关于「实践与证伪:定价试算器、经营者清单与预注册检验」,《责任图渗流与担保尾部破产》给出的定量结论是什么?
结论:第八章给出定价试算器(默认年度破产概率 0.00711,校正因子 1.25)、经营者清单与预注册检验方案。书中以标注框区分已发表的证据、可运行的 Python 计算(17 段程序)、原典与作者的假说。
数据口径:书中的资产负债表、损失分布与法律参数均为示意;全部计算由书中程序生成;本书不是个人化的法律、税务或财务建议,各法域规则不同,书中也不提供任何规避合法债务或逃避执行的方法。 书中以标注框区分已发表的证据、可运行的 Python 计算(17 段程序)、原典与作者的假说。
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: 0.00711, 1.25
| defaults: M | 8 s=0.1667 a=1.5 C=8.0 W=6.0 T=10 k=4 kg=1 pi=0.25 |
| formula yearly 0.00711, over T years 0.0689; Monte Carlo 0.00889 | |
| correction factor Monte Carlo / formula | 1.25 |
| yearly ruin, unlimited guarantees kg | 1 kg = k (all) |
| k | 1 pi = 0.00 0.01039 0.01039 |
| k | 1 pi = 0.25 0.01039 0.01039 |
| k | 2 pi = 0.00 0.00861 0.01721 |
| k | 2 pi = 0.25 0.00905 0.01551 |
| k | 4 pi = 0.00 0.00601 0.02406 |
| k | 4 pi = 0.25 0.00711 0.02064 |
| k | 8 pi = 0.00 0.00367 0.02939 |
| k | 8 pi = 0.25 0.00535 0.02464 |
| fair yearly fee rate, cap 2, one of 4 buckets | 0.0282 |
| fair yearly fee rate, cap 6, one of 4 buckets | 0.0160 |
查看完整验算程序
# Chapter 8. p81: the pricing calculator. Owner's yearly ruin chance from
# personal guarantees, by single-big-jump algebra for Pareto line losses
# P(X > x) = (s / x)^a, M lines split into k buckets of equity C / k:
# p = (1 - pi) * kg * (M / k) * (s / (C / k + W))^a * [G >= W]
# + pi * M * (s / (C + W))^a
# kg = buckets the owner guarantees, W = owner's liquid estate, pi =
# chance the group is treated as one estate. Fair yearly fee on a cap G
# for one bucket = (M / k) s^a [x^(1 - a) / (1 - a)] from C/k to C/k + G.
# A seeded Monte Carlo checks the formula at the defaults. Illustrative.
import math
import random
M, S, A, C, W, T = 8, 1 / 6, 1.5, 8.0, 6.0, 10
DEF = {"k": 4, "kg": 1, "G": math.inf, "pi": 0.25}
def p_year(k, kg, g, pi):
one = (M / k) * (S / (C / k + W)) ** A if g >= W else 0.0
return (1 - pi) * kg * one + pi * M * (S / (C + W)) ** A
def fee(k, g):
lo, hi = C / k, C / k + g
f = lambda x: x ** (1 - A) / (1 - A)
return (M / k) * S ** A * (f(hi) - f(lo)) / g
def mc(k, kg, pi, paths=200000):
rng = random.Random(20261112)
per, cap, hits = M // k, C / k, 0
for _ in range(paths):
xs = [S * (1.0 - rng.random()) ** (-1 / A) for _ in range(M)]
if rng.random() < pi:
pay = max(math.fsum(xs) - C, 0.0)
else:
pay = math.fsum(max(math.fsum(xs[b * per:(b + 1) * per]) -
cap, 0.0) for b in range(kg))
hits += pay > W
return hits / paths
pd = p_year(DEF["k"], DEF["kg"], DEF["G"], DEF["pi"])
print("defaults: M={} s={:.4f} a={} C={} W={} T={} k={} kg={} pi={}"
.format(M, S, A, C, W, T, DEF["k"], DEF["kg"], DEF["pi"]))
sim = mc(DEF["k"], DEF["kg"], DEF["pi"])
print("formula yearly {:.5f}, over T years {:.4f}; Monte Carlo {:.5f}"
.format(pd, 1 - (1 - pd) ** T, sim))
print("correction factor Monte Carlo / formula = {:.2f}".format(sim / pd))
print("yearly ruin, unlimited guarantees kg = 1 kg = k (all)")
for k in (1, 2, 4, 8):
for pi in (0.0, 0.25):
print(" k = {} pi = {:.2f} {:21.5f} {:12.5f}".format(
k, pi, p_year(k, 1, math.inf, pi), p_year(k, k, math.inf, pi)))
for g in (2.0, 6.0):
print("fair yearly fee rate, cap {:.0f}, one of 4 buckets: {:.4f}"
.format(g, fee(4, g)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《多子公司责任隔离图渗流、个人连带担保尾部破产放大与经营资产风险仓切分》第 8 章「实践与证伪:定价试算器、经营者清单与预注册检验」(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-177 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026177,
author = {Fan, Yafu},
title = {Liability Percolation across Subsidiary Graphs, Tail-Ruin Amplification from Personal Guarantees and Risk-Bucket Partitioning of Operating Assets},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}