Simkin: Calibrating forecasts requires frequent, uncorrelated repetitions and feedback
Todd Simkin · #119 Todd Simkin: Making Better Decisions · Sep 7, 2021 · at 35:54
Todd Simkin, trader and educator at Susquehanna International Group (SIG), discusses how decision-makers can calibrate probabilistic estimates.
0:00 / 0:19exact quote · 19.3s
720p mp4 · rendered on demand · StarZero watermark
“If the only thing that you predict are very hard to predict outcomes and you predict them very rarely, you're not going to be able to calibrate very well at all. On the other hand, if you predict many different things that particularly if they're uncorrelated, so you have a lot of repetitions where you're going to get feedback, then you're going to have learning”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Todd Simkin
Simkin: Call-out culture incentivizes finger-pointing regardless of expertise
“The big problem with call out culture is that we're rewarding people just for the call out. That it doesn't matter what their other behavior it is. It doesn't matter what other positions they have, what knowledge they have, what expertise they have. They just …”
Simkin: Probabilistic forecasts cannot be judged wrong unless 0% or 100%
“It's really hard to say anybody's wrong about a single probability prediction, unless they said that something was zero percent or a hundred percent.”
Simkin: Constructive communication beats probabilistic thinking for truth-finding
“Talk more is number one. That beats probabilistic thinking that, you know, that, that beats sort of anything else in truth finding is, is being able to bring in other people in the decision process in a constructive way.”
Simkin: Traders should not back limited opinions against informed counterparties
“Our approach is to recognize that we are often on the bad end of informational asymmetry that other people who are looking to trade in the markets frequently know more than we do. And as a result, we don't want to back our own limited opinion against a more in…”
Simkin: Decision-makers must update opinions in a Bayesian way
“The most important thing from our perspective is to be willing to update that opinion in a Bayesian way, to use approaches that are going to be inclusive of all the information available to us, and only when we get to mutually exclusive information or mutually…”
Simkin: Use related instruments to find disconfirming evidence before taking trades
“And again, in isolation, it's really hard for me just to say this person's wrong. But if I have a similar looking option, similar because it's the same expiration, or because it's the same strike and a different expiration, or that there's something about the …”