BRYBOSCHAN (Bry-Boschan Business Cycle Dating)

[TomDoan]

The attached is an implementation of the Bry-Boschan business cycle dating algorithm. For quarterly data, it implements the Pagan-Harding(2002) version.

Bry and Boschan (1971). "Cyclical Analysis of Time Series: Selected Procedures and Computer Programs", NBER, New York.
Pagan and Harding (2002) "Dissecting the cycle: a methodological investigation", Journal of Monetary Economics, Volume 49, Issue 2, 365-381.

bryboschan.src

Detailed description

Updated 26 April 2017 to include UP and DOWN options for the ranges of expansions and recessions.


Examples

These examples are based upon Mark Watson(1994), "Business Cycle Durations and Postwar Stabilization of the U.S. Economy", American Economic Review, vol 84, no 1, 24-46. The monthly example is from the paper itself, while the quarterly uses the same data set, but is compacted to quarterly.

bryboschan.rpf

Example for monthly data

(463 Bytes) Downloaded 934 times

bryboschanquarterly.rpf

Example for quarterly data

(563 Bytes) Downloaded 718 times

watsonaer.rat

Data file

(190.5 KiB) Downloaded 646 times

[m joaorodrigues]

Hi there,

Just wondering if there are any rats codes that calculate the bry-boschan algorithm but for deviation cycles (eg kalman filter, band-pass, etc.) instead of the usual classical cycle.

Artis, Michael J., Massimiliano Marcellino and Tommaso Proietti (2003), “Dating the Euro Area Business Cycle”, CEPR Discussion Paper no. 3696, devised something of that kind!

Thanks,
Best regards

[zulkas]

Good evening,

I´d like to apologize about my english. I´m from Spain. I´m trying to use winrats to can run the bry boschan algorhitm, but i have no idea how can i do it. I have downloaded the file bryboschan.src but i don´t know what to do with it, i even don´t know how can i import the dates to winrats in a proper format. I have an excel file with the dates in the format 30/12/2011 and with a lot of columns with labels in the first cell and numbers in the rest of them. Even if i was able to import this file to winrats, then what? how could i run the bry boschan algorhitm to my variables? please i really need help.

i attach you the excel file

[moderator]

The first step is of course to get the data read in. If you have no idea how to do that, please start by working through at least the first few sections of the "Introduction to RATS" book included with the software. That should tell you everything you need to know about reading in data.

In terms of applying the Bry Boschan algorithm, that is available in RATS via a pre-written "procedure" included with RATS called @BryBoschan.

Assuming you are new to using procedures in RATS, please see pages 32-33 of the Introduction for some general information on using procedures. You can find more on using procedures on pages 451-452 of the "User's Guide".

Next, open up the BryBoschan.src file (which contains the code for the procedure). You will find comment lines at the top of the file describing the syntax (options, parameters, etc.) and other details of the procedure.

I hope this will help you get started.

[zulkas]

I have created an excel with clear turning points to try to understand how the algorithm works. i´m using these commands:

CAL 2005 6 12
ALL 2011 12
OPEN DATA ADFDF..XLS
DATA(FORMAT=XLS,ORG=COLS)
CALENDAR(M) 2005:6

@BRYBOSCHAN PEAK=2007:11 TROUGH=2005:10 PRUEBA 2005:6 2011:12

i don´t know what i´m doing wrong, can anyone explain me which commands do i have to use to obtain the results i´m looking for?

[TomDoan]

There are examples of the use of BryBoschan above.

The PEAK and TROUGH options are for output, not input. Bry-Boschan is designed to locate the business cycle peaks and troughs, so you don't want to input them. The most basic use with your data would be:

@BRYBOSCHAN PRUEBA 2005:6 2011:12

If you want to save dummy series with the peaks and troughs, you would do that with

@BRYBOSCHAN(PEAKS=PEAKS,TROUGHS=TROUGHS) PRUEBA 2005:6 2011:12

[hardmann]

Dear Tom:

While we call the @BryBoschan, whether its input (series) is level, or detrend?

Best regard.
Hardmann

[TomDoan]

Dear Tom:

While we call the @BryBoschan, whether its input (series) is level, or detrend?

Best regard.
Hardmann

Levels. It uses its own definition for the trend.

[luching]

Hi Tom: I have a small clarification on BBQ. I thought it picks out pairs of turning points, a peak and a trough. However, for some series while running the code, I could get only a peak but not a trough. Does this mean that I should ignore that point? Or can I play around with the censoring rules to see if I can get all pairs?

[TomDoan]

Hi Tom: I have a small clarification on BBQ. I thought it picks out pairs of turning points, a peak and a trough. However, for some series while running the code, I could get only a peak but not a trough. Does this mean that I should ignore that point? Or can I play around with the censoring rules to see if I can get all pairs?

It shouldn't have consecutive peaks or consecutive troughs by the time it's done. It's possible to have those at an intermediate stage, but those get eliminated as it works through the algorithm.