How harbor prices
Four pricing models, replayed on the same withdrawal history.
Historical simulationEthereum · Lido withdrawals01 Jul 2024 – 08 Mar 2025
LP profit316.61WETHbest of 4 models
Trades below cost3.0%best of 4 models
Valuation error0.92bpbest of 4 models
Trades simulated5,98516,452 receipts valued
LP profit
total, WETH- harbor
- 316.61
- fixed-delay
- 216.05
- age-based
- 257.58
- queue-aware
- 221.22
- 5% hurdle
- 342.77
Profit per trade
trades below cost · lower is better- harbor
- 3.0%
- fixed-delay
- 14.5%
- age-based
- 5.4%
- queue-aware
- 3.9%
Valuation error
avg miss · lower is better- harbor / queue-aware
- 0.92 bp
- fixed-delay
- 1.46 bp
- age-based
- 1.37 bp
Who gains
LP gain vs queue-aware, WETH- harbor
- +95.39
- fixed-delay
- −5.18
- age-based
- +36.36
- queue-aware
- 0.00
How it was tested
- harbor
- Queue-aware valuation with FACE capacity pricing.
- fixed-delay
- One historical settlement-time estimate for every receipt.
- age-based
- Remaining settlement time estimated from receipt age.
- queue-aware
- Receipt age and queue conditions, without a capacity adjustment.
- Capital
- 10,000 WETH
- Funding
- 5% / yr
- LP margin
- 2 bp
- External fee
- 10 bp
- Operations / receipt
- 0.0002 WETH
- Seller gas / receipt
- 0.0002 WETH
- Settlement lag
- +6 h
| Model | LP profit | After funding | Seller proceeds |
|---|---|---|---|
| harbor | 316.6089 | 214.3689 | 596,940.6266 |
| fixed-delay | 216.0479 | 113.7800 | 597,041.0871 |
| age-based | 257.5842 | 155.3233 | 596,999.5924 |
| queue-aware | 221.2229 | 118.9635 | 597,035.9172 |
Simulated on history, not observed harbor trades or future returns. harbor capacity: 60% utilization target, kappa 25 bp, up to 64 receipts. The test period was already inspected.
Reconciled research data · Graph indexing pending
- History
- 73,760 requests · 69,309 claims · block 22,000,000
- Dataset
- 5ae28ad25cbf235f25e2439fb698c42d4bf3aafdacaef75cc4b11b238c626344
- Benchmark
- 221cee8b0441f20bff514f0e8fde54f0fba0fbe2f1c9d45404d8b34892b38c57