关于「「持而盈之,不如其已」:东西方原典里的停止与离场」,《期权提前止盈与最优停时》给出的定量结论是什么?

结论:取第一类错误 0.05、第二类错误 0.20,上界常数是 16.000,下界是 0.2105。效应量为 0.10 个标准差时,固定样本检验需要 619 笔配对交易;序贯检验在原假设下的平均样本量约 268 笔,在备择假设下约 381 笔,模拟分别得到 295 笔与 444 笔,实际第一类错误 0.045,检验力 0.790。效应量 0.20 时,固定样本 155 笔,序贯平均 67 笔与 95 笔;0.40 时分别是 39 笔、17 笔与 24 笔,检验力 0.848。序贯检验平均省下三到六成的样本,代价是单次检验的样本量不固定。第八章的预注册方案会采用它。

数据口径:书中价格一律以指数点表示,路径全部合成,参数为示意;全部计算由书中程序生成。 书中以标注框区分已发表的证据、可运行的 Python 计算(21 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: 0.045, 0.10, 0.20, 0.2105, 0.40, 0.790

Wald bounds A 16.000, B 0.2105 (log 2.773, -1.558)
effectfixed nASN H0ASN H1sim n H0sim n H1alphapower
0.106192683812954440.0450.790
0.20155679574990.0480.797
0.4039172419280.0430.848
查看完整验算程序
# Chapter 7. p72: Wald's sequential test, applied to "does the rule beat
# holding?" Paired per-trade differences are taken as normal with
# standard deviation 1 (so effects are in standard-deviation units).
# H0: mean 0; H1: mean e. Error rates alpha 0.05, beta 0.20. Wald's
# bounds A = (1 - beta) / alpha, B = beta / (1 - alpha); his average
# sample numbers under H0 and H1, set against the fixed-sample size of
# a one-sided z-test, and a seeded simulation of 400 runs each.
import math
import random
import statistics

ALPHA, BETA = 0.05, 0.20
z = statistics.NormalDist()
la, lb = math.log((1 - BETA) / ALPHA), math.log(BETA / (1 - ALPHA))
print("Wald bounds A %.3f, B %.4f (log %.3f, %.3f)"
      % ((1 - BETA) / ALPHA, BETA / (1 - ALPHA), la, lb))


def sprt(e, true_mean, rng):
    llr, n = 0.0, 0
    while lb < llr < la:
        x = rng.gauss(true_mean, 1.0)
        llr += e * (x - 0.5 * e)
        n += 1
    return n, llr >= la


rng = random.Random(42672)
fixed_z = z.inv_cdf(1 - ALPHA) + z.inv_cdf(1 - BETA)
print("effect  fixed n  ASN H0  ASN H1  sim n H0  sim n H1  "
      "alpha  power")
for e in (0.10, 0.20, 0.40):
    nfix = math.ceil((fixed_z / e) ** 2)
    asn0 = ((1 - ALPHA) * lb + ALPHA * la) / (-0.5 * e * e)
    asn1 = ((1 - BETA) * la + BETA * lb) / (0.5 * e * e)
    r0 = [sprt(e, 0.0, rng) for _ in range(400)]
    r1 = [sprt(e, e, rng) for _ in range(400)]
    print("%5.2f   %6d   %6.0f  %6.0f   %6.0f    %6.0f   %.3f  %.3f"
          % (e, nfix, asn0, asn1,
             statistics.fmean([n for n, _ in r0]),
             statistics.fmean([n for n, _ in r1]),
             len([1 for _, h in r0 if h]) / 400,
             len([1 for _, h in r1 if h]) / 400))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《期权提前止盈最优停时、恐慌脉冲波动率收割与持有至到期回吐陷阱母典》第 7 章「「持而盈之,不如其已」:东西方原典里的停止与离场」(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-426 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026426,
  author    = {Fan, Yafu},
  title     = {Optimal Early Profit-Taking in Options: Harvesting Volatility Spikes, Twenty-One-Day Decay Cliffs and the Expiration Hold Trap},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论