A financial product delivers not a good but a promise: money today in exchange for money in the future, under certain circumstances. Every promise is made of three elements: time, risk and trust. The return an investor actually receives from that promise can be split into five terms: time compensation, risk compensation, information advantage, trading costs and luck, on the premise that the promise is kept. The core of financial product design is to state the promise clearly, match it to the user's needs, and make it credible.
1. Today's money for tomorrow's promise
When you buy a loaf at a bakery, you get the bread as you pay, and you know on the spot whether it tastes good. When you put money into a fixed-term deposit at a bank, what you get after paying is only a promise: in three years the bank will return the principal with interest.
Every financial product has this structure: pay now, receive later under certain circumstances. Understanding a financial product means answering three questions:
When is it paid out? When does the money paid today come back? This is time.
How much is paid in which circumstances? The future has many possibilities; how much does the promise pay in each? This is risk.
Who pays, and why believe them? Who is making the promise, and do they have the ability and the will to keep it? This is trust.
Settled on the spot versus a promise across time
The payment is the same; what comes back sits at a different point in time.
Illustrative; amounts in yuan. The deposit earns 2% a year. The three outcomes for the stock fund only show that there is more than one possible result; they are not probabilities or forecasts.
2. The shape of the promise
The future holds more than one possibility, so a promise must state how much it pays in each. Arrow proposed in 1953 (the English translation was published in 1964) that a security can be understood as "a right to payment in different states of the world". From this angle, financial products differ mainly in the shape of their promise:
Fixed claims. Deposits and bonds promise a fixed amount; as long as the promisor does not fail, the outcome is certain.
Residual claims. Shares promise no amount at all. Shareholders receive what is left after the company has kept all its promises to employees, suppliers, creditors and others: a great deal when business is good, possibly nothing when it is bad.
Contingent claims. Insurance and options pay only when a specified event occurs, such as an accident, or a price crossing a certain level.
The creditors and shareholders of the same company hold two promises of different shapes.
One company, two shapes of promise
The company owes its creditors 100 (in units of 10,000 yuan). Drag the company's value at maturity to see how it is divided between creditors and shareholders.
050100150200
Creditors hold a fixed claim: when company value exceeds 100 they receive only 100; when it falls short of 100 they take everything that remains. Shareholders hold a residual claim: when company value falls short of 100 they get nothing, and everything above 100 goes to them. On this basis Merton (1974) treated equity as a call option on the value of the company with the face value of the debt as the strike price.
On this basis Merton pointed out in 1974 that equity is economically equivalent to an option on the value of the company: the higher the company's value, the more shareholders receive; when the company's value is not enough to repay its debts, shareholders' losses are limited to what they put in. Creditors, in effect, give up all gains from company value above the debt in exchange for a fixed promise.
Financial products do not make risk disappear; they only redistribute it among participants. Depositors pass risk to the bank, creditors leave the residual fluctuation to shareholders, and policyholders transfer the losses from accidents to the insurer. The party that bears more risk demands a higher expected return as compensation. This is the risk premium.
Put the same money into promises of different shapes, and the distributions of outcomes differ greatly.
The same money, four promises
Put 10,000 yuan into each of four promises and hold to maturity. Each promise is simulated over 4,000 possible futures; on the left are the promisor and the circumstances in which the promise fails.
Middle 90% of outcomesMedianMeanPrincipal of 10,000 yuan
Illustrative parameters that do not correspond to any real product: deposit rate 2% a year; government bond yield 3% a year, held to maturity; corporate bond coupon 5%, with an annual default probability of 1.5% and 40% of principal recovered on default, interest received and recoveries reinvested at 2%; stock index mean annual log return 5.5% and volatility 18%, before fees. Inflation fluctuates around 2.5% a year (standard deviation 1.2%). If government or corporate bonds are sold before maturity, their prices move with interest rates; the figure shows only hold-to-maturity outcomes.
Three points can be read from the figure:
The more fixed the promise, the more concentrated the outcome, and the lower the average return. The outcome of a deposit or a government bond is a single point; the outcome of shares is a wide range, with the highest average. The higher return is compensation for bearing greater uncertainty.
A fixed promise is not a fixed outcome. A corporate bond promises more interest than a government bond, at the cost of possible default by the issuer, which gives the distribution of outcomes a long downward tail. Measured in purchasing power, even a deposit can "lose money".
The median outcome of shares is below the mean. In most cases the outcome falls below average; a few especially good cases pull the mean up. The longer the horizon, the larger the gap. Presenting a product by its average return alone will leave most users' expectations unmet.
3. What capital markets do: the functional perspective
Banks, exchanges, funds and insurers take different forms, and those forms keep changing. Merton and Bodie proposed in 1995 that the financial system is better understood through its functions than through its institutions. Institutions change; functions are comparatively stable. The financial system has six core functions.
The six functions of the financial system
Institutions change; functions are comparatively stable. Choose an innovation to see which functions it mainly improves.
1Transferring resourcesAcross time, places and industries
2Managing riskDiversifying, hedging and transferring
3Clearing and settling paymentsGetting trades settled
4Pooling funds and subdividing ownershipSplitting large investments into small shares
6Dealing with information asymmetry and incentivesMaking promises checkable
Choose an innovation above.
The division into functions is from Merton and Bodie (1995). The mapping of innovations to functions is this book's own synthesis; the AI agent case is a hypothesis yet to be tested and is shown with dashed lines.
The functional perspective offers a stable standard for judging anything new: which function does a new product form perform better or more cheaply? Once AI enters financial products, it should be measured by the same standard: which function it improves, at what cost, and whether the improvement holds up under measurement.
4. Implications for product design
Design the product around the promise: first state who promises, what is promised, and in which circumstances it fails, then match each of the three elements, time, risk and trust, to the user's needs.
Chapter summary: design the product around the promise
The design requirements for each of the three elements, and the new questions an agent raises.
The promiseWho promises · What is promised · When it fails
Time
Match maturity to when the money is needed; measure returns in purchasing power
Risk
Match the shape of the promise to the capacity to bear risk; do not let the average stand for the outcome
Trust
Disclosure, custody and records are product features; make the process verifiable rather than relying on short-term results as proof
AI agents: a new promisor
What it promises to do, why it is credible, and how problems are detected are answered with the same three elements
References
Arrow, K. (1964). The role of securities in the optimal allocation of risk-bearing. Review of Economic Studies, 31(2), 91–96.
Merton, R. (1974). On the pricing of corporate debt: The risk structure of interest rates. Journal of Finance, 29(2), 449–470.
Merton, R. & Bodie, Z. (1995). A conceptual framework for analyzing the financial environment. In D. Crane et al., The global financial system: A functional perspective (pp. 3–31). Harvard Business School Press.