I am calculating YoY change in macroeconomic variables (inflation, industrial production etc.). However this growth is distorted by low base of last year. Currently I am using 2 year cagr to adjust for the base effect. Is there a better statistical method to separate the base effect and momentum.
For a "stable" base, you can use any average of previous years that "makes sense", that is an average with minimal variance and give an explicit explanation If the last year doesn't deviate from that average then it is unnecessary
To demonstrate the momentum effect you can use a moving average, in that case there is no meaning for the base and edge cases are normalized