关于「报价死区:进价越高,报价反而越低」,《买手定价、费率阶跃与配件加购》给出的定量结论是什么?
结论:第一章证明:低于阈值刷卡抽成、达到阈值改收汇款时,等净利率倒推必然在一段成本区间内倒挂。阈值 13,000、抽成 0.028 时,成本 9,686 的包报 12,808,低于成本 9,650 的包(12,953),利润率只有 0.1884;保住利润率的价格下限是在死区内一律报 13,000 的单调包络线,有 0.0279 的「包·日」落在那里。
数据口径:书中的客户、价格、抽成、运费、退税率与汇率波动率均为示意模型;全部计算由书中程序生成;本书不是财务、税务、海关、法律或投资建议,不描述任何特定金流服务的条款。 第八章把这些合成一年的经营与一份检验方案。
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。 与结论吻合的打印值: 0.028, 0.1884, 12808, 12953, 13000, 9650
| T | 13000, m = 0.20, S1 = 350, S2 = 560 |
| fee g c1 (enter) c2 (leave) width drop at c1 | |
| 0.022 | 9764.0 9840.0 76.0 95.0 |
| 0.025 | 9725.0 9840.0 115.0 143.8 |
| 0.028 | 9686.0 9840.0 154.0 192.5 |
| 0.030 | 9660.0 9840.0 180.0 225.0 |
| fee g | 0.028: quotes across the deadzone |
| cost naive route true margin envelope margin | |
| 9600 | 12889 card 0.2000 12889 0.2000 |
| 9650 | 12953 card 0.2000 12953 0.2000 |
| 9686 | 12808 bank 0.1884 13000 0.2118 |
| 9700 | 12825 bank 0.1884 13000 0.2108 |
| 9750 | 12888 bank 0.1883 13000 0.2069 |
| 9800 | 12950 bank 0.1882 13000 0.2031 |
| 9840 | 13000 bank 0.2000 13000 0.2000 |
| 9900 | 13075 bank 0.2000 13075 0.2000 |
| grid 9000 to 10500, step 0.5 (3001 costs): | |
| naive quote below that of a cheaper item | 307 |
| naive bank quote below T (so paid by card) | 308 |
| envelope nondecreasing: True; margin >= m: True |
查看完整验算程序
# Chapter 1. p01: The quote-inversion deadzone of a stepped payment fee.
# Below the threshold T a customer pays by card and the shop loses the
# fee g; at T or above the shop asks for a bank transfer, saves the fee,
# but pays S2 = S1 + dS + h (insured home delivery plus manual work).
# The naive rule targets the same net margin m on both routes:
# card P1(c) = (c + S1) / (1 - g - m) while P1 < T,
# bank P2(c) = (c + S2) / (1 - m) once P1 >= T.
# We print the deadzone [c1, c2], the drop of the quote at c1, and a
# table across it: the naive quote, the margin it really earns (a quote
# below T is paid by card), and the monotone envelope. No sampling.
T, M, S1, DS, H = 13000.0, 0.20, 350.0, 150.0, 60.0
S2 = S1 + DS + H
def naive(c, g):
p1 = (c + S1) / (1 - g - M)
if p1 < T:
return p1, "card"
return (c + S2) / (1 - M), "bank"
def margin(p, c, g):
if p < T: # bank transfer is not offered below T
return (p * (1 - g) - c - S1) / p
return (p - c - S2) / p
def envelope(c, g):
c1 = T * (1 - g - M) - S1
c2 = T * (1 - M) - S2
if c < c1:
return (c + S1) / (1 - g - M)
if c <= c2:
return T
return (c + S2) / (1 - M)
if __name__ == "__main__":
print(" T = {:.0f}, m = {:.2f}, S1 = {:.0f}, S2 = {:.0f}".format(
T, M, S1, S2))
print(" fee g c1 (enter) c2 (leave) width drop at c1")
for g in (0.022, 0.025, 0.028, 0.030):
c1 = T * (1 - g - M) - S1
c2 = T * (1 - M) - S2
drop = (g * T - (S2 - S1)) / (1 - M)
print(" {:.3f} {:10.1f} {:10.1f} {:5.1f} {:10.1f}".format(
g, c1, c2, c2 - c1, drop))
g = 0.028
print(" fee g = 0.028: quotes across the deadzone")
print(" cost naive route true margin envelope margin")
for c in (9600, 9650, 9686, 9700, 9750, 9800, 9840, 9900):
p, r = naive(c, g)
e = envelope(c, g)
print(" {:4d} {:5.0f} {:5s} {:11.4f} {:8.0f} {:6.4f}"
.format(c, p, r, margin(p, c, g), e, margin(e, c, g)))
grid = [9000 + 0.5 * i for i in range(3001)]
top, under = 0.0, 0
for c in grid:
p = naive(c, g)[0]
under += p < top
top = max(top, p)
low = sum(1 for c in grid if naive(c, g)[0] < T and
naive(c, g)[1] == "bank")
mono = all(envelope(b, g) >= envelope(a, g)
for a, b in zip(grid, grid[1:]))
ok = all(margin(envelope(c, g), c, g) >= M - 1e-12 for c in grid)
print(" grid 9000 to 10500, step 0.5 ({} costs):".format(len(grid)))
print(" naive quote below that of a cheaper item: {}".format(under))
print(" naive bank quote below T (so paid by card): {}".format(low))
print(" envelope nondecreasing: {}; margin >= m: {}".format(mono, ok))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《跨境买手动态定价、阶跃费率支付路由与配件连带加购经济学》第 1 章「报价死区:进价越高,报价反而越低」(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-110 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026110,
author = {Fan, Yafu},
title = {Cross-Border Buyer Dynamic Pricing, Fee-Threshold Payment Routing and Accessory Attach Economics},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}