What does LVR, Sandwich Bounds and Vault Risk conclude, in numbers, about “The σ²/8 Rate of Loss-Versus-Rebalancing, Discounted by Fees and Block Times”?

Answer: Chapter 2 derives σ²/8 and measures the discount from fees and block times.

Scope of the numbers: Parameters are illustrative and all calculations are generated by the book's programs. This is general investment risk analysis, not investment advice; it recommends no token, protocol or trade, and crypto-assets are high risk and can be lost in full.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python only

A. sigma = 0.8 a year; arbitrage loss per year per unit value
stepsteps/yrexactMonte Carlosigma^2/8
1 day3650.0799910.0802530.080000
1 hour8,7600.0800000.0799170.080000
12 sec2,628,0000.0800000.0799720.080000
B. LVR budget sigma^2/8 at several volatilities
sigma/yrper yearbp per daydaily sigmaon 1,000,000/yr
0.2000.00500.1370.01055,000
0.4000.02000.5480.020920,000
0.6000.04501.2330.031445,000
0.8000.08002.1920.041980,000
0.9550.11413.1250.0500114,063
1.2000.18004.9320.0628180,000
1.5000.28127.7050.0785281,250
Show the full program
# Chapter 2. p21: LVR of a constant-product pool is sigma^2 / 8 per unit
# of pool value per unit time.
# Model: the market log price moves by a normal step with mean -s^2/2 and
# variance s^2 = sigma^2 dt (a martingale price, geometric Brownian
# motion). Arbitrage resets the pool each step and earns, per unit of
# pool value, (sqrt(q/p) - 1)^2 / 2. Its exact mean is 1 - exp(-s^2/8),
# which tends to s^2 / 8 as the step shrinks. A fixed-seed Monte Carlo
# checks the exact mean. Volatilities are illustrative annual figures.
import math
import random


def exact_step(s):
    return 1.0 - math.exp(-s * s / 8.0)


def mc_step(rng, s, n):
    out = []
    for _ in range(n):
        z = rng.gauss(-s * s / 2.0, s)
        out.append((math.exp(z / 2.0) - 1.0) ** 2 / 2.0)
    return math.fsum(out) / n


if __name__ == "__main__":
    rng = random.Random(20261120)
    sigma = 0.8
    print("A. sigma = 0.8 a year; arbitrage loss per year per unit value")
    print("  step       steps/yr  exact      Monte Carlo  sigma^2/8")
    for name, k in (("1 day", 365), ("1 hour", 8760),
                    ("12 sec", 2628000)):
        s = sigma * math.sqrt(1.0 / k)
        ex = exact_step(s) * k
        mc = mc_step(rng, s, 200000) * k
        print("  {:<9} {:>10,}  {:.6f}   {:.6f}     {:.6f}".format(
            name, k, ex, mc, sigma ** 2 / 8))
    print("B. LVR budget sigma^2/8 at several volatilities")
    print("  sigma/yr  per year  bp per day  daily sigma  on 1,000,000/yr")
    for sig in (0.2, 0.4, 0.6, 0.8, 0.05 * math.sqrt(365), 1.2, 1.5):
        yr = sig ** 2 / 8.0
        print("  {:>7.3f}  {:>8.4f}  {:>10.3f}  {:>11.4f}  {:>15,.0f}".format(
            sig, yr, yr / 365 * 1e4, sig / math.sqrt(365), yr * 1e6))

Single source and citation

Fan, Y. (2026). Concentrated-Liquidity Loss-Versus-Rebalancing, On-Chain Sandwich Bounds and Decentralised Vault Risk Control (Version 1.0.0), Chapter 2: The σ²/8 Rate of Loss-Versus-Rebalancing, Discounted by Fees and Block Times. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 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}
}

Other chapter answers in this book