Opinion by: Kevin Loaec, CEO of Wizardsardine
For decades, institutions have followed a familiar pattern when managing assets. They choose a large, regulated custodian. Then, institutions transfer responsibility. Institutions rely on the assumption that scale, compliance and insurance equate to safety.
In traditional finance, this approach holds. Transactions are reversible, central banks provide backstops and regulators can intervene. When something breaks, there are mechanisms to absorb, unwind or redistribute the damage.
Bitcoin changes those assumptions completely because it is a bearer asset. Control is defined by cryptographic keys, and not account credentials. Every single transaction is final. There is no authority that can freeze, reverse, or recover funds once they move onchain. Yet, many institutions still approach Bitcoin using the same mental model they apply to more traditional assets.
The result is a quiet contradiction. Institutions pay custodians large fees for the appearance of safety. They also accept the risks that Bitcoin was designed to mitigate.
<h2 id="when-control-is-outsourced-risk-concentrates">When control is outsourced, risk concentrates</h2>
Custodial models are built on delegation. Assets are pooled. Keys are shared, abstracted or held behind layers of internal controls. Governance lives offchain. It's enforced through policies, approvals and service agreements rather than the asset itself.
From an organizational perspective, this can feel sensible because responsibility is externalized. Liability appears contained and insurance is cited as a backstop.
Bitcoin does not recognize delegation. If keys are compromised, lost or misused, there is no external authority that can intervene. Insurance coverage is often partial, capped or conditional.
As a result, in a systemic failure, clients face the same bottleneck. There is a single custodian holding assets for many parties, with limited ability to make everyone whole.
This is not a theoretical concern. Concentrated custody creates honeypots. Honeypots attract failure. Failures can occur through technical compromise, internal error, regulatory action or operational breakdown. In Bitcoin, concentrating control does not reduce risk. It does the opposite: Risk is amplified.
The industry has already seen how this plays out. Large, <a href="https://cointelegraph.com/news/three-years-ftx-collapse-crypto-creditors-exchangeearns-back-trust" rel="noopener noreferrer nofollow">centralized custody</a> models have failed before. They’ve left consumers, businesses and counterparties tied up in lengthy recovery processes. Limited visibility, with uneven outcomes.&nbsp;
<h2 id="governance-cannot-live-outside-the-asset">Governance cannot live outside the asset</h2>
The core misunderstanding is not technical. It is organizational. Institutions are accustomed to enforcing governance through accounts, permissions, emails and internal workflows. That approach works when assets themselves are controlled by intermediaries. In Bitcoin, governance that lives outside the asset is, at best, advisory.
If an institution does not control the keys, it does not control the asset. Boards and auditors are right to be wary of fragile set-ups. A model where one individual can move funds is indefensible. Regulators are also right to push back against unclear control structures.
The choice is not between a single-key wallet and full custodial outsourcing. Bitcoin allows governance to be enforced directly at the protocol level. Spending conditions, approval thresholds, delays and recovery paths can be encoded into the wallet. Control becomes structural rather than procedural. The network enforces the rules, not a vendor’s backend or a support desk.
<h2 id="policy-driven-custody-changes-the-risk-model">Policy-driven custody changes the risk model</h2>
Modern Bitcoin scripting makes it possible to design custody around real organizational needs.
An institution can require multiple stakeholders to approve transactions. It can enforce time delays. It can define recovery paths if keys are lost or personnel change. It can separate day-to-day operations from emergency controls. These rules are enforced onchain, deterministically, every time. All of this fundamentally alters the risk profile.
Related: <a href="https://cointelegraph.com/explained/the-crypto-events-that-reshaped-the-industry-in-2025" rel="noopener noreferrer nofollow">The crypto events that reshaped the industry in 2025</a>
Instead of trusting a custodian to behave correctly under stress, institutions rely on systems that behave predictably by design. Instead of outsourcing risk to insurance policies, they reduce the likelihood of catastrophic failure in the first place. It is a matter of engineering.&nbsp;
<h2 id="the-insurance-narrative-deserves-scrutiny">The insurance narrative deserves scrutiny</h2>
Custodial insurance is often presented as the ultimate safeguard when in practice, it is frequently misunderstood. Several <a href="https://cointelegraph.com/news/sam-bankman-fried-ftx-goes-on-trial-week-review" rel="noopener noreferrer nofollow">high-profile custody failures</a> have shown that insurance coverage often falls short of client expectations, either due to coverage caps, exclusions or prolonged claims processes.
Large custodians insure pooled assets, and coverage limits rarely scale linearly with assets under custody. Exclusions are also common and payouts depend largely on the nature of the incident, and the custodian’s internal controls. In a systemic event, insurance does not eliminate risk, it distributes a fraction of it.
By contrast, individually controlled, policy-driven Bitcoin wallets are far easier to underwrite. Risk is isolated, controls are transparent and failure scenarios are bounded. For insurers, this is a simpler and more predictable model. The process of insurance works best when it complements strong controls, not when it compensates for their absence.
<h2 id="sovereignty-is-operational-not-philosophical">Sovereignty is operational, not philosophical</h2>
Vendor dependence introduces another layer of institutional risk that is not often known. Custodial outages, policy changes, or regulatory interventions can leave funds temporarily inaccessible. Exiting a custodian relationship can be slow, expensive and operationally complex, particularly for organizations operating across jurisdictions.
In practice, this has already happened through <a href="https://www.cnbc.com/2022/06/13/crypto-lender-celsius-pauses-withdrawals-bitcoin-slides.html" rel="noopener noreferrer nofollow">withdrawal freezes</a>, compliance-driven access restrictions and service outages that left clients unable to move assets precisely when timing mattered most.
With onchain, open-source custody systems, the software provider is not the gatekeeper. If a service disappears, the institution retains control. Interfaces can change and providers can be replaced. The asset remains accessible because control lives on the blockchain, not inside a company’s infrastructure. This is not an argument against service providers but an argument for removing them from the critical path of asset control.
<h2 id="trust-the-protocol-not-the-promise">Trust the protocol, not the promise</h2>
Bitcoin offers institutions something rare: the ability to hold a high-value asset with rules that are transparent, enforceable and independent of any single counterparty.
Yet many institutions still prefer familiar narratives over structural safety. Log-in screens feel safer than scripts. Brands feel safer than math, and insurance sounds safer than prevention.&nbsp;
This level of comfort can come at a huge cost.&nbsp;
Institutions should not pay for the illusion of safety while absorbing unnecessary counterparty risk. Bitcoin allows governance, recoverability and control to be built directly into how assets are held. The technology is mature. The tools exist.
What remains is the willingness to abandon custody models that belong to a different financial system.
Opinion by: Kevin Loaec, CEO of Wizardsardine.

Institutions are paying Bitcoin custodians for the privilege of added risk

Table of Contents

Cointelegraph

Other articles published on Mar 29, 2026

Join our free newsletter for daily crypto updates!