关于「先抵后转:多边净额与最少付款笔数」,《跨境结算摩擦与结汇批量》给出的定量结论是什么?

结论:第四章说明直接多边净额的成本只是逐张付款的 0.363。

数据口径:书中的费率、汇率、波动率、金额与主体结构均为示意;全部计算由书中程序生成;本书是非个人化法律、税务或理财建议,只讨论在合法持牌通道上、如实申报前提下降低摩擦,不写任何规避外汇管制、拆分规避申报、地下钱庄或洗钱的手法,外汇管理规定只述大意,以上架前人工复核为准。 第一章用一张月账比较六种结汇规则:示意参数下逐单结汇的摩擦是平方根批量的 10.93 倍;并修正蓝图:平方根律借自鲍莫尔与托宾的现金存货模型,只在有固定费用时成立。

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: 0.363

net position kUSD: CN +480 HK -40 US -355 DE +40 SG -40 UK -85
settlementpayskUSDcost USDbp
gross3012653137.524.80
bilateral136251537.524.60
hub via HK510001592.515.93
direct net45201140.021.92
direct-net cost / gross cost = 0.363; hub / gross = 0.508
fixed fees: gross 750 USD, direct net 100 USD
查看完整验算程序
# Chapter 4. p41: gross, bilateral and multilateral settlement of one month.
# Six group entities owe each other for goods and services under real
# contracts (matrix in kUSD-equivalent, illustrative). Each entity keeps
# one functional currency. A payment costs a fixed 25 USD, 5 bp for the
# transfer and, when the two currencies differ, a 15 bp conversion.
# Gross: every invoice paid. Bilateral: each pair nets to one payment.
# Hub: a centre in HK (USD) receives or pays each entity's net position
# once. Direct: the same net positions settled payer to receiver,
# largest first, so money never passes through the centre. Netting is
# lawful only where the rules of every jurisdiction involved permit it
# and every underlying invoice is recorded and declared in full; the
# model does not decide that.
import math

ENT = ["CN", "HK", "US", "DE", "SG", "UK"]
CCY = {"CN": "CNY", "HK": "USD", "US": "USD", "DE": "EUR",
       "SG": "SGD", "UK": "GBP"}
PAY = [  # PAY[i][j]: what entity i owes entity j this month
    [0, 40, 25, 30, 10, 15],
    [120, 0, 35, 20, 15, 10],
    [260, 60, 0, 120, 20, 30],
    [60, 25, 40, 0, 10, 35],
    [70, 15, 10, 5, 0, 5],
    [90, 20, 25, 35, 10, 0]]
FIX, WIRE, FX = 25.0, 5, 15


def pay_cost(a, b, amount):
    bp = WIRE + (FX if CCY[a] != CCY[b] else 0)
    return FIX + bp * amount / 10.0


def report(name, payments):
    vol = math.fsum(p[2] for p in payments)
    cost = math.fsum(pay_cost(*p) for p in payments)
    print("%-13s %5d %9.0f %10.1f %8.2f"
          % (name, len(payments), vol, cost, cost / vol * 10.0
             if vol else 0.0))
    return cost


n = len(ENT)
gross = [(ENT[i], ENT[j], PAY[i][j]) for i in range(n) for j in range(n)
         if PAY[i][j]]
bilat = []
for i in range(n):
    for j in range(i + 1, n):
        d = PAY[i][j] - PAY[j][i]
        if d > 0:
            bilat.append((ENT[i], ENT[j], d))
        elif d < 0:
            bilat.append((ENT[j], ENT[i], -d))
net = {e: 0 for e in ENT}
for a, b, x in gross:
    net[a] -= x  # step
    net[b] += x  # step
hub = []
for e in ENT:
    if e != "HK" and net[e] < 0:
        hub.append((e, "HK", -net[e]))
    elif e != "HK" and net[e] > 0:
        hub.append(("HK", e, net[e]))
owe = {e: -v for e, v in net.items() if v < 0}
get = {e: v for e, v in net.items() if v > 0}
direct = []
while owe:
    a = max(sorted(owe), key=lambda e: owe[e])
    b = max(sorted(get), key=lambda e: get[e])
    x = min(owe[a], get[b])
    direct.append((a, b, x))
    for book, e in ((owe, a), (get, b)):
        book[e] -= x  # step
        if book[e] == 0:
            del book[e]
print("net position kUSD: " + "  ".join("%s %+d" % (e, net[e])
                                         for e in ENT))
print("%-13s %5s %9s %10s %8s"
      % ("settlement", "pays", "kUSD", "cost USD", "bp"))
g = report("gross", gross)
b = report("bilateral", bilat)
h = report("hub via HK", hub)
d = report("direct net", direct)
print("direct-net cost / gross cost = %.3f; hub / gross = %.3f"
      % (d / g, h / g))
print("fixed fees: gross %d USD, direct net %d USD"
      % (FIX * len(gross), FIX * len(direct)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《跨国多主体结算摩擦与合规资金池调度、多币种通道费率换汇滑点最小损耗网络流母典》第 4 章「先抵后转:多边净额与最少付款笔数」(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-280 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026280,
  author    = {Fan, Yafu},
  title     = {Cross-Border Settlement Frictions and Compliant Treasury Scheduling: Minimum-Loss Network Flows over Multi-Currency Channel Fees and Conversion Slippage},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论