Bayesian Updating: Revising Probability Estimates with New Evidence
Summary
This article introduces Bayesian reasoning as a way to estimate an unknown probability from limited observations and revise that estimate when new evidence arrives. It distinguishes a prior probability, based on information available beforehand, from a posterior probability updated using observed evidence. Examples involving a ball’s location and complaints directed at two shopping centers illustrate how base rates and differing complaint rates affect the updated estimate.
The discussion connects this iterative process to risk management: decisions in changing markets depend on incorporating current information and revising beliefs as evidence changes. The examples are explanatory rather than empirical trading studies, and the article does not give a market model, quantitative strategy, or method for choosing priors. Its central lesson is conceptual: probability estimates are conditional on available information and should be updated when relevant evidence appears.
Key ideas
- Bayesian reasoning updates an initial probability estimate when new evidence becomes available.
- A prior reflects information before the evidence, while a posterior incorporates the evidence.
- Base rates and observation likelihoods can change which explanation is more probable.
- The article applies this updating concept to risk management in dynamic markets.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.