Hello dear community,

I use the following regression model of an article in JFE, this is the exact notation how the author represents his model:


Ln(Issunance Yield) = ßLn(Climate) +ßLn(Controls) + ßState-Year FE + u


My dependent variable Ln(Issuance Yield) shows a decreasing trend in a graphical representation. I think we all know that the interest rate level has decreased significantly over the last few years.

How do I take this fact into account in an OLS regression?


Some background: i cannot use A fixed effects model, since my variable of interest Climate is time invariant, it measures an average effect for every country.


I have a population of 24 countries (GER, USA, UK etc) with about 16.000 observations between 2005 - 2017.
I would like to investigate the relationship between the issue yield of government bonds and climate risk and control variables.

The number of observations per country is very unbalanced, the number of issuances per year varies greatly per country. Furthermore, some countries issue different bonds on the same day.

Kind regards
Tim