Replacing Judgment With Models

The transcript discusses the limitations of computational innovation, particularly in data assembly and hypothesis testing. It critiques the over-reliance on diversification in venture capital and banking, emphasizing the need for case-by-case analysis. The text also highlights the misunderstanding of transaction costs in financial models, arguing that mechanistic models overlook long-term credit losses.

What are the flaws mentioned in the underlying logic of modern financial approaches?

How does diversification serve as a second line of defense in venture capital?

What is the misconception about transaction costs when using mechanistic models in credit extension?

In what way do models fail to account for the real transaction costs of loans?

What is the impact of credit losses on the long-term effectiveness of extending consumer credit?