Hi Zeeps and Shielded Labs Team,
I have been reviewing the recent dynamic fee proposal based on the 50-block median. While it elegantly addresses short-term mempool congestion and sandblasting attacks, it suffers from a fundamental macroeconomic flaw: It operates in a vacuum decoupled from real-world fiat purchasing power.
If the market exchange rate of ZEC increases by 10x or 100x, a static or congestion-only median fee format arbitrarily balloons user fees in real-world fiat value. This creates an economic paradox—the higher the token price, the more everyday utility and peer-to-peer commerce are suppressed because checking balances or shielded transfers become too expensive in real-world terms.
To solve this without violating Zcash’s absolute sovereignty (i.e., avoiding centralized third-party fiat oracles or untrusted price feeds), I propose a native, self-regulating feedback mechanism: An Endogenous Adaptive Fee Model based on Long-Term Macro Metrics and Differential Noise.
1. Core Economic Engine: The Network Wealth Density Formula
Instead of tying fees to historical fee rates (which creates a feedback loop), the protocol should derive its fee structure from the slow-moving, rigid macro variables representing the absolute productive floor of the economic network: The Market Maintenance Total Cost (\(C_{m}\)).
\(C_{m}=\text{Hardware/Node\ Operational\ Baselines}+\text{Staking/Emission\ Yield\ Demands}\)
Because \(C_{m}\) is determined by massive physical hardware setups, capital locked in staking pools, and long-term smooth emission parameters (ZIP 234), its value is rigid and changes slowly over time.
The Theoretical Base Fee (\(F_{theory}\)) required for a standard transaction is computed natively by every node at a given block height using a 30-Day Weighted Moving Average (30-day WMA) against the current smooth log-emission pool base, applying a mandatory 1.3 Premium Multiplier to guarantee a permanent 30% operational profit margin for network validators:
\(F_{theory}=\frac{\text{30-Day\ WMA}(C_{m})}{\text{Current\ Block\ Emission\ Base}}\times 1.3\times K\)
- Self-Stabilizing Velocity: As the speculative fiat value of ZEC climbs, the internal accounting parameters reflecting the network’s asset velocity and production thresholds expand. The denominator (emission base matching current log-decay) scales proportionately, forcing the nominal fraction of ZEC needed per transaction to mathematically compress. This locks the real-world user fee to a perfectly flat, predictable physical horizontal line.
- Anti-Manipulation Shielding: Attempting to manipulate this baseline by flooding the network for 24 or 48 hours is rendered economically obsolete. Because the mathematical anchor is stretched over a monthly weighted macro window, short-term data spikes are smoothed out, causing malicious spam vectors to be financially self-destructive to the attacker.
2. Defending zk-SNARKs via Randomized Privacy Amortization (Noise Injection)
In a zero-knowledge architecture (zk-SNARKs), charging a precise mathematical fee (e.g., 0.00012345 ZEC) creates an explicit privacy side-channel. Malicious chain-analysis nodes can parse public blocks and link transaction graphs across the shielded pool by matching specific fee remainders.
To permanently break side-channel tracking without adopting rigid, unscalable rounding rules, this proposal deploys a dual-layer verification protocol:
- Wallet-Level Noise Injection: The native user wallet calculates \(F_{theory}\) and automatically injects a localized differential privacy noise variable (ε). The final fee embedded in the payload becomes a randomized approximate number:
\(F_{final}=F_{theory}\pm \epsilon \) - Consensus Interval Verification: Validation nodes do not check for an exact scalar match, which would break the consensus state machine. Instead, the consensus code executes a strict deterministic range check. A transaction’s fee payload is deemed valid if and only if it falls precisely within a set margin around the premium target (e.g., between the 1.25 and 1.35 bounds):
\(\text{IF\ }(F_{final}\ge F_{theory}\times \frac{1.25}{1.30})\text{\ AND\ }(F_{final}\le F_{theory}\times \frac{1.35}{1.30})\rightarrow \text{VALID}\)
3. Amortization and Network Integrity
Because the injected noise (ε) alternates symmetrically, it functions as a zero-mean statistical variable across macro horizons. Over millions of network transfers, individual deviations wash out perfectly. User expenses and validator staking yields converge smoothly to the exact 1.3 premium target target mapped out by the core formula. Meanwhile, to external blockchain analysis tools, every fee payload appears as chaotic, uncorrelated statistical noise, permanently blinding metadata tracking.
Conclusion
This model achieves total economic sovereignty. It does not look at the Fed, it does not call a USD Oracle, and it does not rely on crude heuristics like mining difficulty. It reads Zcash’s own internal organic heartbeats to protect everyday users from being priced out during mega-bull runs, while turning staking yields into a reliable asset anchor.
I would love to hear thoughts from the core developers and cryptographic researchers on integrating a monthly endogenous macro average paired with wallet-level differential privacy noise into our upcoming roadmap.
Best regards,
EndoArchitect