关于「赝品先验:鉴定证据的似然比与信息价值」,《收藏品回报去偏》给出的定量结论是什么?

结论:赝品是收藏品回报里看不见的一项扣减:它不出现在指数里,因为被揭穿的赝品直接从数据中消失。用贝叶斯定理看,同一条证据在不同先验下分量完全不同;在先验极低的品类里,即使四条证据全部通过,后验仍可能只有六成多。残余的赝品概率可以折算成每年的回报拖累,在 0.20 时与选择偏差的修正幅度相当。检测的价值呈倒 U 形,只在决定悬而未决时最高。下一章转向另一项确定发生的扣减:佣金。

数据口径:书中参数为示意;全部计算由书中程序生成;本书非个人化法律或理财建议,费率与法规以上架前人工复核为准。 书中以标注框区分已发表的证据、可运行的 Python 计算(20 段程序)、原典与作者的假说。

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: 0.20

genuine growth 0.063 a year, fake keeps 0.05
F5 yrs10 yrs20 yrsdrag@10
0.000.06300.06300.06300.0000
0.010.06090.06200.06250.0010
0.030.05680.05990.06140.0031
0.050.05260.05770.06030.0053
0.100.04170.05220.05750.0108
0.200.01850.04020.05140.0228
F that wipes out ten years of growth: 0.470
查看完整验算程序
# Chapter 4. p42: a residual forgery probability as a yearly return drag
# Model: a piece bought at the price of a genuine work grows at G a year
# if genuine; if it is a fake (probability F after all checks) it is
# worth K of a genuine one when the truth comes out. The expected multiple
# after N years is (1 - F)*(1 + G)**N + F*K; the table turns it into an
# annual rate. G = 0.063 borrows the corrected Korteweg et al. figure;
# K = 0.05 and the grid of F are illustrative.
import math

G, K = 0.063, 0.05
FS = [0.00, 0.01, 0.03, 0.05, 0.10, 0.20]
NS = [5, 10, 20]


def annual(f, n):
    mult = math.fsum([(1.0 - f) * (1.0 + G) ** n, f * K])
    return mult ** (1.0 / n) - 1.0


def main():
    print("genuine growth {:.3f} a year, fake keeps {:.2f}".format(G, K))
    print("{:>6}".format("F") + "".join("{:>10}".format("{} yrs".format(n))
                                        for n in NS) + "{:>10}".format(
                                            "drag@10"))
    for f in FS:
        row = [annual(f, n) for n in NS]
        print("{:>6.2f}".format(f) + "".join("{:>10.4f}".format(x)
                                             for x in row) +
              "{:>10.4f}".format(G - row[1]))
    lo, hi = 0.0, 1.0
    for _ in range(60):
        mid = 0.5 * (lo + hi)
        if annual(mid, 10) > 0.0:
            lo = mid
        else:
            hi = mid
    print("F that wipes out ten years of growth: {:.3f}".format(lo))


if __name__ == "__main__":
    main()

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《收藏品重复成交指数选择偏差、赝品先验与另类资产流动性折价母典》第 4 章「赝品先验:鉴定证据的似然比与信息价值」(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-208 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026208,
  author    = {Fan, Yafu},
  title     = {Repeat-Sales Selection Bias in Collectibles, Forgery Priors and the Illiquidity Discount of Alternative Assets},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论