I kept staring at the same number and wondering who it actually belonged to. Twenty-eight billion dollars is the kind of figure that makes a timeline stop scrolling. Pair it with a household asset manager and an onchain vault, and the brain fills in a story the press release never quite told. A tokenized Fidelity bond ETF sounds, on first read, like the whole fund packed its bags and moved onto a blockchain. It did not. What launched is narrower, and in my view more interesting, once you strip the headline down to the plumbing.
On October 5, 2026, Plume introduced nBND, a vault whose primary reserve is shares of Fidelity Total Bond ETF, the fund known by the ticker FBND. The pitch is simple enough to repeat at a dinner table. Onchain investment products began with short bills and cash-like instruments. Capital allocators, the argument goes, also want duration and a manager who can move between government paper, corporate credit, and emerging-market debt. nBND is Plume’s attempt to put that mix inside a tokenized wrapper.
What Actually Launched, And What Did Not
Start with the sentence that should sit above every chart. FBND is an actively managed U.S.-listed bond ETF. nBND is a Plume vault that uses shares of that ETF as its main reserve asset. Those are two different objects. One has a decade-plus record, thousands of holdings, and an asset base measured in the tens of billions. The other is a new onchain claim on a slice of that existing product, and Plume has not published how large that slice is.
I have found that the fastest way to misread these launches is to treat the underlying fund’s size as the vault’s size. Official fund data put FBND assets at about $26.6 billion as of June 30, 2026. Later third-party snapshots pushed the estimate toward $28.2 billion. That money sits inside the traditional ETF. It is not capital that has migrated onto Plume. No launch figure for deposits, shares reserved, or total value locked was disclosed. If you need a single caution before you go further, that is it.
The vault is tied to an existing fund. It is not the fund itself, relocated.
Holding nBND should not be described as becoming a registered shareholder of FBND. The announcement frames the product as a tokenized vault backed by ETF shares. That is a reserve structure, not a direct brokerage account with your name on the transfer books of the ETF. The difference matters for voting, for tax lots, for what happens in a redemption crunch, and for who you call when something breaks. Perhaps the most useful mental model is a claim on a claim. You hold the vault token. The vault holds the ETF shares. The ETF holds the bonds.
The Fund Behind The Wrapper
FBND launched on October 6, 2014. Expense ratio listed in official materials: 0.36%. Mandate: a high level of current income, with a portfolio that can reach beyond plain government debt. Managers may allocate across Treasuries, investment-grade corporates, high-yield bonds, mortgage-backed securities, and emerging-market exposure. As of August, the portfolio ran to thousands of line items. That is active management in the old-fashioned sense. Someone is choosing, trimming, and rotating. You are not buying a single Treasury bill with a maturity date printed on the face.
A small correction is worth making, because track-record talk gets sloppy. Reports that give FBND a twenty-year ETF history appear to blend it with an older mutual-fund strategy in the same family. The ETF’s own inception date is October 2014. Twelve years is a real record. It is not two decades of exchange-traded history. If you are comparing drawdowns, fee drag, or behavior in 2022’s rate shock, use the ETF’s own window.
Fidelity’s public description positions the fund for investors who want income and some cushion against equity swings. Cushion is not the same as safety. High yield and emerging-market debt can gap. Mortgage-backed paper has its own prepayment moods. Duration means the price can fall when yields rise, even if the coupons keep arriving. Anyone treating nBND as a cash substitute is importing the wrong product into the wrong slot.
Why The Short-Bill Era Felt Too Small
Tokenized fixed income grew up on instruments a lawyer could explain in one breath. Treasury bills. Money-market funds. Short repos. Maturities were close. Structures were plain. Redemption stories were easier to underwrite. Plume’s chief executive, Chris Yin, called those products the starting points, and said allocators now want duration and active management sitting next to them.
There is a market-size argument underneath that line, and it is worth keeping in proportion. Tokenized U.S. Treasuries were cited as growing from roughly $12 billion in April to about $15 billion in June. The global fixed-income stock is described as larger than $100 trillion. Those numbers sketch an addressable world. They are not assets committed to this vault, and they are not a forecast. I tend to read them as a map of the gap, not a promise that the gap closes on any particular timetable.
A Clean Split Between Headline And Balance Sheet
Early summaries blurred the ETF and the vault into one object. A cleaner split looks like this.
| Item | What it is | What the number refers to |
| FBND | Actively managed bond ETF | About $26.6B at June 30, 2026; later estimates near $28.2B |
| nBND | Onchain vault reserved mainly by FBND shares | Launch deposits not disclosed |
| Holder status | Vault token, not automatic ETF registration | Not a direct shareholder claim |
| Mandate | Income across credit and government debt | Not a short-bill cash product |
If a dashboard later prints a vault balance, that figure is the one that belongs next to nBND. Until then, repeating the ETF’s asset total as if it had moved is a category error. Harmless in a headline. Expensive in a risk memo.
How The Reserve Idea Is Supposed To Work
The economic story, stripped of branding, is familiar from other tokenized funds. A vehicle acquires shares of a traditional product. It records ownership in a way the onchain system can reference. Users receive a token that is meant to track the vault’s claim on those shares, minus fees and operational friction. Programmable portfolio tools are the selling point on the crypto side: collateral, composability, faster internal transfers, the chance to plug a bond sleeve into a broader onchain book.
Cynthia Lo Bessette, who leads digital asset management at Fidelity Investments, confirmed the firm’s participation and described the work as bringing financial products onchain so investors can use more programmable tools. She pointed at collateral and access to capital as possible uses. The announcement did not set adoption targets, name a second fund, or give a timetable. Plume called nBND the start of the relationship and said other assets could follow. Could is doing real work in that sentence.
I keep coming back to a practical question. If the token is useful as collateral, who accepts it, at what haircut, and under what legal opinion? A press line about collateral is not a lending market. Until a venue, a margin schedule, and a default waterfall exist, the feature is a design intention. Interesting. Not yet a rate you can underwrite.
What You Are Actually Exposed To
Stack the risks instead of blending them. The bond book can lose mark-to-market value. Credit spreads can widen. Emerging-market lines can gap on a currency or policy shock. The ETF wrapper has its own tracking and liquidity traits. The vault adds operational, smart-contract, custody, and legal-structure risk that the ETF shareholder in a normal brokerage account does not carry in the same form. Then there is basis risk between the token’s traded price and the net asset value of the reserved shares, especially if secondary liquidity is thin.
- Rate risk from duration, not from a bill maturing next month
- Credit risk across investment-grade, high-yield, and emerging-market sleeves
- Structural risk because the token is not the same as registered ETF ownership
- Liquidity risk if redemptions or secondary markets are narrower than the slogan
- Disclosure risk while deposit size and fee stack remain unpublished
None of that makes the product unserious. It makes it a bond product wearing new rails. Bonds still behave like bonds.
Active Management Is The Feature And The Variable
Passive Treasury tokenization is easy to narrate. You know the issuer. You know the maturity ladder. Active credit is a different animal. FBND’s managers can lean into corporates, mortgages, or emerging markets when they see value, and step back when they do not. That flexibility is why someone might prefer it to a stack of bills. It is also why two “bond tokens” can diverge hard in a stress month.
In my experience reading these launches, people underweight manager risk because the brand is familiar. Familiar is not the same as static. An active fund’s past sector mix is a snapshot, not a covenant. If you care whether the sleeve is mostly government paper or a fatter credit bet, you have to keep reading the holdings, not the ticker alone.
The 0.36% expense ratio is the published ETF cost, not necessarily the all-in cost of holding the vault. Wrappers can add administration, mint, redeem, or performance-related charges. Until that schedule is public, comparing nBND with buying FBND in a brokerage account is incomplete. Sometimes the onchain route wins on access or collateral utility. Sometimes it loses on fees and friction. Both can be true for different users.
Where This Sits In Plume’s Longer Build
nBND did not appear in a vacuum. Through 2025 and 2026 Plume assembled pieces that look, from the outside, like an attempt to make tokenized funds legible to regulated intermediaries. Its Kimber unit holds a U.S. Securities and Exchange Commission transfer-agent registration, secured in October 2025. A transfer agent keeps ownership records and processes changes in ownership. That is unglamorous work. It is also the work securities law actually cares about.
In May the firm received a digital asset business license from the Bermuda Monetary Authority, described as covering regulated onchain vault management. In August it joined a DTCC digital-assets working group that also included names such as Nasdaq and Charles Schwab. Membership is not an integration. It does not place DTCC-held assets on Plume. I would treat it as a seat in a room, not a pipe already connected.
Other distribution experiments fill in the pattern. A vault effort with Ether.fi earlier in the year spoke of a planned $100 million allocation across bond ETFs, credit pools, and collateralized loan exposure. Separate fixed-income vaults reached Bybit users through products tied to PIMCO and CMBI. An nBASIS vault later placed Bitwise and Invesco-linked products inside Binance Wallet. A proof of concept with Shinhan Asset Management tested a tokenized Korean-won bond fund offshore, focused on whitelist controls and identity checks, without offering tokens to Korean residents. The through-line is distribution plus controls, not a single hero product.
Why Fidelity’s Name Changes The Conversation
Brand is not a risk model, but it changes who will take the meeting. A vault reserved by shares of a large, actively managed bond ETF from a firm investors already use in retirement accounts is easier to explain than a bespoke credit pool. Lo Bessette’s on-record participation matters for that reason. It signals the asset manager is in the room, not merely name-checked after the fact.
Still, participation is not the same as a balance-sheet guarantee, and it is not the same as moving the ETF’s creation and redemption onto new rails. Fidelity runs FBND. Plume runs the vault structure around shares of FBND. If you blur those roles, you will mis-assign both credit and accountability. When something operational fails, the first question is which entity actually performs that function. The announcement does not hand you that org chart in full.
Programmable collateral is a use case only after someone will lend against the token.
A practical test, not a slogan
There is a quieter institutional angle too. Allocators who already own FBND in a custody account may not need a token at all. The users who might are the ones who want the exposure inside an onchain portfolio, or who cannot easily hold the ETF in their current setup, or who hope the token becomes acceptable collateral. Those are different jobs. A product can be a good answer to one and a poor answer to the others.
Duration Is The Point, And The Discomfort
Short bills were the training wheels of tokenized fixed income because price volatility was modest and the story ended on a known date. Duration removes that comfort. A multi-sector bond fund can post a negative year even while it pays income. Anyone who lived through the rate reset of 2022 does not need a lecture on that. They need a reminder that a new wrapper does not repeal it.
Yin’s line about wanting duration next to the cash products is, I think, the right product insight. Portfolios are not one instrument. A treasury sleeve and a credit sleeve do different work. The gap in onchain markets has been the second sleeve, offered by a manager with a public record rather than a newly assembled book. nBND tries to fill that gap by pointing at a book that already exists. Clever, if the legal chain from token to shares stays intact. Fragile, if any link in that chain is vague.
Ask a blunt question before you size it. Do you want bond exposure, or do you want a yield token that feels like cash? If the second, this is the wrong shelf. If the first, you still have to decide whether the extra rails are worth the extra unknowns versus holding FBND the ordinary way.
Questions The Launch Left Open
Good launches answer the boring questions early. This one left several of them for later, which is common and still unsatisfying.
- How many FBND shares were reserved at launch, and who custodies them?
- What is the all-in fee, and how does it compare with holding the ETF directly?
- Who can mint and redeem, and on what settlement timeline?
- Is the token permissioned, and what identity checks apply?
- What happens to the token if the vault, the custodian, or the transfer record fails?
- Which venues, if any, will accept nBND as collateral, and at what haircut?
Until those answers land, the honest description is a structure announcement, not a fully documented fund factsheet. That does not make it vapor. It makes it early. Early products reward readers who separate design from deposits.
How I Would Read The Next Headline
More Fidelity assets may follow. That line is explicit. It is also non-specific. No second ticker, no date, no size. If a follow-up arrives, the useful test is the same one that applies here. Is the new token a vault reserved by shares of an existing fund, or something closer to native issuance? Are deposits disclosed? Is the holder a shareholder or a claimant on a reserve? Does the fee stack fit in a sentence?
I would also watch whether nBND stays a showcase or becomes a balance. Showcase launches can be technically real and economically tiny. Balance is deposits, secondary volume, and redemption that works on a dull Tuesday. The dull Tuesday is the whole product.
Quick filter before sharing a figure: ETF assets are not vault assets. A working group is not an integration. Collateral talk is not a loan market. Inception of the ETF is 2014, not a 20-year ETF record.
There is room, genuinely, for onchain fixed income that is not a pile of bills. Global bond markets dwarf the tokenized Treasury pocket, and allocators already mix cash, intermediate credit, and longer government paper in ordinary accounts. Copying that mix onto new rails is a reasonable project. The mistake is announcing the copy as if the original pile of assets had moved. It has not. A door opened. The room behind it is still being furnished, and the furniture, so far, is a reserve of shares rather than a relocated fund.
If you remember one distinction from the day nBND went live, remember that one. The bond ETF is large. The vault’s claim on it is a separate, still unpublished quantity. Everything else in the story hangs off that split.
