Table of Contents
If you hold alpha in a subnet, you're part of a competition for Bittensor's newly minted TAO. Every block, the protocol divides its new supply and routes more to some subnets, less to others. The formula running that competition, TaoFlow, has an opening: operators who understand how to time liquidity flows extract more emission than their subnet's actual market value warrants.
Const published a technical design for a new formula in PR #2781. The proposal introduces three scoring components: the subnet's smoothed alpha price, a multiplier that favors newer subnets, and a penalty for subnets that withhold miner rewards. The documentation states the upgrade "does not have any effect on TAO/Alpha holders, wallet interfaces etc." if and when it deploys.
We will explain below how the new formula is designed and how it determines which subnets earn more of Bittensor's new supply.
Why TaoFlow Rewards Timing Trades Over Real Value
Every block, Bittensor creates new TAO and distributes it across subnet operators, miners, validators, and stakers. Your staked alpha earns you a portion of what flows to your subnet. TaoFlow determines each subnet's portion by tracking how much TAO moves into and out of its liquidity pool over a rolling 30-day window.
Tracking movement rewards anyone willing to manufacture it artificially. An operator with enough capital buys into their own subnet's pool, pushes the inflow signal positive, and collects the emission lift. Then they sell out before the 30-day window resets and move the same capital to a different subnet. They return when the first pool's memory clears, and repeat.
@unconst put the behavior plainly in the proposal: "Cycling emission between your own subnets was a strategy." The new design closes this loop. A purchase and a sale of the same subnet's alpha move the emission signal in equal and opposite directions, canceling each other out regardless of timing. There's no window to exploit.
How the New Formula Is Designed to Score Subnets
The proposed formula has three parts, and each one does a specific job. Together, they score subnets on things worth measuring.
The first is smoothed price. This is the weighted average price buyers have been paying for your subnet's alpha over recent weeks, not the live tick from one large purchase yesterday. The smoothing works like a store's monthly revenue versus a single big sales day: one unusual transaction doesn't move the picture. Subnets whose alpha has been consistently trading at a high price have earned it, and the formula rewards them proportionally.
The second is root proportion, and this one is designed to give newer subnets a fair shot. It works like draft picks in a sports league: newer competitors get structural help at the start, because they're still building something from scratch. Mathematically, it's a ratio based on how much alpha a subnet has ever issued. A subnet launched last month carries a root proportion close to 1.0.
A two-year-old subnet with a large alpha supply might carry 0.1. As subnets mature, the structural advantage fades. Established subnets compete on price alone.
The third is miner burn. Each accounting window, the design checks what fraction of mining rewards reached real miners versus what the subnet owner kept or burned. Subnets with genuine mining competitions score zero on this and keep their full emission share. Subnets where the owner captures mining rewards lose emission in exact proportion to the amount diverted.
How the Design Would Check If Your Miners Are Getting Paid
Before this proposal, the chain has no way to distinguish a subnet running a real mining competition from one where the owner collects all the rewards. Both look identical from the outside.
Miner burn is designed to change that picture. Think of it like a delivery platform checking whether the restaurant is actually paying out driver tips. If the restaurant pockets the tips instead, the platform sends fewer orders its way. Here, if a subnet owner routes mining rewards to themselves, the formula routes fewer new emissions to the subnet. Validators would enforce this in normal conditions. The Triumvirate, Bittensor's governance oversight body, would act when the validator set is inactive.
One line in the proposal addresses a long-standing imbalance directly: "Mining subnets should no longer sit at a disadvantage to non-mining ones." Under TaoFlow, a subnet with no active mining receives emission without any penalty. The new design addresses this.
What the Proposal Would Mean for the Alpha You Already Hold
The design documentation is direct about scope: "this upgrade only effects subnet owners and dynamic TAO traders, it does not have any effect on TAO/Alpha holders, wallet interfaces etc." Nothing about how you stake or manage your positions would change.
What would shift over time is where Bittensor's new supply concentrates. Subnets with genuine market value in their alpha and active mining competitions would earn a growing share of new TAO. Subnets relying on timing trades or passive ownership arrangements would earn proportionally less. If the subnets you're staked in are building something worth holding, this formula is designed to work in their favor, and by extension, yours.
A Better Competition, By Design
The proposal addresses two limitations of TaoFlow simultaneously: symmetric price accounting removes the timing-based optimization, and miner burn introduces the first protocol-level enforcement against subnets withholding miner rewards.
Under the proposed formula, Bittensor would ask the same two questions about every subnet in the network: does the market believe this subnet's alpha is worth something, and are its miners getting paid? Subnets answering yes to both would earn more. That's the competition the design sets up, and it's the right one.
Disclaimer: This article is for informational purposes only and does not constitute financial, investment, or trading advice. The information provided should not be interpreted as an endorsement of any digital asset, security, or investment strategy. Readers should conduct their own research and consult with a licensed financial professional before making any investment decisions. The publisher and its contributors are not responsible for any losses that may arise from reliance on the information presented.
