Real-world asset tokenization has the asset managers, the regulators, and the capital lined up. What it doesn't have is a custody and wallet layer that institutions can actually use without inheriting retail-crypto risk. That gap is becoming the bottleneck nobody wants to own.
The pitch for tokenizing real-world assets has been remarkably consistent for three years now. Take a bond, a money market fund, a slice of private credit, or a Treasury, put a representation of it on a blockchain, and you get instant settlement, programmable compliance, fractional ownership, and a 24/7 market. BlackRock's BUIDL fund, Franklin Templeton's on-chain money fund, and a growing list of tokenized Treasury products have proven the thesis works on paper and, increasingly, in practice. The asset side is solved. The demand side is real. So why does adoption keep stalling at the same place?
The answer, according to a recent argument from Pharos Network CEO Wish Wu, is unglamorous: the wallet. Or more precisely, the entire custody and key-management layer that sits between an institution's balance sheet and the chain. Tokenization solved how to put the asset on-chain. It did not solve how a regulated entity holds, moves, and accounts for that asset without taking on operational and legal risk it was specifically built to avoid.
The problem institutions actually run into
When a retail user loses a seed phrase, they lose their money and that is the end of it. When a $400 billion asset manager structures a tokenized fund, the same self-custody model is a non-starter. A pension fund cannot explain to its board that a junior operations employee held the private keys to a nine-figure position. A bank cannot reconcile an on-chain holding against its internal ledger if the wallet infrastructure was designed for a single anonymous user, not for a multi-desk institution with segregation requirements, audit trails, and four-eyes approval on every transfer.
This is where the friction lives. Most wallet infrastructure was built for crypto-native users and then bolted onto institutional needs after the fact. The result is fragmentation. An institution dealing with tokenized assets across several chains and several issuers often ends up with a different wallet, a different custody arrangement, and a different compliance integration for each one. There is no single view of positions, no consistent policy engine, and no clean way to enforce who can move what under which conditions.
That fragmentation has real costs. It slows onboarding, because every new asset or chain means another integration and another risk review. It complicates compliance, because know-your-customer and anti-money-laundering checks have to be wired into transfers asset by asset rather than enforced at the wallet layer. And it undermines the original promise of tokenization, since instant settlement means little if moving an asset between two custody systems still takes days of operational coordination.
Why this is a market, not just a complaint
The interesting part for anyone watching the ecosystem is that the wallet problem is creating a distinct category of company. Custody used to mean a vault and an insurance policy. For tokenized assets it means a programmable account layer that can encode compliance rules directly, support multi-party computation or multi-signature key management so no single person holds full control, and present institutions with reporting that maps cleanly to their existing accounting systems.
Several players are positioning here. Established digital-asset custodians like Fireblocks and Anchorage Digital have built MPC-based infrastructure aimed squarely at institutions, abstracting away seed phrases in favor of policy-governed transactions. Hardware security vendors have a stake too, with companies like Yubico supplying the authentication hardware that gatekeeps access to these systems.
Pharos itself is making a more infrastructure-level argument: that the chain and the wallet experience need to be designed together for institutional RWA flows, rather than treating compliance and custody as add-ons to a general-purpose network. Whether a purpose-built approach beats integrating best-of-breed custody onto existing chains is an open question, and a reasonable person can be skeptical of any single network claiming to be the answer. The vendor making the diagnosis also sells the cure, which is worth keeping in mind.
The trade-off nobody escapes
What makes the wallet layer genuinely hard, rather than just neglected, is a tension that does not fully resolve. The entire point of putting assets on a blockchain is to remove intermediaries and gain self-custody and programmability. The entire point of institutional custody is to add controlled intermediation, recovery mechanisms, and accountability. Build a wallet that is too open and institutions will not touch it. Build one that recreates every layer of traditional custody and you have rebuilt the system tokenization was supposed to streamline, now with a blockchain bolted underneath it.
The firms that win this category will be the ones that find a defensible middle. That means key management where loss is recoverable but control is still cryptographically enforced, compliance that runs as code at the wallet level rather than as a manual checklist, and a unified account model that works across the chains and issuers an institution touches. Get that right and the rest of the tokenization stack finally has somewhere to plug in.
What changes if it gets solved
If the custody and wallet layer matures into something institutions trust, the second-order effects are larger than the wallet itself. Onboarding a new tokenized asset becomes a configuration change rather than a months-long integration project. Settlement that is instant on-chain becomes instant in practice, because the asset does not have to traverse incompatible custody silos to move. And the addressable market widens, because the institutions currently sitting out, the ones who like the economics of tokenization but cannot accept the operational risk, get an answer they can show a risk committee.
For now the situation is a familiar pattern in infrastructure markets. The flashy layer, the tokenized asset, arrived first and got the attention and the headlines. The boring layer underneath it, the part that determines whether any of it is usable at scale, is where the actual constraint sits. The companies quietly building that layer are not promising to reinvent finance. They are trying to make a wallet an institution can sign off on, which turns out to be the harder and more valuable problem.
Comments
Please log in or register to join the discussion