In early 2025, Bittensor went through its biggest economic upgrade to date: dynamic TAO, or dTAO. If you want to understand anything about staking on Bittensor today, including why some subnets yield more than others, this is the mechanism to understand.
Before dTAO: one token, one vote
Originally, TAO holders staked to validators on the root network, and those validators voted on how emissions were split between subnets. Influence over the network's resources was concentrated in a relatively small set of large stakers.
After dTAO: every subnet is a market
Dynamic TAO replaced voting with markets. Each subnet now has:
- Its own token, called alpha. Staking TAO into a subnet means swapping it for that subnet's alpha through a pool.
- Its own liquidity pool pairing TAO with alpha (an automated market maker, like a mini decentralized exchange).
- A price, the alpha/TAO rate, set by the flows in and out of the pool.
Emissions now follow these markets: subnets attracting more TAO earn a larger share of network emissions. In other words, capital allocation replaced politics. Stakers collectively decide which subnets deserve resources by putting their TAO where their conviction is.
What this means for your returns
When you stake into a subnet, your position has two engines:
- Staking yield: emissions distributed to the subnet flow to validators and their delegators, block after block, compounding automatically (a tempo is ~72 minutes).
- Alpha price: if the subnet attracts TAO after you, your alpha appreciates against TAO. If TAO flows out, it depreciates.
This is the fundamental trade-off of dTAO: subnet staking behaves less like a savings account and more like investing in an early-stage token with a yield attached. The price component can dominate the yield component in both directions, which is why an advertised APY never tells the whole story (we break that down in where TAO staking rewards come from).
Root staking still exists
Netuid 0, the root network, still lets you stake TAO without alpha exposure. It generally yields less than successful subnets, but you stay fully denominated in TAO. Many investors hold a base of root stake and allocate a risk budget to subnets on top.
The practical consequences
- Diversification matters. One subnet's alpha can crater on a bad narrative; a basket smooths that out. That's why multi-subnet strategies exist.
- Liquidity matters. Thin pools mean slippage when entering and exiting. Position size should respect pool depth.
- Timing matters less than discipline. Flows drive prices; spreading entries over time (DCA) avoids buying a spike.
dTAO turned Bittensor into an economy of 128+ competing markets. That is what makes it fascinating, and what makes tooling for it necessary. If you'd rather not manage the mechanics yourself, TaoScale wraps subnet selection, validator routing and execution into one-tap strategies, always non-custodial.