Recession analysis

[spaceniuzai]

Hello- I have a question on data manipulation. I would like to study the recovery path after each U.S. economic recession. Say, I have a monthly data series on the U.S. industrial output. What is the easiest way for me to convert this data into several series that index the trough month of each recession as 100 and the subsequent values following the trough? As illustrated in the example below:

DateID Recession#1 Recession#2
month0 100 100
month+1 105 102
month+2 108 103

Thanks a lot!

[EViews Gareth]

There are two interpretations to that question. The first is "what is the mathematical way of defining the troughs", the second is "How do I do that in EViews".

I can't figure out which one you're asking.

[spaceniuzai]

There are two interpretations to that question. The first is "what is the mathematical way of defining the troughs", the second is "How do I do that in EViews".

I can't figure out which one you're asking.

Sorry for not being clear. What I am asking is the second part of the questions - "How do I do it in Eviews". US economic cycles are declared by the National Bureau of Economic Research (http://www.nber.org/cycles/cyclesmain.html). Thanks.

[EViews Gareth]

So in essence you're really asking how to create an index series based upon a certain date:

Code: Select all

series xidx = x/@elem(x, "1991m1")*100

[spaceniuzai]

So in essence you're really asking how to create an index series based upon a certain date:

Code: Select all

series xidx = x/@elem(x, "1991m1")*100

Thanks. This is helpful. I also want to compare the recovery paths after several recessions. In other words, the identifier for each series is not a date, but the number of months after the respective economic trough. Based on my limited experience of Eviews, I have only used workfiles based on date, e.g. annual, monthly, etc. Thanks for your help.