Skip to content →

Model “Risk”

I was having a conversation about model risk the other day, and I think it’s a misnomer.

Risk to me means the possibility that something bad (or maybe good) will happen tomorrow. It implies future uncertainty. If I’m standing on the corner, I know exactly where I am now, but I don’t know what will happen when I start to cross against the light. If I own Google right now, I know what it’s worth; risk involves what it will be worth later.

Model “risk” is different. It’s not a future risk, it’s a PRESENT uncertainty. You don’t know what a hybrid product is worth now because there is no liquid market, and you don’t know which model to use in order to tell you what someone will pay for it NOW. It’s not model risk, it’s model uncertainty.

(I just googled risk and came upon a reference to a book by Frank Knight called Risk, Uncertainty, and Profit written in 1921. I am obviously fairly ignorant because this is apparently a famous book. Anyhow, he seems to define In it “risk” as randomness with knowable probabilities and “uncertainty” as randomness with unknowable probabilities. I’m not sure I agree — in economics all probabilities are pretty much unknowable, so there is only uncertainty in that case.+

Published in Models