ucodegen, using long term US stock market returns has serious survivorship bias issues.
How did Russian stocks do from 1871? Farmland in Poland? Confederate and Third Reich bonds?
To put it another way, any asset class we have long term historical data on is cherry-picked and better than average, simply because so many asset classes had drastically negative, -100% returns.
What number to use then?
In my view, it is mistaken and often hubris to assume one’s investments will do better than treasuries. So about 2%.
