The Institutional Series: What the Wrapper Reveals
When a new asset class gets absorbed into a regulated product structure, the economics suddenly become visible. Digital assets are going through that process now, and the filings tell you about the state of the market.
There is a moment in the life of every asset class when it stops being something institutions build for themselves and becomes something they buy. The activity does not change but what sits inside a product structure with disclosure obligations, and the economics of everything underneath that structure become legible to anyone who reads the document.
This same thing happened when fee disclosure arrived in fund management, and what happened in structured credit when rating agency frameworks and standardised documentation replaced bilateral negotiation. In each case the effect was similar and reasonably quick: within a couple of cycles, the economics of the entire chain had reorganised around the fact that they were now visible.
What Is Now on the Record
The clearest example is in the proof-of-stake products, because they are the ones where the wrapper has to account for an operational activity rather than simply hold an asset.
Grayscale's Ethereum Staking ETF distributed the proceeds of staking rewards to shareholders in early January 2026, the first time a US spot crypto ETP had done so. BlackRock and Grayscale now run staking-enabled Ethereum products, and Fidelity filed in August 2026 to add staking to its own spot Ethereum fund. The SEC issued a Commission-level interpretation in March 2026 setting out a taxonomy for digital assets and addressing the application of the securities laws to protocol staking specifically.
The same process is running in parallel elsewhere in the market, on a different timeline. Tokenised government money market funds have moved from novelty to cash management line item. Deposit tokens are letting banks digitise their own liabilities on programmable rails. Each of these is the same transition: an activity that was operationally bespoke becoming a product with published terms.
I want to focus on the staking products for the rest of this piece, not because they are the most important part of the market, but because they are the furthest along this particular curve and therefore show most clearly what the process does.
The First Public Price for the Layer Underneath
The most consequential disclosure in these filings is the reward split.
Fidelity's August filing sets out that the fund retains 85% of gross staking rewards, with the remaining 15% divided between the sponsor, the custodians and the node operators, with quarterly cash distributions planned but not guaranteed. Morgan Stanley proposed a 95/5 split for its Ethereum and Solana products in June.
Set aside which is better for an investor, which depends on the rest of the fee stack. The structural point is that the cost of the infrastructure layer is now a disclosed, comparable figure in a regulated document. Almost nothing else in digital assets works this way. The rest of the stack is priced privately, bilaterally, and without reference to any benchmark.
Anyone who lived through fee transparency arriving in traditional asset management knows what follows. Issuers compete on the visible number, because a spread that wide between two large managers does not survive a comparison table. The providers underneath get compressed accordingly. And the compression favours whoever can operate at institutional standard at scale, which is to say it consolidates.
That is the same dynamic I described in the second piece in this series as a prediction. It is now observable in public filings, which is a considerably better basis for it.
The Named Counterparty
The second disclosure matters more for how institutions should actually evaluate these products. Fidelity's filing names its intended node operators.
This is where the digital asset version of the wrapper differs from most of its predecessors, and it is worth being precise about why. A gold ETF absorbed the operational problem of vaulting and assay, and once absorbed, it was genuinely gone: the metal sits still. A private credit vehicle absorbed underwriting, but the underwriting was the manager's own. In a staking product, the wrapper is passing through a return that is being generated continuously by a named third party operating infrastructure, where performance determines how much return arrives and where failure has a tail that is not symmetric with the upside.
So the diligence perimeter has moved rather than closed. An allocator buying a spot bitcoin ETF could reasonably treat the operational layer as the issuer's problem. An allocator buying a reward-bearing product is taking a view, whether consciously or not, on a counterparty named in the prospectus.
This is an argument that reading the operational disclosure has become part of evaluating them, and that the discipline required is the ordinary discipline of counterparty assessment rather than anything exotic.
The Arithmetic That Does Not Appear in the Marketing
Here is the calculation an allocator should run, and that no issuer's materials set out in one place.
Start with the gross network reward. Ethereum's staking APR has been running at roughly 2.78%, with about 38.9 million ETH staked, close to a third of supply, across roughly 897,000 active validators as at May 2026. By mid-August 2026 the active staked balance had risen further, to around 42.2 million ETH.
From that, deduct the reward split. Deduct the sponsor's management fee, which sits on top and is charged against assets rather than against rewards. Then deduct the drag from whatever proportion of the fund cannot be staked at all, because it is held back to meet redemptions, expenses and liquidity. Fidelity's filing contemplates staking up to 100% of holdings under normal conditions, retaining only what is needed for those purposes but "up to" carries real weight in that sentence, and the proportion actually staked is the variable that decides the outcome.
Run those deductions and what reaches the end investor is materially below the headline network rate. It remains a return, and it remains operationally generated rather than the product of a directional view. But an institution underwriting a 3% to 6% figure taken from a two-year-old presentation is underwriting a number that no longer describes this market.
I would rather set that out plainly than have an allocator discover it after committing. Whether the resulting net figure is attractive is a question about alternatives, and it is the subject of the next piece.
A Familiar Liquidity Problem in Unfamiliar Clothing
The feature of these products that gets almost no attention, and that I think is the most interesting design question in the market, is liquidity.
The structure offers daily creation and redemption. The asset underneath is subject to protocol-level rate limiting in both directions. Following the Pectra upgrade, Ethereum's exit capacity is capped at 256 ETH per epoch, roughly 57,600 ETH per day, and the queue runs strictly first in, first out, with no priority processing regardless of the size of the stake or the standing of the institution behind it. Once a validator exits, a further delay of around 27 hours follows before the assets become available, during which nothing is earned.
Any allocator who has looked at an open-ended property fund, or a daily-dealing credit fund, will recognise the shape of this immediately. A liquid wrapper over a less liquid underlying is one of the oldest structural problems in fund management, and the industry has a long and occasionally painful record of learning where the limits sit.
Two things make the digital asset version different, and on balance I think they make it more tractable rather than less. The constraint is protocol-level and mechanical rather than discretionary, so it can be modelled precisely instead of estimated. And it is publicly observable in real time, by anyone, at any point.
What the observation shows this year is the opposite of what most institutional readers assume. In May 2026 the entry queue held roughly 3.59 million ETH with a wait of over 62 days, while the exit queue sat at zero. By mid-August the entry wait had fallen to around 39 days, with two validators and 64 ETH between them on the exit side. Demand from spot Ethereum ETFs and corporate treasury holders has been a significant driver of that entry pressure.
The received view is that staking is illiquid on the way out. At present the binding constraint is on the way in: capital committed today waits over a month before it earns anything, which is a genuine drag on a fund taking continuous inflows. The exit side is currently clear. It has not always been, and it is the side that would congest under stress, at precisely the moment a fund most needs it not to. April 2026 showed the mechanism, when security incidents in restaking and DeFi pushed volume onto the exit side quickly.
So the design question for any reward-bearing product is what proportion is staked, what buffer is held unstaked, and what happens if redemptions exceed the buffer while the queue is congested. That is an ordinary liquidity mismatch question. It has good answers but it is simply not being asked in most of the coverage of these launches.
What to Take From This
Three things, which apply well beyond staking.
When an asset class gets wrapped, the wrapper does not eliminate the operational complexity underneath. It relocates it to a named counterparty and, if the disclosure regime is doing its job, prices it in public.
The relevant comparison between competing products is rarely the headline number. In this case it is the reward split, the fee, the proportion actually deployed, and the quality of the operators named in the filing. Three of those four are now published.
And the analytical tools required are the ones institutions already have. Fee stack analysis, counterparty assessment, liquidity mismatch modelling. New context but same discipline.
Which leaves the question the whole exercise is in service of, and which I have deliberately not answered here: once you have run the arithmetic, is the resulting return worth having, set against what you already own?
Next week: I'll be taking a look at staking rewards against the fixed income alternatives in a cutting cycle, and why the direction of travel matters more than the level.
