Open the product catalogue of any financial institution and you will find deposits, loans, insurance, bonds, stocks, funds, investment advice and pensions. Memorizing them by type is not hard. The hard part is explaining why they look the way they do: why a deposit can be withdrawn at any time, why a fund's assets are held in custody by a different institution, why insurance requires a premium in advance and pays only after something goes wrong.
When the physicist Richard Feynman died, his blackboard carried the sentence "What I cannot create, I do not understand." The best way to understand a set of institutions is to build it from zero. Arranged in the order in which they arise, these products stop being a list and become a chain: each one exists to solve a problem left by the one before.
Finance turns today's value into credible claims on the future. Every financial product is a way, under some constraint, to make such claims more accessible, more credible or cheaper.
Each of the three words has a specific meaning. Accessible means that people who could not make the exchange before now can, as when a young person without savings borrows to pay for school. Credible means that the promise is kept in the future, as when deposit insurance spares depositors the worry that their bank will fail. Cheaper means that the exchange costs less, as when an index fund lets ordinary people hold the whole market for a very low fee. Whether a product has value can be tested with these three words.
The derivation begins with one person, one plot of land and one year's harvest. Each chapter introduces just one new constraint. It first asks what happens without a solution, then looks at the mechanism that grows out of it. The mechanism solves the old problem and creates a new one, which becomes the constraint of the next chapter. Problems off the main line are marked as side branches and are not pursued.
Economics textbooks often begin their reasoning with Robinson Crusoe on his island, taken from Defoe's 1719 novel. This book follows that practice but carries the derivation all the way to today's financial markets. The derivation is not history. Historically, institutions appeared in a tangled order across many places; the derivation keeps only the logical order, in which each step presupposes the one before.
A simulator accompanies the whole book. In Chapter 1 the world contains one person and one year's harvest; each later chapter unlocks a rule: lending, commitment, money, insurance, banks, shares, markets. Readers can switch off any rule at any time and see what the world falls back to. All numbers in the simulator are illustrative. They show the direction and relative size of each mechanism and do not represent any real economy.
Part One derives the world as it is today: ordinary people hand their money to professional institutions, and disclosure, custody and regulation keep those institutions' promises credible. Every step has a cost. One of these is the cost of having someone else think and act on your behalf, that is, the cost of agency. Large language models and AI agents make this cost fall sharply.
Part Two changes this parameter and reruns the steps of Part One that it affects: which institutions become redundant, and which become more important. Time, risk and irreversible consequences do not disappear as a result, and a separate chapter explains why. Part Three uses the same method, starting from one person and one sum of money and adding one user-side constraint at each step, to derive a financial product for the age of AI.
The book also exists in a Decomposition version, which starts from four equations (for returns, for the outcome a user actually receives, for the amplification factor and for net value) and takes them apart term by term. Decomposition answers which terms make up an outcome and is suited to checking what a product lacks. Derivation answers why an institution arises and is suited to imagining new products when constraints change. The two versions are independent and can be read on their own.