Every term of the return equation assumes that promises will be kept. Financial products are credence goods: buyers find it hard to judge their quality, so their markets tend to shrink. What makes these markets work is a set of institutions that make promises credible: disclosure and audit, ratings, custody and segregation of assets, deposit insurance, regulation and reputation, and financial intermediaries that monitor on others' behalf. After a trade, central counterparties and clearing and settlement see that the promises made in the trade are finally kept.
Following the classification of Nelson (1970) and Darby and Karni (1973), goods can be divided into search goods, whose quality can be judged before purchase; experience goods, whose quality becomes known only after use; and credence goods, whose quality is hard to judge even after use. Financial products are credence goods: results arrive late and are mixed with luck. A fund that beats the market for three years may have a capable manager, or may simply have been lucky.
Akerlof analysed the consequences of this situation in 1970. If buyers cannot distinguish good products from bad ones, they will only pay a price that reflects average quality. Providers of good products find this not worthwhile and leave, the products left on the market grow steadily worse, and in the end trade becomes hard to sustain.
How a market shrinks when buyers cannot tell good from bad
Sixty products, with quality ranging from 0 to 100. Sellers know the quality of their own products; buyers know only the average quality of products on the market, so they will only pay for average quality.
Financial markets can exist because of a set of institutions that make promises credible. They act at different points along the chain of a promise: some make the condition of the promising party verifiable, some lower the cost of verification, and some provide a backstop when a promise fails.
Tap any mechanism.
These mechanisms are not add-ons to financial products; they are part of the product itself. What users buy is not only an expected return but also a judgment of how credible that promise is.
In 1984 Diamond identified a fundamental reason for financial intermediaries to exist: delegated monitoring. One bank monitoring many borrowers centrally is far cheaper than each depositor monitoring them separately. But this immediately raises a new problem: who monitors the monitor? The bank's answer is to spread its loans across a large number of borrowers so that its own promise is stable enough, and depositors therefore need not monitor the bank.
Why delegated monitoring is cheaper
Six depositors' money is lent to eight borrowers. Each line is one act of monitoring.
A trade is only a promise; for money and securities actually to change hands, the trade must go through clearing and settlement. In modern markets, a central counterparty typically takes on every trade: it becomes the buyer to every seller and the seller to every buyer, so each participant faces only the central counterparty and need not worry about whether thousands of counterparties will default. The central counterparty also offsets each participant's purchases against its sales and settles only the net amount. Settlement uses "delivery versus payment": money and securities are exchanged simultaneously, so neither side has to hand over its part first.
Central counterparty and netting: from a web to a few lines
Four brokers do 10 trades with one another in a day. Switch between the three stages to see how the number and value of settlements change.
Settlement cycles are getting shorter. From 28 May 2024, US stocks moved from settlement on the second trading day after the trade to settlement on the first trading day after the trade; Hong Kong stocks still settle on the second trading day after the trade. The shorter the cycle, the fewer unsettled promises there are and the lower the counterparty risk and the margin required, but the greater the demands on back-office operations.
The preconditions for keeping promises are often invisible in a product's interface, yet they decide whether users actually receive their returns.
Provide the reasoning, records and verifiable data, rather than relying on short-term results alone to prove the product's quality.
State who holds the assets in custody, whether they are segregated from the firm's own assets, and what protections apply and up to what limits.
Explain where the intermediary makes its money and whether that conflicts with the user's interests.
Show when money and securities actually arrive and when they can be withdrawn.