interest rates?

Of course, many of them realize that the payrolls are encumbered with a significant error. However, the market is aware that very weak, as well as exceptionally positive data may cause a mass reaction. Playing against it may bring definitively more losses than gains. This is even taking into consideration that in the end it appears that those more inquisitive investors were right.

Return to the main list


See also: