3
$\begingroup$

I am reading a paper by Michal Andrle, Jan Brůha, Serhat Solmaz (On the sources of business cycles: implications for DSGE models). The authors use data at a quarterly frequency for their empirical analysis. But they also have this statement in their paper, 'Our empirical approach boils down to a multi-country dynamic principal component analysis of data at business cycle frequencies'.

What does the term 'business cycle frequency' mean? And what does it mean to analyze data at business cycle frequency? I have yet found a simple definition for it.

Improve this question
$\endgroup$
1
  • $\begingroup$ Yeah, sure. I have edited my post. Thanks!! $\endgroup$
    – Emmanuel Ameyaw
    Feb 25, 2021 at 14:24

3 Answers 3

Reset to default
1
$\begingroup$

Business cycle frequency is a frequency that has the length of one business cycle. Business cycle is:

economy-wide fluctuations in production, trade, and general economic activity. From a conceptual perspective, the business cycle is the upward and downward movements of levels of GDP (gross domestic product) and refers to the period of expansions and contractions in the level of economic activities (business fluctuations) around a long-term growth trend.

The length of the business cycle varies but usually, it is at least a few years.

Improve this answer
$\endgroup$
7
  • $\begingroup$ If I get it right, it is usually a few years irrespective of whether one uses quarterly, monthly or annual data? $\endgroup$
    – Emmanuel Ameyaw
    Feb 25, 2021 at 14:27
  • $\begingroup$ Yes, the sampling frequency of the data - monthly, quarterly, annual - does not matter. The problem with annual data is that it might be low a frequency, and you might miss recessions. As for the length, one might note that business cycles in the developed countries are longer-lived since the 1990s (recessions a decade apart). $\endgroup$
    – Brian Romanchuk
    Feb 25, 2021 at 18:48
  • $\begingroup$ @EmmanuelAmeyaw yes it is several years no matter what data are available $\endgroup$
    – csilvia
    Feb 26, 2021 at 9:08
  • $\begingroup$ Many thanks. So to 'analyze data at business cycle frequency' as used in research papers don't actually change anything if it is omitted. Whether data is annual, monthly, or quarterly, you can still say we analyze data at business cycle frequency, right? Anyway, I don't really know what that statement means. $\endgroup$
    – Emmanuel Ameyaw
    Mar 1, 2021 at 8:26
  • $\begingroup$ @EmmanuelAmeyaw analyzing data at business cycle frequency means that no matter what data you have your time unit will be 1 business cycle. For example, even if you use daily data you can analyze the data on quarterly frequency by aggregating daily observations across a quarter. The same goes for business cycle frequency. Frequency is not about what data you have but what is the length of t used in the study $\endgroup$
    – csilvia
    Mar 1, 2021 at 9:13
1
$\begingroup$

Number of peaks in GDP(gross domestic product) per year

Improve this answer
$\endgroup$
0
$\begingroup$

The consensus is to define the business cycle frequency to be between 1.5 and 8 years (see Christiano and Fitzgerald (2003)).

Intuitively, analyzing the data at the business cycle frequency means looking at the variation in the data that happens at the frequency mentioned above.

Practically, this means that you first filter your data (e.g. using a band-pass filter) and then conduct your analysis on the filtered data.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.