Trust is infrastructure

There is a mantra the digital-asset sector has repeated for years: institutional adoption will arrive when the technology matures. Once standards consolidate, interoperability works and custody is solved, institutions will come in force. It is, according to this narrative, merely a matter of engineering and time.

It is a defensive, comfortable position. And, to a large extent, a false one.

The more honest version starts with two figures. As of July 2026, around $345B in assets are committed to tokenization. Of that total, only $33.5B actually circulates on-chain. More than 90% of the value gets stuck somewhere in the funnel between the decision to tokenize and the asset operating in the market.

If the problem were technological, this funnel would be inexplicable. Tokenized Treasuries manage close to $16B and distribute yield on-chain daily. BlackRock runs BUIDL with billions under management. The DTCC —the entity that settles practically all US equities and custodies over $114 trillion in securities— is piloting tokenized-securities trading with more than fifty firms, JPMorgan and Goldman Sachs among them. The technology already settles, custodies, distributes coupons and passes audits. It has been doing so, soundly, for years.

The real bottleneck is not technical. It is human.

When a tokenized-asset proposal reaches a risk, investment or product committee, nobody in the room asks whether the smart contract works. It is taken for granted, the way email delivering messages is taken for granted. The questions today are different: who is accountable if something goes wrong, where does the regulator stand, who has done this before and how did it go. They are legitimate questions. No whitepaper answers them.

A track record answers them. Years of verifiable public positions. Data published even when the market was up and silence was more profitable. Identifiable executives who defended their thesis before regulators instead of tweeting against them, and who kept their face visible through the down cycles too. That record appears on no balance sheet or income statement, yet it is what decides which proposals clear the committee and which die in it.

My thesis is that this record deserves a more accurate name than “reputation” or “brand”: in digital assets, trust is infrastructure.

Infrastructure in the literal sense. A layer of the system, as indispensable as custody or settlement, without which the whole cannot operate at institutional scale no matter how perfect the code or the corporate marketing. And it behaves as such in three observable properties:

First, it is designed. The market is full of excellent technology nobody trusts yet, and of mediocre technology moving billions because someone built credibility first. If trust were an automatic by-product of technical merit, this distribution would be impossible.

Second, it is built ahead of need.Nobody builds a bridge while crossing the river. Regulatory and institutional credibility accumulates over months of sustained public work. The company that starts communicating when it needs the “yes” is late by definition. Communications stops being a budget question and becomes a strategy decision: what matters is not how much you invest, but how long before you need it.

Third, it is maintained. A one-off campaign is not infrastructure, and neither is a spike of visibility. What a risk committee values is continuity: the voice that was there last quarter, last year and through the previous cycle. That sustained presence is the one thing that cannot be improvised.

These three properties offer a practical test for any executive: if a risk or evaluation committee asked the three key questions tomorrow (who is accountable, where does the regulator stand, who did it before), does public, dated, signed material exist that answers them? If the answer depends on documents you would have to draft this week, that trust layer simply does not exist.

The stablecoin case validates this thesis with clarity, precisely because the technological variable is solved: a 2026 stablecoin does, in essence, the same thing as a 2019 one. Even so, the category has gone from the discredit of 2022 to more than $300B operating under regulatory frameworks such as the GENIUS Act in the United States or MiCA in Europe.

What changed was neither the code nor the technology. It was a decade of hard institutional work: issuers who published the composition of their reserves when nobody demanded it, executives who testified before lawmakers and held public positions under their own name. Those who did that work lead the category today. Those who did not are footnotes.

This reality carries a consequence for the sector’s boards: if trust is infrastructure, executive communications is not a support function. It is critical-layer engineering, and it must be decided at the same level as the other essential layers. Delegating it to “whoever runs the socials” is like delegating custody to an intern. Nobody would do that with the assets. Most still do it with their credibility.

There is precedent for doing it well. One of the most influential venture capital firms in the world (a16z) was built, in practice, with three people in the room: two investors and a communications strategist who designed the narrative before the firm even existed operationally. Others had the capital. The trust infrastructure, they did not.

Fifteen years later, their competitors still attribute the success to the returns, which took time to arrive and were disputed. The real explanation was that third chair, built on communications — but that story is for another installment.

The gap between the $345B committed and the $33.5B that actually circulates will not be closed by another round of technical development. It will be closed by the companies that understand they have a layer left to build, and by the executives willing to build it with their name, their face and enough lead time.

The technology already crossed the river. Trust is still building the bridge.