Financial Products in the Age of AI
A financial product delivers a promise. AI allows financial products, for the first time, to act on the user's behalf, and it also makes the questions "what is promised, and why is it credible" harder to answer. This book places three disciplines, finance, product and AI, in a single framework to discuss how financial products should be designed in the age of AI.
The book can be read in two ways, which are two views of the same book.
Decomposition starts from four equations and takes them apart term by term. The first three parts each cover one set of constraints: the nature of markets (where returns and risks come from), the nature of users (what people want and what they can bear), and the nature of AI (how much an agent can do on people's behalf). Part Four shows how to design financial products under these three sets of constraints. It answers which terms make up an outcome, and suits checking what a product is missing.
Derivation starts from a world without finance and adds one constraint at a time, working out how financial institutions and products emerge. It then lowers one parameter, agency costs, and runs the whole derivation again; finally it uses the same method to derive a financial product for the age of AI. It answers why an institution emerges, and suits imagining new products when constraints change. You can switch between the two at any time at the top of the contents.
In both ways of reading, the finance follows the line that runs from Bachelier and Fama to multi-factor scientific investing, with economic analysis as the main thread and as little complex mathematics as possible. Mechanisms are handed to figures, animations and simulations the reader can operate: adjust the conditions yourself and see how the conclusions change.
Choose a reading order by background
Pick a background: solid circles are disciplines you already have, dashed circles are ones to fill in.
The book is updated continually; chapters not yet finished appear grey in the contents. It discusses principles and methods, does not recommend any specific security or product, and does not constitute investment advice.