Reduce Babylon Unbonding Fee Rate from 100 to 30 sats vByte

Reduce Babylon Unbonding Fee Rate from 100 to 30 sats vByte - Parameter Change Proposal

Summary

This proposal recommends reducing the Babylon unbonding transaction fee rate from 100 to 30 sats/vByte. Based on comprehensive analysis of Bitcoin transaction data spanning April 2023 to April 2025, this change would maintain reliable transaction processing while reducing costs by 70%, making the protocol more accessible and cost-efficient for users.

Parameters Upgrade Proposal

The unbonding fee will be reduced through a parameter upgrade proposal introducing the following new version of the x/btcstaking module parameters:

{
  "covenant_pks": [
    "d45c70d28f169e1f0c7f4a78e2bc73497afe585b70aa897955989068f3350aaa",
    "4b15848e495a3a62283daaadb3f458a00859fe48e321f0121ebabbdd6698f9fa",
    "23b29f89b45f4af41588dcaf0ca572ada32872a88224f311373917f1b37d08d1",
    "d3c79b99ac4d265c2f97ac11e3232c07a598b020cf56c6f055472c893c0967ae",
    "8242640732773249312c47ca7bdb50ca79f15f2ecc32b9c83ceebba44fb74df7",
    "e36200aaa8dce9453567bba108bdc51f7f1174b97a65e4dc4402fc5de779d41c",
    "f178fcce82f95c524b53b077e6180bd2d779a9057fdff4255a0af95af918cee0",
    "de13fc96ea6899acbdc5db3afaa683f62fe35b60ff6eb723dad28a11d2b12f8c",
    "cbdd028cfe32c1c1f2d84bfec71e19f92df509bba7b8ad31ca6c1a134fe09204"
  ],
  "covenant_quorum": 6,
  "min_staking_value_sat": 500000,
  "max_staking_value_sat": 500000000000,
  "min_staking_time_blocks": 64000,
  "max_staking_time_blocks": 64000,
  "slashing_pk_script": "agdiYWJ5bG9u",
  "min_slashing_tx_fee_sat": 100000,
  "slashing_rate": "0.001",
  "unbonding_time_blocks": 1008,
  "unbonding_fee_sat": 9600,
  "min_commission_rate": "0.03",
  "delegation_creation_base_gas_fee": 1095000,
  "allow_list_expiration_height": 139920,
  "btc_activation_height": 893362
}

This will be the fifth version of the staking parameters, following the three sets used in Phase-1 and the unused parameters introduced at Babylon Genesis (as staking registrations have not yet been enabled).

The only difference between the fourth and fifth versions are a reduced unbonding_fee_sat and a new btc_activation_height. The activation height was selected based on estimated Bitcoin block times to align with Monday, 21 July at 7am UTC.

Important Clarification: This parameter change will only affect new Phase-2 stakes. Existing Phase-1 CAP-1, CAP-2, and CAP-3 stakes will continue to use their original unbonding fee rates as established when they were created. This ensures that the economic parameters for existing stakes remain consistent with their initial conditions.

Dataset and Methodology

Our analysis examined a robust dataset comprising:

  • 106,816 Bitcoin blocks (2023-04-14 to 2025-04-14)
  • Filtered to focus on blocks with 100+ transactions (to exclude block anomalies)
  • Using the 10th percentile fee rate from each block as our inclusion metric
  • Multiple statistical approaches including rolling window analysis, CDF modeling, and worst-case wait time calculations

The 10th percentile fee rate represents a conservative threshold for transaction inclusion while avoiding distortions from private mempools and off-chain agreements. This metric provides a reliable indicator of the minimum fee needed for transaction inclusion in standard blocks.

Key Findings

Network Fee Distribution

  • Median inclusion fee rate: 9.03 sat/vByte
  • Mean inclusion fee rate: 25.29 sat/vByte
  • Standard deviation: 55.88 sat/vByte
  • Network conditions: Fees operate below 30 sat/vByte approximately 80% of the time!

The substantial difference between median and mean indicates a right-skewed distribution where occasional high-fee periods inflate the average. This suggests that most transactions confirm at significantly lower rates than our current 100 sat/vByte setting.

Confirmation Probability Analysis

For a 48-hour confirmation window (โ‰ˆ 288 blocks):

Probability Required Fee Rate (sat/vByte)
50% 8.02
75% 22.00
90% 50.15
95% 91.06
99% 201.00

This analysis shows the fee rates required for different probability thresholds within a 48-hour window. A 30 sat/vByte fee rate would provide approximately 80-85% probability of confirmation within 48 hours, based on interpolation between the 75% and 90% thresholds.

Critical Threshold at 30 sat/vByte

The data reveals a significant inflection point at 30 sat/vByte where worst-case wait times drop dramatically:

Fee Rate (sat/vByte) Longest Wait (days) Time Above This Rate (%)
10 113.1 47.6
20 96.1 29.6
30 7.7 20.0
40 4.8 14.6
50 4.7 11.4
100 (current) 2.9 5.1

The maximum wait time decreases from 96.1 days at 20 sat/vByte to just 7.7 days at 30 sat/vByte - a 92% reduction. This represents an optimal threshold where additional fee increases yield diminishing returns in confirmation reliability.

Rationale and Impact

The 30 sat/vByte rate offers an optimal balance between cost and reliability:

  • Cost Reduction: 70% lower fees compared to current 100 sat/vByte rate
  • Confirmation Probability: ~80-85% likelihood of inclusion within 48 hours
  • Efficiency Threshold: Dramatic improvement in worst-case wait times (96.1 โ†’ 7.7 days)
  • Network Compatibility: Operates above required network fee 80% of the time

The primary trade-offs are a slightly lower confirmation probability (80-85% vs. 95-97%) and longer maximum wait times during rare congestion periods (7.7 days vs. 2.9 days). For non-time-critical unbonding operations, these represent acceptable compromises given the significant cost savings and the infrequency of high-congestion events (< 20% of time).

Security Implications

This parameter change maintains the security properties of the protocol for several reasons:

  1. Slashability During Unbonding: Stakers remain slashable throughout the unbonding period, even if their unbonding transaction is in the mempool.

  2. Effective CPFP Mechanism: If an unbonding transaction becomes stuck due to congestion, the protocolโ€™s Child-Pays-For-Parent (CPFP) mechanism ensures slashing can still be enforced. Calculations show that the effective fee rate of the combined transaction packet would be approximately 156 sat/vByte (5x higher than the unbonding rate), ensuring the slashing mechanism remains effective even in high-congestion periods.

  3. Sufficient Time Buffer: The 7-day unbonding period provides adequate buffer for transaction confirmation even in worst-case scenarios (maximum observed wait time of 7.7 days at 30 sat/vByte).

While lowering the fee rate slightly reduces the slashing transaction fee margin, the high effective fee rate of the combined transaction packet and the extended unbonding period maintain robust protection against mempool congestion.

Conclusion

Based on comprehensive analysis of Bitcoin fee dynamics over a 2-year period, reducing the unbonding fee rate from 100 to 30 sat/vByte represents an optimal balance between cost and confirmation reliability. This change would significantly reduce transaction costs for users while maintaining reasonable confirmation times for the vast majority of network conditions.

The data strongly supports that 30 sat/vByte represents a strategic inflection point in the fee landscape, where wait times drop dramatically while still providing approximately 80-85% probability of confirmation within 48 hours.

4 Likes

Iโ€™ve seen some complaints this week on twitter about the unbonding fee rate being too high, so this is a timely proposal

great data to back up this 30 sats/vbyte metric!

starting from zero and working upwards, increasing the fee past 30 has quickly diminishing returns on Wait Time and is not worth it. The 30 number looks great

though now Iโ€™m curious what the drop off looks like between 20-30 sats/vbyte at smaller intervals; but for the sake of this proposal, I donโ€™t think we need to get that granular.

1 Like

Lowering the fee introduces a possibility that unbonding transactions may not be included in a block immediately. Historical data suggests this scenario is relatively rare, and the protocol includes mechanisms such as CPFP to address such cases. However, users may still experience inconvenience if their unbonding is delayed and BTC is not received in a timely manner. Despite this, a 70% reduction in transaction fees offers a significant benefit in terms of cost savings. This seems like a reasonable tradeoff, particularly for unbonding operations that are not time-critical.

One question I have is about the fee discrepancy between legacy stakers (CAP-1 to CAP-3) and new stakers in Phase-2. Since the new fee applies only to future stakes, it introduces an indirect imbalance. I would like to understand whether this was considered internally, or if it is not regarded as a major concern.

1 Like

@HwangjaeLee Thanks for the thoughts!

Regarding, your question, TL;DR: The Babylon Genesis chain design considers all Bitcoin Staking parameters versions as static, and therefore past versions utilized for Phase-1 cap-1, cap-2, and cap-3 stakers and the unbonding fees they contain cannot be altered without breaking key system assumptions.

The Babylon Genesis chain defines versioned parameter sets, each covering a range of Bitcoin block heights. A version specifies all required data for specifying the validity of a Bitcoin Staking transaction including the covenant members, the unbonding time & fee, min/max staking amount and more. The versions for a specific Bitcoin block height range are designed as immutable for the following key reasons:

  • Historical validity: A transaction that is mined in a specific Bitcoin block and is valid under the original BTC Staking parameters for this block could become retroactively invalid if those parameters changed (e.g., cap-1 allowed stakes up to 0.5BTC, whereas cap-3 allowed stakes up to 5000BTC).
  • Slashability guarantees: When a staker registers their stake on Babylon Genesis, they must supply two pre-signed slashing transactions, one spending their staking transaction, and the other the on-demand unbonding one. This happens to ensure slashability both during staking and the on-demand unbonding period. Changing the unbonding fee alters the entire unbonding transaction, leaving the pre-signed unbonding slashing transaction potentially pointing to a different transaction than the one the staker has previously collected signatures for.
1 Like

Caliber supports this proposal. We agree that the unbonding fee should be reduced to create a better experience for stakers.

1 Like

Thanks for being proactive and laying out such a detailed and well-supported proposal, @domo!!

During the early launch and growth stages, keeping fees low is crucial to onboarding users and creating a smooth experience. While a lower fee may slightly increase confirmation times, the reality is that most users today prioritize cost over speed.

As infrastructure operators, itโ€™s important that we advocate for user friendly defaults that also maintain protocol integrity. The data you provided shows 30 sat/vByte hits the right balance between affordability and reliability, without compromising slashing enforcement or protocol security.

Weโ€™ll be voting in favor of this change and support reducing the unbonding fee rate to 30 sats/vByte. This is a smart move for network usability and long-term adoption. :saluting_face:

1 Like

Got it, the situation is much clearer now.
Thanks so much for the detailed and kind explanation!