5D Library › 多阶段最后通牒议价、应收账款衰减动力学与最优施压序列 · English

第五章 可信的威胁与贯彻到底的价值

起诉的威胁何时可信?

谢林指出,威胁只有在发出威胁的人到时候真的会执行,或者已经安排得非执行不可时才起作用 (Schelling 1960)。债权人起诉的威胁,只有在债务人拒付之后起诉仍然值得时才可信。p09 把这写成一个比较:起诉成本为 CC;债权人以 0.9 的机会胜诉,之后能收回判决金额的 0.7,所以只有当 0.63A>C0.63 A > C 时起诉才划算。

推导

# Chapter 5. p09: When is the threat to sue credible? An unpaid invoice
# of amount A. Suing costs C; the creditor wins with chance 0.9 and can
# then collect 0.7 of the judgment. Suing is worth doing after the
# debtor refuses only if 0.9 x 0.7 x A > C, so a debtor who knows the
# numbers can ignore the threat below A* = C / 0.63. Three regimes:
# ordinary litigation (C = 3,000), a small-claims track (C = 300), and
# a lawyer already paid on retainer, so that only 1,000 more is spent.
if __name__ == "__main__":
    regimes = (("ordinary court", 3000), ("small claims", 300),
               ("retainer paid", 1000))
    for lab, c in regimes:
        print("   {:<15} threshold A* = {:8,.0f}".format(lab, c / 0.63))
    print("   amount     ordinary   small claims   retainer paid")
    for a in (500, 2000, 5000, 20000):
        row = ["credible" if 0.63 * a > c else "empty" for _, c in regimes]
        print("   {:6,d}     {:>8}   {:>12}   {:>13}".format(a, *row))
#    ordinary court  threshold A* =    4,762
#    small claims    threshold A* =      476
#    retainer paid   threshold A* =    1,587
#    amount     ordinary   small claims   retainer paid
#       500        empty       credible           empty
#     2,000        empty       credible        credible
#     5,000     credible       credible        credible
#    20,000     credible       credible        credible

在成本 3,000 的普通诉讼里,威胁只对 4,762 以上的发票可信;欠 2,000、又清楚这些数字的债务人可以不理它。成本 300 的小额诉讼程序把门槛降到 476。律师已经按聘请费付清、只需再花 1,000 时,门槛降到 1,587。让威胁可信,主要是降低执行它的成本,或者事先已经付掉。

注

威胁必须是真的。 在美国,《公平债务催收作业法》禁止消费债务的催收者以「任何依法不能采取或并不打算采取的行动」相威胁 (United States Congress 1977);其他地方也有类似规定。在消费债务催收中,空洞的威胁不是手段,而是违法。本章的算术讨论的是哪些真实的步骤值得采取。

信誉

面对许多债务人的债权人,即使单个诉讼亏钱,贯彻到底也可能有收益,因为其他债务人会知道威胁是真的。p10 比较这类诉讼的成本与它们促成的额外付款。

推导

# Chapter 5. p10: Following through to be believed. A firm has N
# overdue invoices a year of amount 2,000 each. Suing on one costs
# 3,000 and recovers 0.63 x 2,000 on average, a loss of 1,740 per suit.
# If the firm is known to sue whenever an invoice reaches the last
# stage, the share of debtors who pay before that stage rises from 0.60
# to 0.75; it must then sue the 10 a year who still do not pay (fewer
# when N is small: 10% of N). Net yearly gain from a reputation for
# following through, by N.
if __name__ == "__main__":
    loss = 3000 - 0.63 * 2000
    print("   loss per suit: {:,.0f}".format(loss))
    print("   invoices N   extra paid   suits   net gain")
    for n in (20, 50, 100, 300):
        extra = n * 2000 * (0.75 - 0.60)
        suits = min(10, 0.1 * n)
        print("   {:10d}   {:10,.0f}   {:5.0f}   {:8,.0f}".format(
            n, extra, suits, extra - suits * loss))
    print("   break-even N with 10 suits: {:.1f}".format(
        10 * loss / (2000 * 0.15)))
#    loss per suit: 1,740
#    invoices N   extra paid   suits   net gain
#            20        6,000       2      2,520
#            50       15,000       5      6,300
#           100       30,000      10     12,600
#           300       90,000      10     72,600
#    break-even N with 10 suits: 58.0

对一张 2,000 的发票起诉,平均亏 1,740。但如果贯彻到底的信誉,把在最后阶段之前付款的债务人比例从 0.60 提高到 0.75,一家每年有 20 张逾期发票的公司净得 2,520;100 张时 12,600;300 张时 72,600。如果不论规模大小每年都要打 10 场这样的官司,盈亏平衡点是每年 58.0 张发票。贯彻到底是对之后每一次提醒之可信度的投资;它对债务人众多、彼此会互通消息的债权人有利,对一次性的索赔则利益小得多。

假说 · 尚未验证

作者的诠释(尚未验证): 在小公司中,以把欠款追到法院立案而著称的公司,逾期发票在最后阶段之前付款的比例较高。什么会证明它错: 在一项按行业与客户群配对公司的预先登记研究中,最后阶段之前付款的比例与公司的立案记录无关。

小结

当执行威胁在执行的那一刻是划算的,威胁才可信;较低的成本与事先付出的承诺使它可信。对债务人众多的债权人而言,对不划算的案件贯彻到底,可以通过信誉得到回报;而在消费债务催收中,不打算执行的威胁根本不被允许。

参考文献

Schelling, Thomas C. 1960. The Strategy of Conflict. Harvard University Press.
United States Congress. 1977. Fair Debt Collection Practices Act, 15 U.S.C. 1692e (False or Misleading Representations).

Fan, Y. (2026). Multistage Ultimatum Bargaining, Accounts-Receivable Decay Kinetics and Optimal Escalation Sequences: Ackerman Concession Curves, Credible Threat Thresholds and Default Recovery Bounds [多阶段最后通牒议价、应收账款衰减动力学与最优施压序列] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090245 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).