Institutional Custody Solutions 2026: Digital Asset Platform Comparison

The institutional custody landscape spans three categories: crypto-native specialists (Fireblocks, BitGo), qualified custodians (Anchorage, Copper), and traditional institutions with digital capability (BNY Mellon, State Street). Custody pricing is bespoke and quote-driven rather than publicly posted, but the stakes keep rising: two traditional custody banks alone report more than $120 trillion in assets under custody and/or administration (as of Q2 2026), even as OCC-chartered crypto-native platforms and stablecoin issuers compete for the same institutional mandates.

The Custody Question

When a pension fund's investment committee asks "who holds the keys," they are asking the question that determines whether a digital asset allocation is viable. Custody is not a back-office detail; it is the foundation on which institutional deployment rests. A custody error introduces operational risk that no return can justify.

The good news: institutional-grade custody solutions now exist across multiple categories. The complexity: each category presents distinct trade-offs between cost, regulatory status, operational capability, and asset coverage. Understanding these trade-offs is essential for informed platform selection.

Three categories have emerged to serve institutional needs, each with different DNA and different strengths. The right choice depends on your organization's specific requirements: regulatory constraints, asset mix, operational sophistication, and cost sensitivity.

The Three Categories

Crypto-Native Technology Platforms

Fireblocks leads with $14+ trillion in secured transactions (as of Jun 2026) across 150+ blockchains and 2,400+ institutional clients. MPC (multi-party computation) architecture eliminates single points of failure. Strengths: technology sophistication, DeFi integration, policy engine. Weakness: the core technology platform still typically requires a sub-custody arrangement for fiduciary accounts, though Fireblocks Trust Company (NYDFS-chartered, launched 2024) now offers a qualified-custody path directly, expanded in 2026 to institutional staking (with Figment) and Canton Coin support.

BitGo offers qualified custodian status plus technology platform. The firm went public on the NYSE (BTGO) on January 22, 2026, the first crypto custody firm to do so, and reported approximately $63 billion in assets on platform and 5,569 institutional clients (up 42% year-over-year) as of Q1 2026. Strengths: regulatory status, institutional insurance ($250M via Lloyd's of London), hot/cold wallet flexibility.

Qualified Custodians

Anchorage Digital holds the first federal bank charter for digital assets (OCC-regulated, 2021) and was for years the only digital asset firm operating as a national trust bank. As of September 2026, Circle, Ripple, Paxos, BitGo and Fidelity Digital Assets have all received OCC conditional or final approval for national trust charters of their own, part of the institutional stablecoin custody landscape now taking shape. Circle's final approval, for Circle National Trust, came July 10, 2026, signaling that the OCC-chartered category is widening from a single-provider exception into a competitive field. Strengths: banking-grade regulatory status, staking capability, governance solutions. Weakness: more limited asset coverage.

Copper provides regulated custody (with $500 million of cold-storage insurance arranged by Aon through a Lloyd's-led insurer panel) alongside ClearLoop for off-exchange settlement. Strength: exchange integration. Note: Copper put itself up for sale in May 2026 at an asking price of about $500 million; by late August 2026, reported offers were near $200 million, and no sale had been announced as of September 28, 2026. It is a consolidation signal in the custody-adjacent settlement space worth monitoring, not (yet) a change in custody status.

Traditional Institutions

BNY Mellon (~$62.6 trillion in assets under custody and/or administration, as of Q2 2026 per company filings, up 12% year-over-year, a record) launched digital asset services in 2022. Strengths: balance sheet strength, regulatory relationships, existing client integration. Weakness: limited asset coverage, conservative approach.

State Street (~$57.9 trillion in assets under custody and/or administration, as of Q2 2026 per company filings, a record level, up 18% year-over-year) offers digital custody through its Digital Markets division. Similar profile to BNY Mellon with emphasis on ETF servicing.

Both banks capitalize these custody commitments under existing prudential frameworks. See capital treatment of tokenized and digital assets under Basel III for how that capital treatment is evolving.

The Second Layer: When Custodied Securities Become Tokens

The largest US securities custodian is preparing to add a layer to its own model. The Depository Trust Company (DTC), part of the Depository Trust & Clearing Corporation (DTCC), provides custody for more than $114 trillion in securities. DTCC plans to launch its DTC Tokenization Service in October 2026. The service covers Russell 1000 stocks, ETFs tracking major indices, and US Treasury bills, bonds and notes. SEC staff granted three-year no-action relief for it in December 2025. As of September 28, 2026, the full service has not launched; July brought only initial limited production trades.

The design matters more than the date. DTCC describes each token as a digital twin that carries the same entitlements and investor protections as the underlying security. The security itself stays at DTC. The token sits on a ledger, in a wallet held by a DTC participant. Tokenization in this form does not remove the custodian. It adds a second custody layer on top of it. We cover the market-structure side in When the DTCC Tokenizes: What Changes for Issuers.

For a treasurer, that means two custody questions where a treasury policy usually asks one. The first is familiar: who holds the security, and under which regime. The second is new: who controls the wallet that holds the token, and how the token record reconciles to the position at DTC. A policy that answers only the first question leaves the second to whoever runs the wallet.

The July trades show why this is a treasury matter. On July 15, 2026, DTCC reported that more than 30 firms had used DTC-tokenized assets in production trades. The workflows included collateral pledges, securities lending, US Treasury repo settled delivery-versus-payment, and central counterparty margin. Those are treasury workflows, not only trading ones. The trades ran on DTCC's private Besu network and on the Canton Network.

The Buyer's Decision Framework

How to Structure the Custody Decision

Selecting a custodian is not a procurement exercise. It is a governance decision that a board or investment committee will eventually be asked to defend. The following six questions frame that decision without prescribing a single answer, because the right custodian depends on an institution's regulatory posture, asset mix, and risk tolerance. The stakes are rising as corporate treasurers structure 2026 digital asset allocations across a wider set of custody providers.

1. What is the custodian's actual regulatory status, and does it match your fiduciary obligation? "Regulated" is not a single category. A firm can hold a state trust charter, a federal (OCC) national trust charter, SEC-qualified-custodian status under Rule 206(4)-2, or simply operate as an unregulated technology provider requiring a separate sub-custody arrangement. As of mid-2026, the OCC-chartered category itself is widening. Anchorage Digital was for years the only firm operating as a national trust bank, but Circle, Ripple, Paxos, BitGo and Fidelity Digital Assets have all received OCC conditional or final approval for charters of their own. Ask which specific license the custodian holds, in which jurisdiction, and whether that license actually covers the asset class and account type you intend to use.
2. How are client assets segregated, and what happens to them in a custodian bankruptcy? Segregation is not binary. Ask whether assets are held in individually titled accounts or in an omnibus structure, whether the custodian's own balance sheet ever has a claim on client assets, and whether outside counsel has opined on bankruptcy remoteness in the custodian's specific jurisdiction. The OCC's pending stablecoin-custody rulemaking is instructive here: its draft Subpart C imposes segregation and non-commingling obligations and specifically addresses the "prudent use" of omnibus accounts, a signal of where regulators expect this question to be tested first.
3. What does the insurance actually cover, and at what limit? Headline insurance figures (crime policies, specie coverage, Lloyd's-placed programs) rarely cover every loss scenario. Ask for the policy's exclusions, whether coverage is per-client or aggregate across the custodian's full book, and whether the policy responds to insider theft, key-management failure, and smart-contract risk separately.
4. What operational controls and independent attestations exist, and how current are they? SOC 1 and SOC 2 Type II reports, penetration-test cadence, and key-ceremony documentation are the baseline evidence of operational maturity. Ask for the report date (attestations older than twelve months should prompt questions), the scope of controls tested, and whether any exceptions were noted.
5. Which jurisdiction governs the custody relationship, and how does that interact with your own regulatory regime? A custodian regulated in one jurisdiction (say, the UK or EU) may not satisfy a fiduciary obligation defined under U.S. rules, and vice versa. See how MiCA structures custody obligations in the EU for the European contrast. With U.S. federal stablecoin-custody rules still in proposal form, institutions operating across borders should confirm today which regime actually governs their assets, not which regime the custodian's marketing materials emphasize.
6. If the asset is tokenized, who holds the token, and how does it reconcile to the underlying position? A tokenized security can carry two custody layers: the security at the depository and the token in a participant's wallet. Ask which entity controls the wallet and its keys, and whether it is the same custodian that holds the underlying position. Ask how, and how often, token-layer entitlements are reconciled to the DTC position. Ask what happens to the token if the wallet provider fails while the security stays at DTC.

These six questions are a starting checklist, not a scoring system. The right weighting depends on an institution's own mandate.

The Cost Question

Custody pricing differences reflect different value propositions, not inefficient pricing, but they are difficult to compare because most providers do not publish a rate card. Traditional institutions draw on existing infrastructure and can price standard-asset custody as an extension of existing relationships. Crypto-native and OCC-chartered platforms price on a bespoke, AUC- and service-mix basis that varies by client. See the Buyer's Decision Framework above for how to structure an apples-to-apples RFP.

Because pricing is bespoke, the only reliable way to quantify the cost difference between two specific custodians is a like-for-like RFP covering custody, settlement, insurance, and onboarding fees for your specific asset mix and volume.

But cost is not the only variable. Regulatory risk, operational capability, asset coverage, and integration with existing systems all factor into the total cost of ownership. A lower quoted fee is not automatically the better outcome once regulatory risk, operational capability, and integration are weighed in.

Platform selection is easier to get right when the evaluator has no stake in the outcome. See why independent advisory on an advisory-fee basis matters for platform selection.