Read some previous posts and still have no clue. I understand that the more frequent data will have lower correlation because of of the asynchronism. But why a LONGER TIME period of historical estimate have a temptation to do this instead of SHORTER TIME PERIOD?
Thank you!
you have a longer time period, you do not have enough data to cover that entire period, so you use the shorter period of data that you have and try to fill up the longer period … thus end up with lower correlation of data, and messed up estimates…
Thanks for the reply! Sorry I am still not quite clear on this. Hope you can help answer the one example I listed below. Thanks!
If I have a 5 year yearly return data, so we have 5 numbers
Now we use 10 year yearly return historical data, so we have 10 numbers. So, why I don’t have enough data for the 10 year period? Why would I prefer to use monthly 10 year data (120 numbers) now?
you have a longer time period of data available to you, so you get more confident that “extrapolating” that data will be more accurate. If you had 1 month of Sales data for a new company - would you feel confident saying 1 year’s worth of Sales = 1 month sales * 12 or would you be more confident saying I have had X amount of Sales for 10 years now - so X / 10 is an estimate of my 1 years Sale, or better still Average monthly sales = X/120?
Great to know! Thanks a lot for your help!!