Bittensor’s Root Reborn: Validators Are Now Fund Managers

For most of Bittensor’s life, a quiet piece of plumbing sold the network’s tokens every block. Root stakers earned their yield in the alpha tokens of individual subnets, and the protocol converted that alpha into TAO automatically, block after block, whether or not anyone actually wanted to sell. That mechanical stream was a steady weight on every subnet token. Root Reborn set out to switch it off, and in the process it recast the people who run the network as something closer to fund managers.

As of this autumn it is no longer a proposal. Root Reborn is live, it has already changed shape more than once, and the first validators have published their picks. TAO itself trades near $281, with a market value around $3.2 billion that ranks it about 37th among crypto assets and still sits roughly 63% below the March 2024 high of $757.60, according to CoinGecko. The AI-crypto cohort ran hot through September, and the token rode that rotation higher. But the thing that decides how TAO behaves from here is not an ETF headline. It is the rebuild of the reward machine underneath it.

This piece walks through what Root Reborn is, what actually shipped in August and what stayed switched off, who is curating baskets now, whether any of it reduces the selling that has dogged the token, and why the network’s largest institutional validator, a Digital Currency Group subsidiary confusingly named Yuma, told the community to slow down.

The Selling Machine Root Reborn Was Built to Switch Off

Bittensor pays its block rewards in each subnet’s own alpha token, not in TAO. Within a subnet, the split is fixed: 18% to the subnet owner, 41% to miners, and 41% to validators and the stakers who delegate to them, as the network’s tokenomics lay out. People who stake TAO to the root network (netuid 0) earned a dividend too, denominated in subnet alpha. To pay that dividend in TAO, the protocol sold the alpha on the open market, mechanically, every single block.

The result was what analysts at ownyourmind call a constant, price-insensitive sell stream on subnet alpha: a flow that hit the order books regardless of whether buyers were there to meet it. That matters because Bittensor’s economy is almost entirely minted rather than earned. Value accrues through emissions and the appreciation of alpha tokens, which makes the loop circular rather than externally funded. After the December 2025 halving, the network issues roughly 3,600 TAO a day, about 12% to 13% annual inflation, with around 11.5 million of a capped 21 million TAO now in circulation.

So a great deal of fresh alpha is minted every day, and under the old design a large slice of it was dumped on schedule. That is the chronic sell pressure Root Reborn targets. The proposal projects up to a 33% cut in mechanical selling if the automatic sell-for-TAO cycle goes away, according to Crypto Briefing. Bittensor borrows Bitcoin’s fixed 21 million cap and halving rhythm, the same scarcity logic that governs the miners who compute Bitcoin’s hashes, but unlike Bitcoin it mints into a hundred separate token economies at once, each with its own thin market.

From GitHub Post to Live Protocol in Under Three Months

Root Reborn arrived on 17 June 2026 as a code proposal from a developer using the handle unconst. CoinDesk summed it up as a plan that would turn validators into something like fund managers, picking which subnets to back and reinvesting yield rather than constantly selling subnet tokens to pay stakers. The market’s first read was not relief but fear. TAO slid 7.8% to about $237.60 when the code hit GitHub, per TradersUnion, and by 19 June it was near $225, down roughly 20% from a $283 high four days earlier, with futures open interest falling toward $252 million to $260 million, crypto.news reported. Traders heard validators pick winners and priced in a new layer of discretion.

From there it moved fast. The sell-pressure fix went live on mainnet in August, the discretionary curation layer switched on weeks later, and the mechanism was rewritten again in September. The table below tracks the sequence.

What Root Reborn Actually Does

Strip away the fund-manager language and the core change is mechanical. When you stake TAO to the root network, you now delegate to a validator who runs a single fund. Every epoch, that fund’s root dividend on each subnet is credited straight into the fund’s holding on that subnet, with, in the words of the official documentation, no sell, no redeploy, no target vector. The alpha simply stays as alpha and compounds. The holdings are real stake sitting in a chain-owned escrow account that has no private key, so nothing can be stolen out of it.

What you own is not a specific pile of subnet tokens but a fraction of the fund, tracked as a beta token. The documentation describes a beta balance as a stable count that grows with accruals and shrinks on claim, while its TAO value moves with the holdings. Net asset value is calculated from real pool depth and fees, so a validator cannot inflate the number by pointing at a thin, illiquid market. When you claim, the protocol redeems your owed fraction of every holding pro-rata, preserving the fund’s composition, and converts it into TAO restaked to the same validator’s root position. A small minimum threshold stops the system from grinding through dust-sized payouts.

A concrete example helps. Suppose you stake 100 TAO to a validator whose fund holds alpha across a dozen subnets. Each epoch, your share of that fund’s dividends is added to its holdings and your beta-token balance ticks up, but no TAO changes hands and nothing hits the market. Six months later you decide to exit. The protocol takes your fraction of every position the fund holds, converts it to TAO at live pool prices, and hands it back as TAO restaked to that validator’s root position. Under the old design, those same dividends would have been sold for TAO on your behalf, epoch after epoch, whether you wanted the TAO or not.

There is a tax wrinkle buried in the design. Because alpha compounds inside the fund instead of being sold for TAO, the taxable moment is deferred: as Crypto Briefing notes, unrealized gains sitting in a basket are not taxable events in most jurisdictions until a holder claims. For a long-term staker, that is a genuine efficiency, and part of the reason the design is attractive beyond the sell-pressure story. It also reframes what a root staker is doing, from collecting a yield that is paid out and taxed as it arrives, to holding a position in a compounding fund whose gains are only crystallized on exit.

The Headline Feature Shipped Switched Off

Here is the detail that most launch coverage missed. The version that reached mainnet on 4 August, runtime v441, did not actually turn validators into fund managers. The marketing promised a competitive allocation layer where validators assemble subnet baskets and compete on returns, but that curation layer was gated off at launch. In the analysis published by ownyourmind, every basket fund initially ran a null accumulate-in-place strategy until a later upgrade enabled it. The dividends stopped being sold, but nobody was steering them anywhere.

Crucible Labs put the split plainly in late August: the mechanism intended to reduce mechanical subnet selling was live and testable, while the mechanism intended to direct capital toward productive subnets was not. In other words, at launch Root Reborn changed when yield is realized rather than who allocates it. The structural, market-moving part, ending the block-by-block sale of alpha, arrived immediately and applied to every subnet. The discretionary, fund-manager part was a placeholder waiting to be switched on. That distinction is easy to lose in a headline, and it matters, because the piece that moved the token was shipped and the piece that justified the fund-manager framing was not.

The First Baskets Are In

The switch came weeks later. According to Crucible Labs, a subsequent runtime turned on the allocation mechanism, letting each root validator publish a basket showing where its root yield should be redeployed. A guardrail caps any single destination at one sixteenth of a basket, which forces every curating validator to spread across at least 16 positions rather than piling into one favorite subnet.

By 30 August, 13 validators had filed their initial allocations. Nine of the 13 ranked among Bittensor’s top 25 validators at the time, including four of the top 10: tao.bot, Taostats, Arbos and Datura. More than a third of the network’s largest operators were curating within the first wave. There was also early herding in what they picked: a subnet trading as GM (SN28) appeared in 12 of the 13 baskets, with two others, Engy and Bitcast, in 11 each. The supposedly competitive, differentiated allocation layer opened with most of the big players crowding into the same few names, which is worth remembering when the design is sold as a market for validator skill.

Then the mechanism mutated again. The current runtime removed the passive allocation vector entirely and replaced it with active basket trading, giving validators direct control over composition. As the documentation explains, once validators could trade their holdings directly, a passive target vector that every inflow chased was redundant. Within about two months, curating a Root Reborn fund went from impossible, to publishing a standing list of weights, to actively trading a portfolio. That is a lot of change for anyone trying to model what their stake is doing, and it is the kind of churn that makes the design hard to pin down even for people watching it closely.

Validators as Fund Managers

With curation live, the fund-manager framing stops being a metaphor. A root staker subscribes TAO to a validator’s fund and holds beta-token fractions of it. The staker can redeem to TAO or move to a different validator at any time, so capital is not locked. The validator, meanwhile, now competes on the returns of the basket it builds. Pick subnets whose alpha appreciates and your fund’s net asset value rises, which attracts more root stake; pick badly and stakers walk.

That is a real change in what a validator is for. The old job was to run honest consensus under Yuma Consensus and collect a dividend. The new job layers portfolio management on top: judging which of the network’s subnets will actually deliver, and backing them with other people’s stake. It nudges Bittensor toward resembling a marketplace of on-chain, non-custodial index funds, each run by a validator and each denominated in the network’s own token. For holders weighing how to get exposure to subnet alpha, that is a new option alongside the ones that appeared when subnet tokens started leaving home for Base and centralized exchanges. The difference is that a basket keeps the alpha inside the protocol and compounds it, rather than exporting it to an outside venue.

Does It Actually Cut the Selling?

The honest answer is that it changes the timing and ownership of the selling, not the reason it happens. Removing the automatic block-by-block sale is a genuine improvement, and ownyourmind agrees it takes out a constant, price-insensitive sell stream on subnet alpha. The headline projection is up to a 33% reduction in mechanical sell pressure. But the same review is blunt that the change does not address the underlying emission-subsidy problem.

Think of it this way. Before Root Reborn, newly minted alpha was sold on a fixed schedule. After it, that alpha sits in a fund and is sold only when a staker decides to claim, or when a validator trades the basket. If stakers claim aggressively, the selling returns. If alpha broadly fails to appreciate, compounding simply defers the loss rather than preventing it. The reform is a dam, not a drought: it holds back a flow that will still move whenever holders want their TAO. Whether the dam reduces net selling over a quarter depends on behavior the protocol does not control.

The Emissions-vs-Revenue Gap That Makes Selling Inevitable

The deeper reason alpha gets sold is that the network mints far more value than its subnets sell to outside customers. The SubConnect Bittensor Revenue Index, published in September, counted 24 subnets generating external revenue, totaling $28 million to $35 million a year, with compute and infrastructure subnets supplying roughly 78% to 82% of it. That is a real and growing number, and 14 of the 24 now channel some of that revenue into buying back their own alpha. But set against emissions worth more than $300 million a year at current prices, external revenue covers only about 9% to 12% of what the network pays out.

Independent analysts are harsher still. ownyourmind’s settlement-based review finds only $0.85 million to $5.19 million of demand it can verify on-chain, against $18.9 million subnets report themselves, which it puts at about 4.7% of the minted budget, concluding that participant rewards are overwhelmingly inflation-funded on every tier. The clear exception is the GPU-rental subnet Lium (SN51), whose billing it tracks near $15.5 million annualized and up threefold year on year. The table shows where the money actually comes from.

The point is not that Bittensor earns nothing. It is that as long as minting dwarfs sales to real customers, someone has to sell the newly minted alpha to turn it into spendable TAO. Root Reborn changes who holds that alpha and for how long. It does not change the arithmetic that keeps producing it.

Why Yuma Says Slow Down (and Which Yuma)

First, untangle a name collision that trips up even regular readers. Yuma Consensus is the scoring algorithm at the heart of Bittensor, the math that decides how much each miner and validator is paid. The Yuma raising objections to Root Reborn is a different thing entirely: a company. It is a Digital Currency Group subsidiary launched in November 2024, described by The Block as DCG doubling down on decentralized AI. The firm validates well over 120 subnets and ranks among the network’s largest institutional stakers, so when it raises a flag, the community listens.

Yuma’s warning, delivered during the June debate, was that the change could “fundamentally alter the role of validators,” as crypto.news reported. Its specific concerns read like a risk memo. Turning validators into active capital allocators invites collusion, preferential treatment and frontrunning, the same discretionary-ordering problem that shows up anywhere insiders choose what gets filled first, as our look at who front-runs your trade on perp DEXs explored. Subnet teams might start courting validator relationships instead of building product. Basket positions could prove hard to liquidate in a sell-off, leaving late redeemers worse off. And validators directing capital start to look a lot like investment managers, which drags regulatory scrutiny onto the exchanges, custodians and institutions that deal with them.

There is an irony worth naming. DCG is also the parent of Grayscale, whose proposed spot TAO ETF is the most-watched institutional product in the ecosystem. One DCG arm is building a clean, passive vehicle for TAO, while another DCG arm is warning that the protocol’s new reward design makes validators look too much like asset managers. The same house is long the simple version of Bittensor and skeptical of the complicated one.

A Mechanism That Keeps Changing Is Its Own Risk

Even supporters of the economics have to reckon with the pace. In under two months, Root Reborn was three different mechanisms: an escrow-and-claim system in August, a published-weights allocation layer by late August, and an active basket-trading market by September, plus smaller tweaks to claim costs and, separately, a change that raised the cost of taking over a subnet. For anyone trying to underwrite the value of a subnet token, the goalposts keep moving.

Mark Creaser, chief executive of DSV Fund, made the point sharply earlier this year. A casino, he told tao.media, “at least keeps the same rules all night,” while “Bittensor changes the table mid-hand and calls it an upgrade.” His colleague Siam Kidd went further, calling dTAO basically uninvestable when the rules reprice retroactively. Root Reborn, however well-intentioned, is another data point in that complaint.

Who gets to change the rules is the other half of the story. ownyourmind notes that the Opentensor Foundation still validates every block under proof of authority and controls privileged operations through a single two-of-three multisig; on the median subnet it measured in September, one coldkey held 57.26% of the stake. Who should hold keys like that is exactly the question we put to DAO security councils across crypto. That concentration is also the backdrop to the decentralization-theatre charge that Covenant AI’s Sam Dare leveled when his team quit in April, telling The Block that the promise no single entity controls the network “is a lie.” Founder Jacob Steeves has acknowledged the network is not yet decentralized the way Bitcoin is and points to a handover target around the end of 2027, per Crypto Briefing. Seen in that light, Root Reborn is both a real economic fix and one more example of core-driven change landing fast.

The ETF Angle: A Fix US Funds Cannot Capture

Root Reborn quietly widens the gap between owning TAO on-chain and owning it through a fund. Grayscale has filed to convert its Bittensor trust into a spot ETF on NYSE Arca, with an S-1 lodged on 30 December 2025 and an amendment on 2 April 2026, under SEC filings that bar the product from staking and set a fee around 2.5%, with Coinbase and BitGo as custodians. A US spot holder would get the TAO price and nothing else: no emissions, no alpha dividends, no beta-token fund fractions, and now no share of the compounding baskets Root Reborn creates.

Europe already shows the other path. The Safello Bittensor Staked TAO ETP from Deutsche Digital Assets stakes the token and rolls the yield into net asset value, at a total expense ratio of 1.49%, as ETF Express detailed at launch. That is the staking paradox in one comparison: the European product captures network yield, the proposed US product cannot. Root Reborn sharpens it, because value that used to be sold into the TAO price every block, and therefore at least showed up in the number a spot ETF tracks, now compounds inside validator funds that only on-chain stakers can claim. No US spot TAO product has been approved; a thinned-out SEC and the government shutdown have kept the queue frozen. These are spot crypto-asset questions in the US, much as TAO sits outside the stablecoin rules at the center of MiCA enforcement in Europe.

What It Changes for Alpha Holders

For someone holding subnet alpha, Root Reborn adds a third way to hold exposure. You can keep loose alpha in a wallet, which is fully yours but illiquid and sells into thin pools. You can hold wrapped or exchange-listed alpha, which is liquid but adds bridge and venue risk. Or you can hold a beta-token slice of a validator’s curated, compounding basket, which is diversified and reinvests automatically but hands allocation decisions to a validator and prices off net asset value rather than a live order book.

Each carries a different bet. Loose alpha is a bet on one subnet. Listed alpha is a bet on one subnet plus the rails that carry it. A basket fraction is a bet on a validator’s judgment across many subnets, wrapped in a vehicle that looks, functionally, like an in-protocol index fund. The choice you make now determines not just your upside but who is making decisions on your behalf, and how quickly you can get out if a subnet you never chose starts dragging the basket down.

Open Questions and Risks

• Adoption breadth: only a handful of top validators curate so far, and they crowded into the same subnets. Does curation spread, or stay concentrated among the largest operators?
• Real sell-pressure impact: the 33% figure is a projection. Claim behavior and whether alpha actually appreciates will decide the real number.
• Favoritism and frontrunning: now that validators actively trade basket composition, Yuma’s core worry about preferential ordering has teeth.
• Liquidity and net asset value: baskets price off pool depth, so a rush of claims during stress could hit thin alpha markets hard.
• Still inflation-funded: external revenue is a single-digit to low-teens percentage of emissions, and Root Reborn does not change that.
• Regulatory framing: validators as allocators invite investment-management treatment, a live question for the SEC in the US and for MiCA and MiFID II in Europe.
• Mechanism instability: three versions in two months makes the design hard to underwrite, exactly the critique DSV Fund has pressed.

What to Watch Next

The next few months will show whether Root Reborn delivers. Watch whether curation spreads beyond the first dozen validators and whether active basket trading produces real dispersion in returns, the signal that a competitive market is forming rather than a herd. Watch on-chain data for any measured drop in alpha selling over the fourth quarter. Watch the SEC queue once the shutdown clears, since a Grayscale or Bitwise approval would pull institutional money toward the one version of TAO that captures none of this yield. And watch for Gamma, the inter-subnet credit idea floated at the Exploit Summit in Montreal, which would push the same reinvest-rather-than-liquidate logic from root staking into payments between subnets.

Strip out the noise and Root Reborn is the clearest test yet of a single question: can Bittensor keep its emissions working inside the network instead of bleeding them into the market? The sell stream is off. The fund managers are real. Whether that adds up to a healthier token, or just a more complicated one, is the thing the next two quarters will settle.

Frequently Asked Questions

What is Bittensor Root Reborn?

Root Reborn is a 2026 overhaul of how TAO root stakers earn yield. Instead of the protocol selling subnet alpha for TAO every block to pay a dividend, yield now accrues inside a per-validator fund and is realized only when a staker claims. It went live on mainnet as runtime v441 on 4 August 2026 and aims to cut mechanical sell pressure on subnet tokens.

Does Root Reborn reduce TAO sell pressure?

The proposal projects up to a 33% reduction in mechanical selling by ending the automatic block-by-block sale of subnet alpha. Analysts caution that the real figure depends on how often stakers claim and whether alpha appreciates, and that the change addresses the timing of selling rather than the network’s underlying reliance on emissions.

What does it mean that validators became fund managers?

Under Root Reborn, each root validator runs a fund that holds subnet alpha and compounds it. Stakers hold fractions of that fund as beta tokens and can redeem or switch validators at any time. Validators now choose and trade the basket composition, so they compete on investment returns, not just on running consensus.

Who opposes Root Reborn and why?

Yuma, a Digital Currency Group validator subsidiary and one of the network’s largest institutional stakers, warned the change could fundamentally alter the role of validators, creating risks of favoritism, frontrunning, hard-to-liquidate positions in a sell-off, and regulatory scrutiny that treats validators like investment managers.

How is Root Reborn different from a TAO ETF?

A US spot TAO ETF such as Grayscale’s proposed GTAO tracks the TAO price and bars staking, so holders receive no emissions, alpha dividends or basket yield. Root Reborn keeps that yield compounding inside validator funds that only on-chain root stakers can claim, widening the gap between direct network participation and passive ETF exposure.

By Marcus Okafor, HOGE Wire markets desk.