Is Bittensor SN88 Investing Worth It? A Miner's Guide (2026)

Short answer: probably yes, if you can build a real signal. If you're copying other miners, the similarity penalty will eat you alive.

This is the long answer — written while I was registering and deploying to SN88 UID 180 on April 22, 2026. If you're thinking about joining, here's what I wish someone had handed me before I spent the money.

What is SN88?

SN88 Investing (marketed as Mobius Fund) is a Bittensor subnet where miners submit subnet portfolio allocations and get scored on how well those allocations perform over time. The winning miners earn alpha rewards. Validators score the miners based on actual on-chain portfolio performance.

Think of it as a public alpha competition. Every miner runs a strategy, and the chain keeps score.

How miners get paid (the scoring formula)

After reading through the validator source and a few days of chain-data reverse engineering, the miner incentive boils down to roughly this:

score  =  MAR  ×  LSR  ×  odds%  ×  daily%

Where:

• MAR (MAR ratio): return divided by max drawdown. Rewards smooth winners, punishes miners who hit +50% then crash -40%.
• LSR (loss-scaled return): annualized return penalized by downside volatility.
• odds%: fraction of your daily portfolio values that beat a benchmark (usually BTC or TAO).
• daily%: fraction of days your portfolio was actually live and positioned.

Translation: you need consistent, low-drawdown returns that outperform a benchmark most days. Any single strategy that only works in one regime will lose its odds% multiplier when the regime flips.

The similarity penalty (the one that kills most miners)

SN88 does NOT reward copycats. If your portfolio correlates strongly with other miners, your score gets divided by a similarity factor. Two miners with identical portfolios split one miner's worth of reward between them.

This is the single biggest reason I almost didn't register. If everyone converges on the same "obvious" picks, everyone loses.

How to avoid it: pick subnets other miners aren't picking, or weight them differently. My strategy lands on 8 very small subnets (bottom 25% of the universe by tao_in) — a subset most momentum-chasers ignore entirely.

Registration: the numbers

As of April 22, 2026:

The 256 alpha minimum caught me off guard. You can't just register — you have to also stake a meaningful amount of SN88 alpha into the subnet before the validator will accept your strategy. At the price I paid, that was roughly $245 worth of SN88 alpha.

So the real all-in cost to start mining SN88 is closer to $300–$320, not the $42 registration fee.

The KYM (Know Your Miner) process

SN88 requires miners to prove their coldkey controls a registered account on kym.investing88.ai before strategy submission. The flow:

1. Register hotkey on SN88 via btcli subnet register --netuid 88
2. Sign up at kym.investing88.ai with a web3 wallet. Grab your web3auth ID.
3. Set your Bittensor coldkey's on-chain identity: btcli w set-identity --additional <web3auth_id> (costs 0.1τ)
4. Return to KYM, click "Claim coldkey" — it reads the on-chain identity field and matches.
5. Stake ≥256 alpha to SN88 from your coldkey.
6. KYM shows "Ready." You can now submit portfolios via the API.

Gotcha: the coldkey stake must come from the SAME coldkey that registered the hotkey. External transfers don't count toward the 256 minimum; you must actually stake into SN88.

Is the math worth it?

Here's the napkin version. At the time of writing, SN88 emission per day is roughly 4–6 alpha distributed across ~200 active miners. If you land in the top 20 of SN88's ranking, you can reasonably expect 0.05–0.15 alpha/day, which at current SN88 alpha prices is roughly $10–$30/day.

Break-even on the $300 startup cost: 10–30 days.

After that, it's pure income on a strategy that costs nothing to run once deployed.

But — and this is a huge but — that math assumes you land in the top 20. If your strategy is mediocre (bottom half of rankings), you'll earn dust. If you're in the bottom 10%, your stake might literally lose to inflation.

What kind of strategy actually works?

Three observations from scoring a year of Bittensor data against the SN88 formula:

1. Diversified portfolios beat single-subnet bets. Single subnets have too much drawdown to survive the MAR penalty.
2. Low-volatility wins over high-return volatility. A strategy returning +3% on 90% of days beats one returning +10% on 60% of days.
3. Small-subnet accumulation filters outperform momentum rotation. Momentum rotation gets whipsawed; quiet accumulation filters capture multi-week moves.

My strategy — SN88 Picks, documented in detail here — filters for small, cheap, active, not-yet-pumping subnets. Across 88 historical picks, 67 returned positive over 30 days. That's not guaranteed to carry forward, but it's the hypothesis I'm deploying real money on.

Should you register?

Yes, if:

• You have a real, original strategy (not "momentum rotation")
• You can stomach the $300 startup cost with a realistic 10–60 day break-even
• You're willing to iterate weekly as the similarity penalty shifts
• You accept that ~50% of miners don't break even in month 1

No, if:

• You're planning to copy someone else's portfolio (the dedup penalty will eat you)
• You can't afford to lose the $300 if your strategy turns out not to work
• You're looking for guaranteed passive income (nothing in crypto is that)

Where to start

• SN88 Investing main site: investing88.ai
• KYM portal: kym.investing88.ai
• btcli docs: docs.bittensor.com
• My strategy (live, with public track record): signals.subnetaiq.io

I'm running SN88 Picks publicly on UID 180 starting April 22, 2026. If the strategy survives its first 30 days on-chain, you'll see the results. If it doesn't, you'll see that too. No hype, no hiding.

— SubnetAIQ