Bitcoin as Inheritance Asset
Onramp Custody vs Self-Custody vs Multi-Sig for Bitcoin Inheritance
By Michael Tanguma, Founder & CEO of Heirfolio. Reviewed by Diana Cruz, GIA Graduate Gemologist. Updated May 25, 2026.
TL;DR. Roughly 20% of all mined Bitcoin is permanently lost. A meaningful fraction of that loss is inheritance failure — holders who died without a workable handoff plan. Self-custody is the cheapest and most sovereign option but the highest inheritance risk. Single-key institutional custody (Coinbase, Kraken) is the lowest inheritance friction but the highest counterparty risk. Multi-sig (especially multi-institution multi-sig, like Onramp's architecture) is the middle that solves both at once. Here is how each works, who it fits, and what to do.
A reader emailed about her father, who died in March 2026 with an estimated 4.2 BTC across two hardware wallets. The family found the wallets. They could not find the seed phrases. The Bitcoin is still there — visible on the blockchain at addresses the family can identify — and it is permanently inaccessible.
Disclosure: I'm the founder of Onramp, which is one of the products compared in this article. I'll be explicit about that throughout. This is the comparison I'd write if I were not the founder, because the failure mode I'm describing is real and the solutions are scarce.
The failure mode to name first: most Bitcoin inheritance plans fail at the same point — the moment the holder is no longer there to explain the seed phrase. Beneficiaries inherit a hardware wallet they can't unlock and a passphrase they can't find. Onramp's published research found over 70 documented physical-attack cases against Bitcoin holders in early 2025 alone, and the inheritance failure category is consistently larger than the attack category in lost-coin estimates.
This article walks through the three custody architectures that survive that moment.
Document your Bitcoin position in Heir Protocol — works with any custody architecture
The three architectures
| Architecture | Who holds keys | Number of keys | Counterparty risk | Inheritance risk |
|---|---|---|---|---|
| Self-custody (single signature) | You only | 1 | None | High — total loss if keys aren't recoverable |
| Single-institution custody (Coinbase, Kraken) | The exchange | 1 (theirs) | Highest — full custodial control | Low — beneficiary form and standard estate process |
| Single-institution multi-sig | You + the exchange | 2 of 3 typically | Moderate — exchange holds majority influence | Moderate |
| Multi-institution custody (Onramp, Casa, Unchained) | You + 2+ independent institutions | 2 of 3 (or 3 of 5) | Lowest — no single party can move funds | Lowest with proper inheritance setup |
| Pure multi-sig (you control all keys) | You only, multiple devices/locations | 2 of 3 or 3 of 5 | None | Variable — depends on key distribution plan |
These five categories cover the realistic options for an individual or family holding meaningful Bitcoin in 2026. They are not all in the same risk category.
The full side-by-side comparison
| Feature | Self-custody (single sig) | Single-institution custody | Single-institution multi-sig | Multi-institution custody (Onramp/Casa/Unchained) | Pure multi-sig (you hold all keys) |
|---|---|---|---|---|---|
| Who can move funds | You alone | Exchange alone (with your auth) | You + exchange (typically 2-of-3) | You + 2 of 3 independent institutions | You alone (with multiple keys) |
| Single point of failure | Your seed phrase | The exchange | The exchange | None (no single party) | Any single key location if not redundant |
| Counterparty risk | None | Highest | High (one entity holds majority) | Lowest | None |
| Hack / freeze / bankruptcy risk | None | Yes — exchange-dependent | Yes — but recoverable with your key | Minimal — distributed across 3 institutions | None |
| Privacy | Highest | Lowest (KYC, custodian reporting) | Lower | Moderate (KYC at each institution) | Highest |
| Setup complexity | Low | Lowest | Moderate | Moderate | Highest |
| Setup time | 30 minutes | 10 minutes | 1-2 hours | 1-3 days (KYC, signing ceremony) | 2-4 hours per location |
| Annual cost | $0 (hardware ~$100 one-time) | 0–1% (often 0 for spot holding) | Variable | 0.10–0.50% AUM typical | $0 (hardware costs only) |
| Inheritance complexity (no plan) | Catastrophic (total loss) | Standard probate | Moderate | Moderate | Catastrophic |
| Inheritance complexity (with plan) | Moderate | Low | Low | Low (key recovery service standard) | Moderate |
| Hardware required | 1 hardware wallet | None (web/app) | 1 hardware wallet | 1+ hardware wallets | 2+ hardware wallets, multiple locations |
| Recovery if you lose 1 key | Seed phrase backup required | Account recovery via institution | Yes (2-of-3) | Yes (2-of-3) | Yes if backup keys exist |
| Recovery if you lose 2 keys | Total loss | Account recovery if institution still exists | Total loss | Total loss | Total loss |
| Recovery if institution fails (single) | N/A | Total loss in worst case | Recovery via your key | Recovery via 2 remaining institutions | N/A |
| Recovery if 2 institutions fail simultaneously | N/A | N/A | N/A | Total loss (worst case) | N/A |
| Time-locked or scheduled distribution | Manual via custom scripts | No | No | Yes (institutional inheritance services) | Yes (advanced multi-sig setup) |
| Sovereignty | Highest | Lowest | Low | Moderate | Highest |
| Best for typical 1 BTC household | Yes (with rigorous plan) | Acceptable | Overkill | Overkill | Overkill |
| Best for 10 BTC household | Risky without inheritance plan | Counterparty risk too high | Acceptable | Strong fit | Possible with deep technical comfort |
| Best for 100 BTC household | Insufficient | Insufficient | Borderline | Strong fit | Strong fit (with professional setup) |
| Recommended inheritance product | Heir Protocol + sealed envelope process | Beneficiary form at exchange | Institutional inheritance + Heir Protocol | Institutional inheritance + Heir Protocol | Multi-sig inheritance plan + Heir Protocol |
Where self-custody wins
The original Bitcoin thesis. You hold the keys, no one else can move the funds, no institution can fail or freeze your access. The full power and the full responsibility.
- Sovereignty. No counterparty can prevent you from moving your Bitcoin. This is the central property of the asset and self-custody is the only architecture that fully preserves it.
- Cost. A hardware wallet (Ledger, Trezor, Coldcard, BitBox02) is roughly $100 one-time. No ongoing fees.
- Privacy. Nobody other than you knows what you hold or where it moves. KYC is required only when you buy or sell through regulated channels.
- Permissionless inheritance design. You can design any handoff architecture you want — sealed envelopes, time-locked transactions, shamir-split seeds, M-of-N social recovery. The flexibility is total.
Recommended for: Small to medium positions (under roughly 1 BTC) where the loss-of-self-custody-rights trade-off doesn't justify institutional involvement. Technically capable holders who can design and test an inheritance handoff. Households where the privacy property is meaningful.
Not recommended for: Holders without a tested, written inheritance handoff plan. The single most common cause of permanently lost Bitcoin is a self-custody position whose holder died, was incapacitated, or simply lost their seed phrase. Self-custody without an inheritance plan is not a plan — it is a path to lost coins.
Where single-institution custody wins (and why it's usually wrong for inheritance)
The easiest option. You buy Bitcoin on Coinbase, Kraken, or another regulated exchange, and they hold the keys. Beneficiary forms are standard. Estate process is the same as any brokerage account.
- Lowest friction at every stage. Setup is 10 minutes. Inheritance is a standard beneficiary form. Probate handles the rest.
- Recovery is institutional. If you lose your password, the exchange has process to restore access.
- Tax reporting is automatic. 1099s, cost basis tracking, the works.
- No hardware to lose. No seed phrase to protect. No technical skill required.
The problem is that single-institution custody re-introduces every counterparty risk that Bitcoin was designed to eliminate.
- The exchange can fail. Mt. Gox, FTX, Celsius, BlockFi, Voyager, Genesis — the list of failed custodians in the 2014–2024 window is long enough that "exchange risk" is a real category. When an exchange fails, customers typically recover cents on the dollar after years of bankruptcy proceedings.
- The exchange can freeze your account. For compliance reasons, account disputes, regulatory pressure, or sometimes for reasons the exchange itself can't explain to support.
- The exchange knows what you hold. Every transaction, every balance, every withdrawal. Privacy is effectively zero.
- Geographic and regulatory risk. A US exchange operating under FDIC-style rules in 2026 may not exist or may operate differently in 2036.
Recommended for: Holders who treat Bitcoin as a tradable asset rather than a long-term store of value. Households that prioritize inheritance simplicity above sovereignty.
Not recommended for: Long-term holders. Anyone whose Bitcoin position is a meaningful fraction of their net worth. Anyone whose reason for holding Bitcoin includes the sovereignty property — because that property is exactly what single-institution custody trades away.
Where multi-institution custody wins
This is the architecture Onramp built and the one I think most clearly solves the inheritance problem for meaningful positions. Disclosure: I founded Onramp; the architecture below is the one our institutional clients have been using since 2021.
Multi-institution custody distributes the three keys of a 2-of-3 multi-signature wallet across three independent institutions. To move any Bitcoin from the wallet, two of the three signers must agree. No single institution can move funds unilaterally; the client also cannot move funds without the institutions' participation, but the client always has access to at least one key and can sign one of the two required signatures.
The Onramp implementation specifically uses:
- One key held by Onramp (the platform)
- One key held by BitGo (the custodian)
- One key held by Coincover (the recovery specialist)
Plus a fourth key held by the client (or in a multi-sig within the client's institution).
The architecture is published in detail on Onramp's Multi-Institution Custody page.
Why this matters for inheritance:
- No single party can lose your Bitcoin. Onramp can be hacked, fail, or be subpoenaed — your Bitcoin is still safe because BitGo and Coincover still hold two of the three institutional keys.
- No single party can move your Bitcoin. Including you. Two of three institutional signers must agree to move funds. The client's involvement is required for legitimate transactions and is precisely what blocks any single party from acting unilaterally.
- Inheritance is a structured workflow. When the principal dies, the institutions have an established process for beneficiary verification and key reassignment. The Bitcoin doesn't get stuck in the limbo where a self-custody seed phrase would be lost.
- No physical handoff of seed phrases is required. Your beneficiaries inherit the relationship with the institutions, not a piece of metal with words engraved on it.
Other multi-institution providers in this category include Casa (their Gold and Diamond plans), Unchained Capital (their concierge service), and a small handful of others. The architecture concept is similar; the institutional partners differ; the inheritance workflows differ.
Recommended for: Holders with positions of roughly 1 BTC or more who want institutional-grade security without single-institution counterparty risk. Households where the holder is the only Bitcoin-literate person and the beneficiaries will need a structured handoff. Family offices and high-net-worth individuals.
Not recommended for: Holders with positions under roughly 0.5 BTC where the annual cost (0.10–0.50% AUM) is meaningful relative to the position. Holders for whom maximum sovereignty (no third party can ever block any action) is a non-negotiable requirement.
Where pure multi-sig (DIY) wins
For technically sophisticated holders, pure multi-sig — where you hold all keys yourself, distributed across multiple devices and locations — combines the sovereignty of self-custody with the redundancy of multi-institution architectures.
A typical 2-of-3 setup might be:
- One hardware wallet at home
- One hardware wallet in a safe deposit box
- One hardware wallet with a trusted family member or attorney
To move funds, you need any two of the three. Loss of any one key is recoverable; loss of two simultaneously is total loss.
- Maximum sovereignty. No third-party institution involvement. You hold everything.
- Inheritance via key distribution. Your beneficiary inherits one of the keys; another goes to a co-trustee or attorney; the third stays with you until the handoff is needed.
- Cost is hardware only. No annual fees.
Recommended for: Bitcoin-native households with deep technical comfort, multi-decade horizons, and an attorney or trusted family member capable of managing one of the keys.
Not recommended for: Most holders. The operational complexity is real — you have to manage multiple devices in multiple locations, you have to test the recovery process annually, and you have to design the inheritance handoff yourself without an institutional partner. Most pure multi-sig setups that fail in inheritance do so because the holder never tested the recovery process and the key locations weren't documented.
The inheritance handoff: what actually needs to happen
For any custody architecture to survive the holder, four things need to be in place. This is the single most important section of this article.
1. Documentation of the position
Your family needs to know:
- That you hold Bitcoin
- How much, in BTC and current USD value
- At what addresses or in what accounts (without exposing keys)
- Which custody architecture you use
- Where any hardware wallets, seed phrases, or institutional contacts live
Most inheritance failures happen at this layer. The Bitcoin is recoverable; the family never learns it exists.
2. Legal documentation that names Bitcoin as an asset
Your will or trust should name Bitcoin specifically as an asset class, designate a beneficiary, and either include the institutional contact or reference the inventory you've documented elsewhere. Saying "my digital assets go to X" in a will is necessary but nowhere near sufficient.
3. Operational handoff plan
The actual mechanics of how the beneficiary takes possession. For each architecture:
- Self-custody: Sealed envelope with seed phrase in a known secure location, plus a written guide that explains what to do with it.
- Single-institution custody: Beneficiary form on file at the exchange.
- Single-institution multi-sig: Beneficiary form at the institution, plus secure transfer of your key.
- Multi-institution custody: Beneficiary form at the platform; the institutions handle the rest.
- Pure multi-sig: Documented locations of all keys, plus a written plan for which beneficiary inherits which key.
4. Tested recovery
The single most important thing you can do, and the thing almost no Bitcoin holder does: test the recovery process while you're still alive. Walk a trusted family member through the actual recovery — opening the hardware wallet, entering the seed phrase, signing a small transaction, or contacting the institutions. If they can't do it with you watching, they certainly can't do it after you're gone.
The Heir Protocol, our product, was built specifically to handle layers 1, 2, and 3 above for Bitcoin alongside physical heirlooms. It does not replace the legal documents, but it documents the holding, generates the letter of intent that points your beneficiary to it, and walks through the operational steps for whatever architecture you use.
Build a free Heir Protocol — covers Bitcoin and physical assets together
The cost comparison
Annual cost for a 5 BTC position (~$471,000 as of May 25, 2026 at $94,200/BTC):
| Architecture | Annual cost | Total 10-year cost |
|---|---|---|
| Self-custody | $0 (hardware ~$100 one-time) | ~$100 |
| Single-institution custody (Coinbase, Kraken) | $0–$4,710 (varies, often free for spot) | $0–$47,100 |
| Multi-institution custody (typical) | $470–$2,355 (0.10–0.50% AUM) | $4,700–$23,550 |
| Pure multi-sig | $0 (hardware ~$300–$500 one-time across devices) | ~$500 |
| Multi-institution custody + Heirfolio Vault Pro | $470–$2,355 + $1,188/yr | $5,888–$24,738 |
The cost of multi-institution custody is real but small relative to the protection it offers for meaningful positions. For a 5 BTC position, $1,500/year of custody cost is roughly 0.3% of the position — less than most jurisdictional risks the architecture protects against.
For smaller positions (under 1 BTC), the percentage cost of multi-institution custody becomes meaningful. Self-custody with a rigorous inheritance handoff plan is typically the right answer.
The bottom line
Five honest verdicts:
-
For small positions (under 0.5 BTC) where the loss would not materially affect your family: Self-custody with a hardware wallet, sealed-envelope seed phrase, and a Heir Protocol entry that documents the holding. Cost: ~$100 hardware + free Heir Protocol tier.
-
For medium positions (0.5–5 BTC) where the loss would matter: Either multi-institution custody (Onramp, Casa, Unchained) or a tested pure multi-sig setup with documented key distribution and an estate attorney. Cost: $500–$2,500/year + Heir Protocol Vault tier.
-
For large positions (5+ BTC) where the position is a meaningful fraction of net worth: Multi-institution custody is the right starting point. The annual cost is small relative to the counterparty and inheritance risks it eliminates. Cost: $2,500+/year + Heir Protocol Vault Pro tier.
-
For all positions: Document the holding before designing the custody architecture. Most lost Bitcoin is lost because nobody knew it existed, not because the architecture failed.
-
For nobody: Single-institution custody as the long-term home for meaningful Bitcoin. It's fine for trading; it's not appropriate as the architecture you intend to pass to the next generation.
Bitcoin's monetary properties are durable. The custody architecture you choose determines whether those properties actually transfer to your family or get lost at handoff.
Build a free Heir Protocol — covers Bitcoin and physical assets together
Frequently asked questions
Is self-custody safer than an exchange?
Safer in some ways, riskier in others. Self-custody eliminates counterparty risk (no exchange can fail, freeze, or seize your coins) but introduces operational risk (you can lose the seed phrase, forget the passphrase, or die without a workable handoff). For trading and short-term holding, exchanges are fine. For long-term inheritance-grade holding, self-custody or multi-institution custody is generally preferred — but only with a tested inheritance plan.
What is multi-institution custody?
Multi-institution custody distributes the keys of a multi-signature Bitcoin wallet across multiple independent institutions. No single institution can move funds; multiple signers must agree. Onramp's architecture uses three institutional signers (Onramp, BitGo, Coincover) plus the client. Casa's Gold and Diamond plans use a similar pattern. The design eliminates single-institution counterparty risk while maintaining institutional-grade operational support, including structured inheritance workflows.
Can I lose my Bitcoin with multi-institution custody?
Yes, in narrow scenarios. If two of the three institutions fail simultaneously (very low probability — the institutions are jurisdictionally and operationally independent) and the client's key is also unavailable, the Bitcoin would be lost. The architecture is designed to survive any single-institution failure or any single-key loss. Multi-institution simultaneous failure is the rare worst-case.
How does Bitcoin inheritance work in probate?
Bitcoin is treated as personal property under US law. It is subject to probate the same way as other personal property unless held in a trust, jointly titled, or held with a beneficiary designation. The challenge is that probate courts can only distribute Bitcoin that the executor can actually access. If the keys are lost, the Bitcoin shows on the blockchain but cannot be moved — which means the estate has an asset on its balance sheet that it cannot transfer. This is the failure mode multi-institution custody and a documented inheritance plan are designed to prevent.
What is the safest hardware wallet for self-custody?
The four most widely-trusted hardware wallets as of 2026 are Ledger (Nano S Plus, Nano X), Trezor (Safe 3, Safe 5), Coldcard (Mk4, Q), and BitBox02. Each has trade-offs: Coldcard is the most security-focused and Bitcoin-only; Ledger has the broadest multi-asset support; Trezor is the most open-source-pure; BitBox02 is the cleanest interface. All four are appropriate for inheritance-grade holdings if the seed phrase is properly stored and the inheritance handoff is documented.
Can I use multi-sig without an institutional partner?
Yes. Pure multi-sig (you hold all keys, distributed across multiple devices and locations) is the most sovereign architecture available. The operational complexity is significant — you have to manage multiple devices, design and test the recovery process, and create an inheritance plan that survives loss of any single key. For technically capable holders with strong support from an estate attorney, pure multi-sig is appropriate. For most holders, the additional complexity is not worth the additional sovereignty.
How much does multi-institution custody cost?
Typically 0.10–0.50% of assets under custody per year. For a 5 BTC position (~$471,000 as of May 2026), that's $470–$2,355 per year. The cost is real but small relative to the counterparty and inheritance risk it eliminates. For positions under roughly 1 BTC, the percentage cost becomes meaningful and self-custody with a rigorous inheritance plan is typically the better choice. For positions above 5 BTC, the case for multi-institution custody is strong.
What happens if Onramp or Casa goes out of business?
The architecture is designed to survive single-institution failure. In multi-institution custody, you (the client) always have access to at least one key, and the institutions hold others. If any single institution fails, the remaining institutions plus you can move funds to a new architecture. The client onboarding documents at Onramp, Casa, and similar providers include the specific recovery process. The relevant question is "what happens if two institutions fail simultaneously" — and the answer is that the risk is low but not zero, which is why client involvement is part of the design.
What to do next
If you hold Bitcoin and don't have an inheritance plan: stop optimizing the custody architecture and write the plan first. Document the position, the architecture, the location of keys or institutional contacts, and the steps your beneficiary needs to take. The Heir Protocol does this for you in about 12 minutes.
If your position is large enough to justify institutional support: read the multi-institution custody architecture brief (full disclosure: this is Onramp's own page) and compare to Casa's and Unchained's published architectures.
If your position is small enough that self-custody is the right call: use a hardware wallet, store the seed phrase securely (steel plate, geographically separate from the wallet), document the holding in Heir Protocol, and test the recovery process with a trusted family member.
The Bitcoin is the easy part. The handoff is the hard part. Build for the handoff and the rest follows.
Michael Tanguma is the founder and CEO of Heirfolio and the founder of Onramp Bitcoin. Onramp is the multi-institution custody architecture referenced throughout this article; I have an obvious stake in the outcome and have tried to be explicit about the trade-offs and competing options (Casa, Unchained, pure multi-sig, self-custody, exchange custody). This article was reviewed for accuracy by Diana Cruz, a GIA Graduate Gemologist and Heirfolio's Valuation Lead. Custody architecture details verified against published documentation from Onramp, Casa, Unchained Capital, and major exchanges as of May 25, 2026. Bitcoin loss estimates from Chainalysis, Glassnode, and academic research on lost-coin proxies. Pricing data verified against each provider's public fee schedule as of the same date.