Reply
Thu 7 Jan, 2016 04:46 pm
Hi,
I am currently working on a project that requires me to use a cox proportional hazard survival model- I have never used this before. What I am struggling with is how to make a continuous variable such as interest rates or real GDP an exposure variable when it changes over the time period of survival.
Context: I am trying to model a subdivision process with 4 stages, I want to know how economic indicators affect the time to 'death' (time to completion of a stage), this process can take 5-10 years so the interest rate the properties are exposed to changes over this time. Initially I thought about anualizing the data but that does not work when the development survives over several years. I am using excel and Stata.
Any ideas appreciated, I hope I have explained this well enough.
Thanks,
@cmmcle,
You're looking for the impossible. There are no ways to predict the economy that far ahead. Even economists will disagree on interest rates and GDP. There are just too many unmeasurable variables with the world economy always in flux.
The best measure of the economy is always historical. Even then, they are revised.