关于「多层级套餐锚点:自选菜单、价格尾数与真实锚点审计」,《下沉城市定价与套餐锚点》给出的定量结论是什么?

结论:第六章按自选原理给三档菜单定价:第四档相对一线的价格比,基础档 0.592、标准档 0.666、尊享档 0.719,并用五条规则审计锚点,拦下虚构原价与诱饵。

数据口径:书中的城市档位、收入、价格水平、成本与需求参数均为示意,不引用任何具体的官方统计数值;全部计算由书中程序生成;本书是非个人化经营建议,书中小工具给出的每一个价格都须由经营者人工放行并先过定价试算器,价格按城市公开标示,锚点只用真实现行价格。 第一章在四档示意城市上量化命题:第四档城市的平均收入是一线的 0.402,余钱在第 95 百分位却是 0.532;它的最优价 508 元是一线的 0.592,按比例打折到 345 元会少赚 0.077,错在定低。

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: 0.592, 0.666, 0.719

tier basic stand prem | share b s p | prem/basic gain
T185712691693| 0.202 0.069 0.153 |1.9750.278
T267510391426| 0.225 0.062 0.104 |2.1120.233
T35749191293| 0.221 0.049 0.070 |2.2530.207
T45088461218| 0.197 0.035 0.046 |2.4000.192
T4 price / T1 price: basic 0.592, standard 0.666, premium 0.719
查看完整验算程序
# Chapter 6. p61: a three-package menu per tier, priced by self-selection.
# A resident's value of the basic package is W = 0.03 * D(x) * tau (p12
# model); the standard and premium packages are worth V = 1.30 and 1.55
# times as much and cost C = 250 and 450 yuan against 90 for basic. Each
# resident takes the package with the largest V * W - price, or nothing.
# Choice is constant between break points of W, so each share is the
# lognormal taste mass between two cut-offs, averaged over 200 income
# quantiles. Prices are found by rounds of golden-section search: all
# three scaled together, then each in ladder order.
# All parameters are illustrative.
import math
from statistics import NormalDist

N01 = NormalDist()
TIERS = ["T1", "T2", "T3", "T4"]
MED = [72000, 50000, 36000, 26000]
SIG = [0.55, 0.58, 0.64, 0.72]
P = [1.00, 0.90, 0.82, 0.76]
E = [0.38, 0.34, 0.31, 0.29]
V = [1.00, 1.30, 1.55]
C = [90.0, 250.0, 450.0]
NAMES = ["basic", "standard", "premium"]
NQ = 200


def disc(x, k):
    w = 0.28 - 0.09 * math.log(x / (P[k] * 72000))
    w = min(0.65, max(0.10, w))
    return x * max(0.0, 1.0 - w - E[k])


def lwgrid(k):
    # log of 0.03 * D at NQ income quantiles (residents with D > 0)
    out = []
    for i in range(NQ):
        d = disc(MED[k] * math.exp(SIG[k] * N01.inv_cdf((i + 0.5) / NQ)),
                 k)
        if d > 0:
            out.append(math.log(0.03 * d))
    return out


def above(t, lw):
    # share of residents whose basic-package value W exceeds t
    if t <= 0:
        return 1.0
    if t == math.inf:
        return 0.0
    lt = math.log(t)
    return math.fsum(1.0 - N01.cdf((lt - x) / 0.45) for x in lw) / NQ


def shares(pr, lw):
    bps = {pr[j] / V[j] for j in range(3)}
    for a in range(3):
        for b in range(a + 1, 3):
            bps.add((pr[b] - pr[a]) / (V[b] - V[a]))
    cuts = [0.0] + sorted(b for b in bps if b > 0) + [math.inf]
    parts = [[], [], []]
    for lo, hi in zip(cuts[:-1], cuts[1:]):
        rep = lo + 1.0 if hi == math.inf else (lo + hi) / 2.0
        u = [V[j] * rep - pr[j] for j in range(3)]
        j = max(range(3), key=lambda m: u[m])
        if u[j] >= 0:
            parts[j].append(above(lo, lw) - above(hi, lw))
    return [math.fsum(x) for x in parts]


def profit(pr, lw):
    s = shares(pr, lw)
    return 1000.0 * math.fsum(s[j] * (pr[j] - C[j]) for j in range(3))


def golden(f, a, b, n=40):
    r = (math.sqrt(5.0) - 1.0) / 2.0
    for _ in range(n):
        x1, x2 = b - r * (b - a), a + r * (b - a)
        if f(x1) < f(x2):
            a = x1
        else:
            b = x2
    return (a + b) / 2.0


def solve(lw, start):
    # each round: scale all three prices together, then one at a time
    pr = [start * v for v in V]
    for _ in range(6):
        base = list(pr)
        lo = max(0.6, (C[0] + 1.0) / base[0], (C[2] + 1.0) / base[2])
        s = golden(lambda v: profit([p * v for p in base], lw), lo, 1.6)
        pr = [p * s for p in base]
        for j in range(3):
            lo = C[j] + 1.0 if j == 0 else max(C[j] + 1.0, pr[j - 1])
            hi = 6000.0 if j == 2 else pr[j + 1]

            def f(v, j=j):
                q = list(pr)
                q[j] = v
                return profit(q, lw)
            pr[j] = golden(f, lo, max(lo, hi))
    return pr


SINGLE = [857, 675, 574, 508]
print("tier  basic  stand  prem  | share b   s     p   | prem/basic  "
      "gain")
MENUS = []
for k, t in enumerate(TIERS):
    lw = lwgrid(k)
    pr = solve(lw, SINGLE[k])
    MENUS.append(pr)
    s = shares(pr, lw)
    one = profit([SINGLE[k], 1e9, 1e9], lw)
    print("%-4s  %4.0f   %4.0f  %4.0f  | %.3f %.3f %.3f |   %.3f     %.3f"
          % (t, pr[0], pr[1], pr[2], s[0], s[1], s[2], pr[2] / pr[0],
             profit(pr, lw) / one - 1.0))
print("T4 price / T1 price: " + ", ".join(
    "%s %.3f" % (NAMES[j], MENUS[3][j] / MENUS[0][j]) for j in range(3)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《跨省购买力平价与下沉城市定价、收入分布恩格尔系数校准与多层级城市套餐锚点母典》第 6 章「多层级套餐锚点:自选菜单、价格尾数与真实锚点审计」(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-275 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026275,
  author    = {Fan, Yafu},
  title     = {Purchasing-Power Parity across Provinces and Pricing for Lower-Tier Cities: Income Distributions, Engel-Coefficient Calibration and Multi-Tier Package Anchors},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论