Robert Carver: BST: 070: Risk Management - PursuitOfEdge/podcasts GitHub Wiki

measure expected risk and realized risk
e.g. if you expect a loss on a big day to be -2%, but you're losing that every other day, then there is a big difference and you should scale it back
the way you're thinking about risk may be wrong or incomplete
if they actually had good risk management, you'd never hear about them, they'd never be in the news
you should have a volatility measure e.g. MA of standard deviation or ATR e.g. tomorrow this could move +/- 1% and incorporate this into your risk
with vola sizing, the less volatile the bigger your position size, but the problem with vola is that when it is lower it has a tendency to spike back up quickly
so the solution is to not only look at recent vola, but vola over the past few years, and set a limit for how big it can e.g. spike 10% in a day in nightmare scenario in the past
his step #1: size your positions according to how much risk there is in the overall market/instrument you're trading
book recs: When Genius Fails, Against the Gods
your risk should be even on a really, really, really bad day you wouldn't be tempted to turn it off