关于「沉没成本与投资模型:为什么投入越多越难止损」,《远期承诺期权与对等停时》给出的定量结论是什么?
结论:第六章量化按沉没成本继续的损失,平均每次 2.02,并用投资模型元分析解释锁定。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是心理咨询、婚恋咨询或临床判断,也不做任何命理推导。 书中以标注框区分证据、可运行计算(21 段程序)、原典与作者假说。
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: 2.02
| cases 20,000; outside option 50; sunk cut 60 | |
| decisions that differ | 0.201 |
| mean realised value, forward rule | 54.78 |
| mean realised value, sunk rule | 52.76 |
| loss per case from the sunk rule | 2.02 |
| loss per differing case | 10.04 |
查看完整验算程序
# Chapter 6. p61: continuing by sunk cost versus by forward value
# Synthetic decisions: past investment S ~ U(0, 100) is already spent.
# The forward value of continuing is V ~ N(mu, 15), with mu drawn per
# case; the value of the best outside use of the same time is O = 50.
# Rule F (forward) continues iff E[V] > O. Rule S (sunk) continues iff
# E[V] > O or S > 60 ("too much invested to stop"). Illustrative.
import math
import random
rng = random.Random(42361)
N = 20000
O, CUT = 50.0, 60.0
diff_cases, loss_f, loss_s = 0, [], []
for _ in range(N):
s = rng.uniform(0.0, 100.0)
mu = rng.gauss(50.0, 12.0)
v = rng.gauss(mu, 15.0)
go_f = mu > O
go_s = mu > O or s > CUT
loss_f.append(v if go_f else O)
loss_s.append(v if go_s else O)
if go_f != go_s:
diff_cases += 1
mf = math.fsum(loss_f) / N
ms = math.fsum(loss_s) / N
print("cases {:,}; outside option 50; sunk cut 60".format(N))
print("decisions that differ: {:.3f}".format(diff_cases / N))
print("mean realised value, forward rule: {:.2f}".format(mf))
print("mean realised value, sunk rule: {:.2f}".format(ms))
print("loss per case from the sunk rule: {:.2f}".format(mf - ms))
print("loss per differing case: {:.2f}".format(
(mf - ms) * N / diff_cases))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《亲职基因优选与排他恋爱解耦、远期空头期权钩子与对等底线停时母典》第 6 章「沉没成本与投资模型:为什么投入越多越难止损」(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-423 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026423,
author = {Fan, Yafu},
title = {Unbundling Paternal Selection from Romantic Commitment: Deferred-Promise Options, Avoidant Autonomy and Reciprocal Stopping Rules},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}