Net TAO Flow: A Smarter Way to Measure Network Activity

By: vaN

Bittensor’s next update introduces a quiet but major shift. It changes how the network decides where TAO emissions go, moving from static price metrics to real movement on-chain.

The Problem with Price

Until now, subnets earned emissions based on price, the amount of TAO staked in a subnet’s pool relative to the total network.

It looks dynamic, but it’s built on history.

Price shows how much TAO has flowed in over time, not what’s happening right now. That means:

• A subnet can still appear healthy even if TAO is leaving.
• Emissions can reward past performance instead of live participation.
• “Price strength” often lags behind reality.

The result is an incentive layer that’s slightly out of sync with network behavior.

Flow Tells the Real Story

The new algorithm focuses on Net TAO Flow, how much TAO is entering or leaving each subnet in real time.

Each block, it:

• Measures the net flow (buys minus sells).
• Applies an exponential moving average to smooth out noise.
• Normalizes results across all subnets.
• Allocates emissions based on these live inflows.

If TAO is flowing in, emissions increase. If TAO is flowing out, emissions decline.

This simple feedback loop keeps rewards tied to real network movement.

Why It Matters

This isn’t about punishing volatility. It’s about aligning emissions with genuine demand.

In the example shown, around t = 90:

• TAO flow turned negative.
• The price metric stayed high because of older inflows.
• Under the current system, that subnet still earns emissions.
• Under the new one, it doesn’t.

That single change makes emissions fairer, faster, and harder to game.

Healthier Incentives

Flow-based emissions reward subnets that are actually attracting and holding TAO. It encourages builders to maintain consistent inflows instead of relying on early hype.

Key effects:

• More responsiveness: emissions reflect live activity.
• Less bias: historical price no longer dominates.
• Tunable behavior: parameters like smoothing and normalization can be adjusted without manual fixes.

It’s a small algorithmic change with wide-reaching impact on subnet economics.

The Takeaway

The Net TAO Flow algorithm values movement over memory. It brings the incentive layer closer to how real markets behave.

• Subnets attracting TAO in real time will gain emissions.
• Subnets losing TAO will see rewards taper naturally.
• The network’s emissions become a live indicator of capital flow, not a snapshot of the past.

Simple, transparent, and aligned with where Bittensor is heading.