Why I stopped trusting "average return" for my FIRE number, and what I use instead
A while back I posted here about Barista FIRE and the bridge strategy. Since then I've gone deeper on the thing that breaks a lot of FIRE plans in practice: sequence-of-returns risk.
The trap: a 5% average return over 30 years sounds safe. But averages hide the order of returns, and order is everything once you're withdrawing. If years 1–2 of retirement are −30%, you're selling in a crash and the portfolio can fail even though the long-run average was fine.
So a simple "X% average" calculator gives false confidence.
I now run Monte Carlo simulations that address this (using block sampling), a Student-t distribution (fatter tails; crashed happen more often than 'normal' assumes), volatility clustering, and a hard capital floor to report a succes probability across thousand of simulated paths of historical returns.
It folds in state pension, wealth tax (and contribution room), so the numbers are after tax. Curious how people handle the withdrawal phase specifically.
Let me know if you want me to run the (Barista) FIRE analysis on your situation/portfolio. I am a financial econometrician so I love these kinds of analyses. Hope that I can help others with it!