关于「三种冲击下的资金链断裂概率与圈存准备金率」,《负营运资金飞轮与预收安全边界》给出的定量结论是什么?

结论:第四至六章用固定种子蒙特卡洛说明,单一冲击很少压垮预售生意,叠加冲击才是准备金要防的东西,示意参数下审慎的准备金率红线是 0.474;每单定金交付前可动用于获客的比例上界是 0.475,回本只按已实现毛利计;并给出信任锚、定金比例权衡与逐周检出以新补旧的规则。

数据口径:书名中的「零本金」只指不借债、不融资,不等于没有风险;书中的营收、毛利、账期、退款率、获客成本与价格均为示意参数,全部计算由书中程序生成;本书是非个人化经营建议,不构成法律、会计、税务或投资建议,各地预付款、定金与储值规则以上架前人工复核为准,本书不涉及任何分润或返佣。 第四至六章用固定种子蒙特卡洛说明,单一冲击很少压垮预售生意,叠加冲击才是准备金要防的东西,示意参数下审慎的准备金率红线是 0.474;每单定金交付前可动用于获客的比例上界是 0.475,回本只按已实现毛利计;并给出信任锚、定金比例权衡与逐周检出以新补旧的规则。

English version

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: 0.474

point p <= 0.01rho* = 0.427(p = 0.0100, upper 0.0136)
Wilson upper <= 0.01rho* = 0.474(p = 0.0067, upper 0.0098)
deterministic floor from Chapter 5, (c + d - 1) / d = 0.500
查看完整验算程序
# Chapter 4. p42: the smallest reserve ratio with rupture risk <= 1%.
# Same weekly model and scenarios as p41 (copied so the program stands
# alone): seeds 20261120 + i, 52 weeks, three shocks. Rupture is monotone
# in the reserve ratio path by path, so bisection on rho is exact for the
# simulated sample. Two answers: the point estimate p <= 0.01, and the
# cautious one whose Wilson 95% upper bound is <= 0.01. N = 4,000 paths.
import math
import random

B, D, W, COST, FIX, SIG = 100.0, 0.6, 4, 0.7, 6.0, 0.15
HAZ = {"demand": 0.02, "refund": 0.01, "supplier": 0.01}


def scenario(i, weeks=52):
    rng = random.Random(20261120 + i)
    noise = [rng.lognormvariate(-SIG * SIG / 2, SIG) for _ in range(weeks)]
    hits = {}
    for name in ("demand", "refund", "supplier"):
        for t in range(weeks):
            if rng.random() < HAZ[name]:
                hits[name] = t
                break
    return noise, hits


def ruptured(noise, hits, rho, active=("demand", "refund", "supplier")):
    pipe = [[B, 0.0] for _ in range(W)]           # [value, prepaid share]
    cash = rho * D * W * B
    for t, z in enumerate(noise):
        demand = 0.45 if "demand" in active and t >= hits.get(
            "demand", 99) else 1.0
        pre = 0.5 if "supplier" in active and t >= hits.get(
            "supplier", 99) else 0.0
        n = B * demand * z
        flows = [D * n, -pre * COST * n]
        old = pipe.pop(0)
        flows += [(1 - D) * old[0], -COST * old[0] * (1 - old[1]), -FIX]
        pipe.append([n, pre])
        if "refund" in active and t == hits.get("refund", -1):
            for e in pipe:
                flows.append(-D * 0.30 * e[0])
                e[0] *= 0.70
        cash = math.fsum([cash] + flows)
        if cash < 0:
            return True
        cap = rho * D * math.fsum(e[0] for e in pipe)
        cash = min(cash, cap)
    return False


def wilson_upper(bad, n, z=1.96):
    p = bad / n
    centre = p + z * z / (2 * n)
    half = z * math.sqrt(p * (1 - p) / n + z * z / (4 * n * n))
    return (centre + half) / (1 + z * z / n)


def count(scen, rho):
    bad = 0
    for noise, hits in scen:
        bad += ruptured(noise, hits, rho)
    return bad


N = 4000
scen = [scenario(i) for i in range(N)]
for label, test in (("point p <= 0.01", lambda b: b / N <= 0.01),
                    ("Wilson upper <= 0.01",
                     lambda b: wilson_upper(b, N) <= 0.01)):
    lo, hi = 0.0, 1.0
    for _ in range(10):
        mid = (lo + hi) / 2
        if test(count(scen, mid)):
            hi = mid
        else:
            lo = mid
    b = count(scen, hi)
    print("%-22s rho* = %.3f  (p = %.4f, upper %.4f)" % (
        label, hi, b / N, wilson_upper(b, N)))
print("deterministic floor from Chapter 5, (c + d - 1) / d = %.3f" % (
    (COST + D - 1) / D))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《负营运资金飞轮、预收定金周转动力学与零本金商业放大母典》第 4 章「三种冲击下的资金链断裂概率与圈存准备金率」(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-365 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026365,
  author    = {Fan, Yafu},
  title     = {Negative Working-Capital Flywheels, Upfront Customer Float and Zero-Capital Bootstrap Scaling},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论