Emissions: the budget
Emissions are newly minted TAO distributed every block. Total issuance is fixed by protocol; how it splits across subnets is decided by network consensus. A subnet's emission share is its slice of that budget — the closest thing Bittensor has to revenue.
Emission share is the number to watch over months. It is also the number most exposed to governance: a share can be reweighted, and large shares are the natural source of budget for new subnets.
APY: the budget divided by the crowd
Delegation APY is roughly your share of emissions divided by the stake competing for them, after the validator's take. Two things follow immediately:
- If emissions stay flat and stake doubles, APY roughly halves. Nothing about the subnet got worse.
- If stake leaves while emissions hold, APY rises. Nothing about the subnet got better.
This is why a 29% APY and an 18% APY are not comparable. The 18% is usually attached to the subnet everyone already wants; the 29% is the price of holding something less liquid, more volatile, or both.
Alpha tokens: the exposure
Under dTAO, staking into a subnet converts TAO into that subnet's alpha token through a liquidity pool. Your position is denominated in alpha. Your realised return when you exit is: emissions earned, multiplied by whatever the alpha/TAO rate has done in the meantime.
Worked example. Stake 100 TAO into a subnet at 25% APY. A year later you hold alpha worth 125 TAO at the original rate — but the alpha token is down 20% against TAO. Your position is worth about 100 TAO. You earned a 25% yield and broke even.
Reverse it: the same 25% yield with alpha up 15% against TAO gives you roughly 144 TAO. Price did more work than yield. Over any horizon longer than a few months, alpha price movement usually dominates the yield term. Which is why "which subnet" matters more than "what APY".
How to compare subnets honestly
| Metric | What it tells you | Failure mode |
|---|---|---|
| Emission share | Durable claim on TAO issuance | Can be reweighted by consensus |
| Delegation APY | Current yield after validator take | Rises when stake exits |
| Alpha 7d/30d change | Whether the market is re-rating it | Narrative-driven, mean-reverts |
| Volume ÷ market cap | Whether you can exit at screen prices | Collapses exactly when you need it |
| Validator count | Operational and scoring resilience | Headline count can hide stake concentration |
Run any candidate through all five before you look at the yield again. Then model the compounding — the difference between annual and daily compounding at 25% over three years is not a rounding error.
TAO Compound is an analytics tool, not financial advice. Subnet figures referenced above come from the sample data set in this build; the live TAO price shown elsewhere on the site is from CoinGecko.