3 discount rates

Hallo :) 

I'm trying to use Global Discounting on my Markov model. The problem is the discount rate is cycle dependent, so from cycle 1 to 46 = 3,5%, 47-69 = 2,5% and > 70 = 1,5%. 

I figured that I need a table to make this happen, but regardless of which format i use it just does'nt seem to work. I have tried to enter a fixted discount rate on 3% and it worked fine. 

Anyone familiar with using tables to discount? 

Best regards 




  • Official comment

    The Global Discounting function only supports a fixed interest rate over Markov cycles. There is no way to use this function for a non-fixed interest rate.

    You should be able to modify the custom discounting function in the example model Special Features > Custom Function Discount.trex. The function within that model delays discounting for 5 cycles. You want something different, but the basic structure of the custom function will remain the same.

    Note that if you use this approach, you will have to place the CustomDiscount function around every cost and utility in the model.

    It is also critical that you manually test the values you will see in the Markov Cohort trace because the discounted values will not be shown separately.

    Comment actions Permalink
  • Thanks. 

    I had a chat with a colleague and a simple table without cycle 0 solved the problem with different rates and global discounting. 

    Comment actions Permalink
  • TreeAge Pro's global discounting uses a single interest rate calculated at _stage 0 for every Markov cycle.

    You can use the Discount function to apply a different rate by cycle. However, I recommend against it. Even with the Discount function, discounting would not be applied in a graduated manner (diff rate for diff years). It would be applied using the discount rate set for that specific _stage value.

    I believe that the discount function should generally be applied with a fixed rate.

    Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post