On trust as infrastructure

Every financial system runs on infrastructure that nobody sees until it fails.

Payment rails. Settlement systems. Custody. Identity. In On the rails we argued that machines will need their own versions of these. There is one more layer that sits underneath all of them, and it is harder to build than any protocol.

Trust.

Not trust as a feeling. Trust as a property of a system: the reasonable expectation that a statement it makes is true and that it has no hidden reason to make it.

In most of finance, trust is borrowed. A bank is trusted because a regulator stands behind it. A fund is trusted because an auditor signs its accounts. A rating is trusted, or not, because of who pays for it. The question of who pays has always been the quiet centre of the matter. In On scoring we argued that machines will increasingly do the judging. They will only be trusted if the incentives behind the judging are clean.

Stranded capital is a domain where this is unusually visible.

A person looking for an old pension is, almost by definition, in a weak position. They do not know exactly what exists. They may be abroad, dealing with unfamiliar institutions in an unfamiliar language. They are often older. And the thing they are looking for is money, which attracts people who would like to help themselves to it.

So the market around lost pensions has always contained two kinds of actor. Some help people find what is theirs and then step aside. Others use the search as a way in: to sell advice, to move the money into a product, to take a share of it, or simply to collect personal details. From the outside, the two can look identical. Both say they help.

The difference is not in the search. It is in what happens after it.

That is why independence is not a marketing line in this domain. It is the product. A tracing service that earns from what the person does next has a reason to steer them. One that does not, does not. PensionHunter's checklist on telling a real tracing service from a scam sets out the warning signs, and almost all of them come down to who gets paid and when.

A trustworthy verifier, in this sense, has a few observable properties.

It names the free route. If an official public tool can do the job, it says so first, plainly, with the address.

It takes nothing from the people downstream. No percentage of the pension, no commission from whoever might manage the money next.

It shows its sources. Every statement about a scheme, a deadline or a rule points to the official page it came from, with the date it was read.

It does not hold the money. It identifies, confirms and hands over. The person deals with the institution directly.

It is clear about who runs it. Readers of this publication know that PensionHunter is operated by our publisher, Battersea Park Capital, and we say so whenever we write about it. Disclosure is cheaper than suspicion.

None of these properties is technically difficult. All of them are commercially uncomfortable, because each one closes off an easy source of revenue. That is precisely why they work as trust. A promise that costs nothing to make is worth nothing to hear.

Italy illustrates how good public trust infrastructure can look. Contribution records sit with the national institute, INPS, and free assistance with applications made from abroad is available through patronato offices around the world. A person does not need to pay anyone to be helped. PensionHunter's guide to finding an Italian INPS record from abroad points there first.

Spain is similar in spirit. A person's whole Spanish working life sits on a single contribution record, viewable through the official Tu Seguridad Social account. The obstacle, again, is the digital credential, not the absence of a public route. The Spanish contribution record guide explains what can be done from abroad.

The lesson from both is that trust infrastructure already exists in public form. What is missing is an independent layer that works across borders and does not compete with it.

For our part, the rule is simple. We always point people to the free government tracing route first, and we take nothing from financial advisers.

As machines take over more of the searching and more of the judging, the scarce good will not be intelligence. It will be a credible statement from a party with no reason to lie. That is what infrastructure means when the infrastructure is trust.

The summary, for anyone choosing who to trust with a search:

• Trust in finance is mostly borrowed from whoever stands behind a statement.
• In stranded capital, the search is easy to fake. What happens after it is not.
• A trustworthy verifier names the free route, takes nothing downstream, shows its sources, does not hold the money and discloses who runs it.
• Those rules are commercially uncomfortable. That is why they work.