Page 1 of 1
Specifying a 1% shock
Posted:
Mon Jan 21, 2013 10:50 pmby Carol
Hi,
I am simulating a log-linearized model and I would like to clarify how to specify a 1% shock.
Should a 1% shock be specified as stderr 0.01 OR stderr 1?
Please let me know. Thanks.
Re: Specifying a 1% shock
Posted:
Wed Jan 23, 2013 7:30 amby jpfeifer
It does not matter as long as you interpret the results correctly. If you specify a 1% shock as 0.01 then if you get a let's say a standard deviation of y of 0.015 this means 1.5%. In contrast, if you specify a 1% shock as 1 then everything is just scaled up by 100 and now a standard deviation of y of 1.5 means 1.5%. (Usually the latter is better for numerical accuracy and readability)
Re: Specifying a 1% shock
Posted:
Wed Jan 23, 2013 1:02 pmby Carol
Hi jpfeifer,
Thank you so much for replying and thanks for this information. Have a good day.