5D Library › Cross-Border Payment Routing, Chargeback Defense Games and Treasury Ruin Bounds · 中文

Chapter 3. The Defence Game against Friendly Fraud, and Scoring the Evidence

What a free-rider reckons

“Friendly fraud” is when a buyer who received the goods tells the bank the payment was “not authorised” or the goods “not received”, to get the money back. It is a game: the buyer decides whether to dispute, the merchant whether to fight (submit evidence, known as representment), and the bank decides on the evidence. Before disputing, the buyer estimates (mostly by instinct) the probability ww that the merchant would win. It is a kind of inspection game (Avenhaus et al. 2002).

In the illustrative model the product is a RM 150 digital report. If the merchant fights and wins, the buyer must pay the money back and loses the account (worth RM 50 to him). A buyer’s hassle cost of disputing is uniform between RM 0 and RM 120; he disputes when 150(1−w)−50w150(1-w) – 50w exceeds it. The merchant fights only when 150w150w exceeds the cost of preparing a case: RM 60 by hand, RM 15 with an automatic evidence package. Of every 10,000 orders, 3% come from buyers tempted to try.

Calculation

# Chapter 3. p05: Friendly fraud as a game. A buyer who received the
# product (a RM 150 digital report) can falsely claim "not authorised" or
# "not received". If the merchant fights the dispute with evidence and
# wins (probability w), the buyer pays the money back and loses access to
# the account, a cost to him of RM 50. Buyers differ in the hassle cost h
# of disputing, uniform on RM 0 to 120. A buyer disputes when
#   150 * (1 - w) - 50 * w > h.
# The merchant fights when w * 150 exceeds the cost of preparing the
# case: RM 60 by hand, RM 15 with an automatic evidence package.
# 3% of buyers are tempted; per 10,000 orders.
A, BAN, HMAX, TEMPTED = 150.0, 50.0, 120.0, 0.03

def dispute_share(w):
    gain = A * (1 - w) - BAN * w
    return min(1.0, max(0.0, gain / HMAX))

def loss(w, prep):
    q = dispute_share(w)
    n = 10000 * TEMPTED * q
    if w * A > prep:
        return q, n, n * (prep + (1 - w) * A)
    return q, n, n * A

if __name__ == "__main__":
    print("   win prob  share who dispute  false disputes  loss RM"
          " (by hand / auto)")
    for w in (0.2, 0.35, 0.5, 0.6, 0.75):
        q, n, l1 = loss(w, 60.0)
        _, _, l2 = loss(w, 15.0)
        print("   {:8.2f}  {:17.3f}  {:14.0f}  {:7,.0f} / {:,.0f}".format(
            w, q, n, l1, l2))
    w0 = A / (A + BAN)
    print("   no tempted buyer disputes once w >= {:.2f}".format(w0))
#    win prob  share who dispute  false disputes  loss RM (by hand / auto)
#        0.20              0.917             275   41,250 / 37,125
#        0.35              0.667             200   30,000 / 22,500
#        0.50              0.417             125   16,875 / 11,250
#        0.60              0.250              75    9,000 / 5,625
#        0.75              0.000               0        0 / 0
#    no tempted buyer disputes once w >= 0.75

With a win probability of only 0.20, 0.917 of tempted buyers dispute: 275 false disputes per 10,000 orders; preparing cases by hand, the merchant does not bother to fight and loses RM 41,250; an automatic package makes fighting worthwhile and cuts the loss to RM 37,125. At a win probability of 0.50 the disputers fall to 0.417, with losses of RM 16,875 or RM 11,250; at 0.60 to 0.250. Once the win probability reaches 0.75, no tempted buyer disputes at all: the evidence deters before any dispute is filed.

Share of tempted buyers who dispute, against the merchant’s probability of winning (p05).

How an evidence package raises the chance of winning

Calculation

# Chapter 3. p06: Scoring an evidence package. Illustrative logistic
# model of winning a "fraud" dispute (Visa reason code 10.4): log-odds of
# a win = -1.2 plus the weight of each element present:
#   delivery/download log with timestamps        +0.6
#   terms-accepted tick, with version and time   +0.3
#   IP or device matches the account's history   +0.7
#   messages from the buyer after delivery       +0.5
# Visa Compelling Evidence 3.0 (from April 2023): two earlier undisputed
# card purchases 120 to 365 days old whose data match (two of user ID,
# IP, delivery address, device ID; one of them IP or device) shift the
# liability for a 10.4 dispute; modelled here as a win probability 0.95.
from math import exp

W = {"log": 0.6, "terms": 0.3, "ip/device": 0.7, "messages": 0.5}

def win(items, ce3=False):
    if ce3:
        return 0.95
    z = -1.2 + sum(W[i] for i in items)
    return 1 / (1 + exp(-z))

if __name__ == "__main__":
    pk = ((), ("terms",), ("log",), ("log", "terms"),
          ("log", "terms", "ip/device"),
          ("log", "terms", "ip/device", "messages"))
    for items in pk:
        print("   {:<34} win {:.3f}".format(" + ".join(items) or "nothing",
                                            win(items)))
    print("   {:<34} win {:.3f}".format("CE 3.0 match (10.4 only)",
                                        win((), True)))
    full = ("log", "terms", "ip/device", "messages")
    for i in full:
        rest = tuple(x for x in full if x != i)
        print("   dropping {:<10} costs {:.3f} of win probability".format(
            i, win(full) - win(rest)))
#    nothing                            win 0.231
#    terms                              win 0.289
#    log                                win 0.354
#    log + terms                        win 0.426
#    log + terms + ip/device            win 0.599
#    log + terms + ip/device + messages win 0.711
#    CE 3.0 match (10.4 only)           win 0.950
#    dropping log        costs 0.137 of win probability
#    dropping terms      costs 0.065 of win probability
#    dropping ip/device  costs 0.161 of win probability
#    dropping messages   costs 0.112 of win probability

Below, an illustrative logistic model scores evidence packages for “fraud” disputes (Visa reason code 10.4). Submitting nothing, the win probability is 0.231; only the record of the ticked terms, 0.289; a time-stamped delivery or download log, 0.354; log and terms, 0.426; adding an IP or device matching the account’s history, 0.599; adding messages the buyer sent after delivery, 0.711. Of the four, dropping the IP or device loses most (0.161) and dropping the terms tick least (0.065).

Visa’s Compelling Evidence 3.0 (from April 2023) adds a rule for reason code 10.4 disputes: if the same card has two undisputed transactions 120 to 365 days before the dispute whose data match the disputed one (at least two of user ID, IP, delivery address and device ID, one of which must be IP or device), liability shifts to the issuer (Visa 2023). The model records this case as a win probability of 0.950.

Note

A “considered purchase” tick box carries little weight. In the illustrative model the record of ticked terms is the weakest of the four kinds of evidence; banks weigh objective records of delivery and use, and identity data that match the account’s history, far more. The tick record is still worth keeping, but it cannot replace the logs.

Summary

Friendly fraud is a game of deterrence: the higher the merchant’s chance of winning, the fewer tempted buyers dispute, and at a win probability of 0.75 false disputes vanish in the illustrative model. The weightiest parts of an evidence package are the delivery log and a matching IP or device; matching transactions that meet Compelling Evidence 3.0 shift the liability for fraud disputes.

References

Avenhaus, Rudolf, Bernhard von Stengel, and Shmuel Zamir. 2002. “Inspection Games.” Chap. 51 in Handbook of Game Theory with Economic Applications, edited by Robert J. Aumann and Sergiu Hart, vol. 3. Elsevier.
Visa. 2023. Visa Core Rules: Compelling Evidence 3.0 for Dispute Condition 10.4 (Other Fraud, Card-Absent Environment). Visa Inc.

Fan, Y. (2026). Cross-Border Payment Routing, Chargeback Defense Games and Treasury Ruin Bounds: Multi-Acquirer Success Optimization, Dispute Evidence Scoring and Cramér-Lundberg Cash Reserves [跨境支付智能路由、拒付抗辩博弈与资金池破产边界] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090225 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).