What does Buyer Pricing, Fee Steps and Accessory Attach conclude, in numbers, about “The Refund That Never Quite Arrives”?

Answer: Chapter 6 shows a promised refund paying 0.9550 of its face by card and 0.8775 in cash, and a joint buffer of 0.0276 pushing a bag into the deadzone.

Scope of the numbers: Marked boxes separate published evidence, runnable Python calculations (16 programs), classical sources and the author's hypotheses.

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: 0.8775, 0.9550

nominal refund 3360 (0.12 of the shelf price)
channelmeanmean/nominal5% worstshortfall 95
card32090.95503160200
cash29480.87753038322
expected loss to failed claims: 101
Show the full program
# Chapter 6. p11: What a tax refund is really worth. A bag's shelf price
# abroad is 28000 TWD at today's rate, tax included; the refund firm
# promises 0.12 of it (an illustrative net rate). Card refund: paid in
# the foreign currency after 30 to 50 trading days, converted by the
# card issuer with a 0.015 fee at the rate of that day (daily 0.0045).
# Cash at the airport: paid at once, less a fixed fee of 120 TWD and a
# booth spread of 0.06. Either way the claim fails with chance 0.03 (no
# customs stamp, a queue missed). 50000 trips. Seed 110.
import math
import random

PRICE, RATE, FAIL = 28000.0, 0.12, 0.03
S = 0.0045

def card(rng):
    if rng.random() < FAIL:
        return 0.0
    days = rng.randint(30, 50)
    move = math.expm1(rng.gauss(0.0, S * math.sqrt(days)))
    return PRICE * RATE * (1 + move) * (1 - 0.015)

def cash(rng):
    if rng.random() < FAIL:
        return 0.0
    return PRICE * RATE * (1 - 0.06) - 120.0

if __name__ == "__main__":
    rng = random.Random(110)
    nominal = PRICE * RATE
    print("   nominal refund {:.0f} ({:.2f} of the shelf price)".format(
        nominal, RATE))
    print("     channel   mean     mean/nominal   5% worst   shortfall 95")
    for name, f in (("card", card), ("cash", cash)):
        xs = sorted(f(rng) for _ in range(50000))
        mean = math.fsum(xs) / len(xs)
        ok = [x for x in xs if x > 0]
        w5 = ok[int(0.05 * len(ok))]
        print("     {:7s}  {:6.0f}   {:12.4f}   {:8.0f}   {:12.0f}".format(
            name, mean, mean / nominal, w5, nominal - w5))
    print("   expected loss to failed claims: {:.0f}".format(
        FAIL * nominal))

Single source and citation

Fan, Y. (2026). Cross-Border Buyer Dynamic Pricing, Fee-Threshold Payment Routing and Accessory Attach Economics (Version 1.0.0), Chapter 6: The Refund That Never Quite Arrives. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 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}
}

Other chapter answers in this book