关于「手续费与 LVR 的盈亏平衡:覆盖率、测量噪声与拥挤的集中」,《做市损耗与三明治上界》给出的定量结论是什么?
结论:盈亏平衡只要两个数:手续费年化 \(f\,u\,365\) 与 \(\sigma^2/8\)。门槛波动率是 \(\sqrt{8\,f\,u\,365}\);30 个基点、每天换手 0.05 倍时,无摩擦门槛是 0.662,年化波动 0.8 的资产覆盖率只有 0.684。出块与手续费会抬高门槛,所以「手续费常常盖不住 LVR」要换成更诚实的说法:高波动、低不知情换手的池子,手续费很可能盖不住。测量本身也很吵,厚尾收益下一个月的数据有 0.416 的机会把亏损仓位读成盈利。集中在个体层面划算、在行业层面亏损。
数据口径:书中参数为示意,全部计算由书中程序生成。 本书是非个人化投资风险分析,不构成投资建议,不推荐任何代币、协议或交易;加密资产高风险,可能全部损失。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.05, 0.662, 0.684, 0.8, 30
| break-even annual volatility (fees = loss to arbitrage) | |
| fee bp turnover/day fee yield/yr sigma* LVR sigma* blocks | |
| 5 | 0.05 0.0091 0.270 0.493 |
| 5 | 0.25 0.0456 0.604 0.909 |
| 5 | 1.00 0.1825 1.208 1.586 |
| 30 | 0.05 0.0548 0.662 1.545 |
| 30 | 0.25 0.2737 1.480 2.764 |
| 30 | 1.00 1.0950 2.960 4.658 |
| 100 | 0.05 0.1825 1.208 3.374 |
| 100 | 0.25 0.9125 2.702 5.952 |
| 100 | 1.00 3.6500 5.404 9.855 |
| coverage ratio fee yield / (sigma^2/8) at sigma | 0.8 |
| 5 bp, turnover 1.00 | 2.281 |
| 30 bp, turnover 0.25 | 3.422 |
| 30 bp, turnover 0.05 | 0.684 |
| 100 bp, turnover 0.05 | 2.281 |
| reference window (Loesch et al.): fees 199.3, IL 260.1 (USD m) | |
| fees / IL | 0.766; share of LPs below zero 0.495 |
核心公式
\[\sigma^* = \sqrt{8\,f\,u\,365}.\]
查看完整验算程序
# Chapter 4. p41: break-even volatility between fee income and LVR.
# Model: a full-range pool turns over u of its value a day in fee-paying
# (uninformed) volume at fee tier f, so fee income is f u 365 a year per
# unit of value. Frictionless LVR is sigma^2 / 8 a year. They balance at
# sigma* = sqrt(8 f u 365).
# Second column: with 12-second blocks the arbitrage loss is about
# LVR * P_trade (chapter 2), so the break-even moves up; we solve
# f u 365 = sigma^2 / 8 * P_trade(sigma) by bisection.
# Turnover and tiers illustrative. The last lines restate the totals
# reported by Loesch et al. (2021) for one five-month window.
import math
DAY = 86400.0
def p_trade(sigma_yr, fee, gap=12.0):
sig_s = sigma_yr / math.sqrt(365.0 * DAY)
return 1.0 / (1.0 + fee / (sig_s * math.sqrt(gap / 2.0)))
def breakeven_frictionless(fee, u):
return math.sqrt(8.0 * fee * u * 365.0)
def breakeven_blocks(fee, u):
lo, hi = 1e-4, 50.0
target = fee * u * 365.0
for _ in range(200):
mid = (lo + hi) / 2.0
if mid ** 2 / 8.0 * p_trade(mid, fee) < target:
lo = mid
else:
hi = mid
return lo
if __name__ == "__main__":
print("break-even annual volatility (fees = loss to arbitrage)")
print(" fee bp turnover/day fee yield/yr sigma* LVR sigma* blocks")
for bp in (5, 30, 100):
for u in (0.05, 0.25, 1.0):
f = bp / 1e4
print(" {:>6} {:>12.2f} {:>12.4f} {:>10.3f} {:>13.3f}"
.format(bp, u, f * u * 365, breakeven_frictionless(f, u),
breakeven_blocks(f, u)))
print("coverage ratio fee yield / (sigma^2/8) at sigma = 0.8")
for bp, u in ((5, 1.0), (30, 0.25), (30, 0.05), (100, 0.05)):
f = bp / 1e4
print(" {:>3} bp, turnover {:.2f}: {:.3f}".format(
bp, u, f * u * 365 / (0.8 ** 2 / 8.0)))
fees, il = 199.3, 260.1
print("reference window (Loesch et al.): fees {}, IL {} (USD m)".format(
fees, il))
print(" fees / IL = {:.3f}; share of LPs below zero 0.495".format(
fees / il))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《集中流动性做市商重平衡损耗(LVR)、链上三明治抢跑边界与去中心化金库风控母典》第 4 章「手续费与 LVR 的盈亏平衡:覆盖率、测量噪声与拥挤的集中」(版本 1.0.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-344 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026344,
author = {Fan, Yafu},
title = {Concentrated-Liquidity Loss-Versus-Rebalancing, On-Chain Sandwich Bounds and Decentralised Vault Risk Control},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}