What does Critical Fractiles for Pre-Season Buying conclude, in numbers, about “Skewed Demand, Salvage and Goodwill: Three Forces on the Fractile, and Why Experience Fails to Correct the Bias”?

Answer: Chapter 2 shows how skewed demand, salvage value and lost goodwill move the fractile, and that the mean order looks better after the fact in 55.90 % of seasons, so that 51.7 seasons are needed to see the difference in profit.

Scope of the numbers: Its goods, stores and demand parameters are illustrative; all calculations are generated by its programs; the book is not individual business advice. Chapter 1 derives the critical fractile from marginal arithmetic; under illustrative parameters the optimal order equals the mean only at a margin of exactly 0.3750, and ordering the mean for an eight-tenths-margin item loses 6.96 % of expected profit.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 51.7

itemCFq*mean diffsd diffmean winsseasons
luxury 0.8333258.0817.92099.90.559051.7
basic0.4286189.221.2367.10.53842360.7
check: fsum of luxury diffs / N = 817.9
Show the full program
# Chapter 2. p22: Why seasonal feedback barely teaches the fractile.
# Paired simulation: the same demand path is served by the optimal
# order q* and by the mean order. Demand ~ Normal(200, 60) truncated
# at 0; luxury item cu 75, co 15; basic item cu 30, co 40.
# We report the mean and s.d. of the per-season profit difference,
# the share of seasons in which the mean order did better ex post, and
# the number of seasons a buyer would need to see the gap at 5 % /
# 80 % power. Fixed seed; parameters illustrative.
import math
import random
from statistics import NormalDist, fmean, stdev

rng = random.Random(20261112)
Z = NormalDist()
MU, SD, N = 200.0, 60.0, 20000
DEMAND = [max(0.0, rng.gauss(MU, SD)) for _ in range(N)]


def margin(q, d, cu, co):
    sold = min(q, d)
    return cu * sold - co * (q - sold)


print("item   CF     q*     mean diff  sd diff  mean wins  seasons")
for name, cu, co in (("luxury", 75.0, 15.0), ("basic", 30.0, 40.0)):
    cf = cu / (cu + co)
    qs = MU + SD * Z.inv_cdf(cf)
    diff = [margin(qs, d, cu, co) - margin(MU, d, cu, co) for d in DEMAND]
    md, sd = fmean(diff), stdev(diff)
    wins = len([x for x in diff if x < 0.0]) / N
    k = (Z.inv_cdf(0.975) + Z.inv_cdf(0.80)) ** 2
    seasons = k * (sd / md) ** 2
    print("{:6} {:.4f} {:6.1f} {:9.1f}  {:7.1f}  {:.4f}     {:5.1f}"
          .format(name, cf, qs, md, sd, wins, seasons))
print("check: fsum of luxury diffs / N = {:.1f}".format(
    math.fsum(margin(MU + SD * Z.inv_cdf(75.0 / 90.0), d, 75.0, 15.0)
              - margin(MU, d, 75.0, 15.0) for d in DEMAND) / N))

Single source and citation

Fan, Y. (2026). Newsvendor Critical Fractiles for Pre-Season Buying, Dirichlet Size-Curve Priors and Dynamic-Programming End-of-Season Markdowns (Version 1.0.0), Chapter 2: Skewed Demand, Salvage and Goodwill: Three Forces on the Fractile, and Why Experience Fails to Correct the Bias. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-211 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026211,
  author    = {Fan, Yafu},
  title     = {Newsvendor Critical Fractiles for Pre-Season Buying, Dirichlet Size-Curve Priors and Dynamic-Programming End-of-Season Markdowns},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book