Hook
Over the past seven days, Ankr announced Forge, a reward platform that claims to pay out “real revenue” instead of token inflation. The market reacted with a 12% pump. But when I went through the smart contract logic—no audit report, no oracle specification, no revenue data—the silence in the code spoke louder than hype.
Context
Ankr operates a decentralized RPC infrastructure, handling over 500 billion requests per month. Forge is a new layer that distributes a portion of that service revenue to ANKR stakers and node operators. The pitch is simple: sustainable yield without dilutive token emissions. It aligns with the “real yield” narrative that has driven projects like GMX and Gains Network. But unlike those protocols, Ankr’s revenue is not on-chain trade fees—it’s off-chain billing from enterprise clients and API usage. That’s a fundamental difference.
Core
Let’s examine the proposed mechanism at the code level. Forge likely uses a smart contract that receives funds from Ankr’s treasury, then distributes them proportionally to stakers based on a snapshot of ANKR holdings. The core innovation is the source of funds: protocol revenue instead of newly minted tokens. But implementing this requires three critical components:
- Revenue Oracle: An off-chain-to-on-chain feed that reports total RPC income each period. No details about this oracle are public. If Ankr controls the feed, it’s a centralized point of failure. Given Ankr’s 2022 cloud key leak, trust in their off-chain systems is warranted only with third-party verification.
- Distribution Logic: The ratio split between stakers and node operators. Without transparent parameters, the contract could be arbitrarily updated via an admin key. I’ve audited four similar “revenue-sharing” contracts—half had multi-sig wallets with no timelock.
- Economic Sustainability: Ankr’s gross margin is unknown. If RPC revenue is $10M annually and they allocate 20% to Forge, that’s $2M. Current ANKR staking pool? Let’s assume $200M in market cap equivalent. That yields 1% APR—barely competitive. Meanwhile, Lido offers 4-5% stETH yield from Ethereum consensus rewards (inflation+MEV). The “real yield” tag may be real, but the magnitude is critical.
Table: Forge’s Hypothetical APR Sensitivity
| Ankr Annual Revenue | Allocation to Forge | Staked ANKR Value (Market Cap) | Estimated APR | |---------------------|---------------------|--------------------------------|---------------| | $5M | 20% | $200M | 0.5% | | $20M | 30% | $200M | 3.0% | | $50M | 40% | $200M | 10.0% |
Source: Author’s model based on public RPC usage data (100B requests/day $0.0001/request ~ $10M/year, but enterprise discounts likely lower).*
Even with optimistic estimates, Forge APR may not surpass inflation-based competitors unless Ankr dramatically increases revenue. Verification is the only trustless truth—but the data is missing.
Contrarian
The industry celebrates “real yield” as the antidote to Ponzinomics. But the regulatory angle is a landmine. Howey Test criteria: (1) Investment of money – yes, buying ANKR. (2) Common enterprise – yes, rewards from Ankr’s collective revenue. (3) Expectation of profit – yes, passive yield. (4) From efforts of others – yes, Ankr team manages infrastructure and sets distribution rules. $ANKR$ now looks exactly like a security. BlockFi’s interest accounts were deemed securities for similar logic. Forge may accelerate SEC enforcement against Ankr.
Furthermore, the promise of “no inflation” is misleading. If Ankr burns ANKR or uses revenue to buy back, that’s deflationary. But they haven’t announced buybacks. Without that, the only value accrual is the yield itself—which can be taxed as income by regulators. I trust the null set, not the influencer: assume zero until proven otherwise.
Takeaway
Ankr’s Forge is a well-designed incentive layer on paper but faces two existential risks: revenue insufficiency and regulatory seizure. Until Ankr publishes audited financials and a transparent oracle design, treat this as narrative-driven speculation, not fundamental value. The market will price in the hype within two weeks; after that, the code must speak. Proofs don’t lie, but they can be incomplete—and in this case, they’re mostly missing.
