关于「双曲贴现与偏好反转:1/(1+kD) 的交叉点」,《截止期与时间动机》给出的定量结论是什么?
结论:第二章推导偏好反转点,并说明在一个急躁程度各异的人群里,今天选择较小较早奖赏的人中有 0.4676 会在提前一个月计划时作出相反选择。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是个人化的经营、法律或理财建议。 书中以标注框区分已发表的证据、可运行的 Python 计算(18 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: 0.4676
| cap on the reversal delay: 60.0 days | |
| front delay | share taking 100 sooner |
| 0 | 0.5723 |
| 7 | 0.5232 |
| 14 | 0.4668 |
| 30 | 0.3047 |
| 45 | 0.1143 |
| 55 | 0.0107 |
| 60 | 0.0000 |
| share who take 100 today but 150 when both are 30 days off: 0.2676 | |
| as a share of those who take 100 today: 0.4676 |
查看完整验算程序
# Chapter 2. p23: A population of hyperbolic discounters. Each person has
# a rate k drawn from a lognormal (median 0.02 per day, log-sd 1.0,
# illustrative) and chooses between 100 after a front delay s and 150
# thirty days later. A person takes the smaller reward iff
# s < G A1 / (A2 - A1) - 1 / k, so no one does once s reaches 60 days,
# whatever k is. Exact shares from the normal integral; no simulation.
import math
A1, A2, G = 100.0, 150.0, 30.0
MED, SD = 0.02, 1.0
CAP = G * A1 / (A2 - A1)
def share_ss(s):
# P(k > 1 / (CAP - s)) for a lognormal k
if s >= CAP:
return 0.0
kcut = 1.0 / (CAP - s)
z = (math.log(kcut) - math.log(MED)) / SD
return 0.5 * math.erfc(z / math.sqrt(2.0))
if __name__ == "__main__":
print("cap on the reversal delay: %.1f days" % CAP)
print("front delay share taking 100 sooner")
for s in (0, 7, 14, 30, 45, 55, 60):
print("%8d %.4f" % (s, share_ss(s)))
flips = share_ss(0) - share_ss(30)
print("share who take 100 today but 150 when both are 30 days off:"
" %.4f" % flips)
print("as a share of those who take 100 today: %.4f"
% (flips / share_ss(0)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《双曲贴现偏好反转、截止期承诺装置与时间动机方程母典》第 2 章「双曲贴现与偏好反转:1/(1+kD) 的交叉点」(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-199 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026199,
author = {Fan, Yafu},
title = {Hyperbolic-Discounting Preference Reversals, Deadline Commitment Devices and the Temporal Motivation Equation},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}