5D Library › 慢崩盘流动性储备、实物资产期权估值与跨代财富保全动力学 · English
第二章 熊市反弹、空头的时间损耗与被迫卖出的时间表
看对方向,仍然赚不到钱
一位看空者完全看对了第一章的慢崩盘:指数会跌一半。他从第 0 个月起每季度花资本的 2% 买三个月期的平值看跌期权,按隐含波动率 35% 定价。另一位看空者等指数从高点下跌 20% 后才开始同样的做法。
推导
# Chapter 2. p03: Being right too early. A bear who is right about the
# whole slow crash of p02 spends 2% of capital each quarter on
# three-month at-the-money puts, priced by Black-Scholes at an implied
# volatility of 35% and a 2% rate, from month 0 to month 36. A second
# bear waits until the index is 20% below its peak and only then starts
# the same programme. Cumulative profit as a share of capital, on the
# 2,000 paths of p02 (seed 88).
import importlib.util
import random
from math import erf, exp, log, sqrt
def load(name):
spec = importlib.util.spec_from_file_location(name, name + ".py")
m = importlib.util.module_from_spec(spec)
spec.loader.exec_module(m)
return m
p02 = load("p02_slow_crash")
def put_frac(vol=0.35, r=0.02, t=0.25):
# at-the-money put price as a fraction of spot
n = lambda z: 0.5 * (1 + erf(z / sqrt(2)))
d1 = (r + vol * vol / 2) * t / (vol * sqrt(t))
d2 = d1 - vol * sqrt(t)
return exp(-r * t) * n(-d2) - n(-d1)
def run(p, start_rule):
cost = put_frac()
pnl, marks, started = 0.0, {}, False
peak = p[0]
for q in range(0, 36, 3):
peak = max(peak, max(p[:q + 1]))
if start_rule == "now" or p[q] <= 0.8 * peak:
started = True
if started:
notional = 0.02 / cost
pnl += notional * max(p[q] - p[q + 3], 0) / p[q] - 0.02
marks[q + 3] = pnl
return marks
if __name__ == "__main__":
rng = random.Random(88)
paths = [p02.path(rng) for _ in range(2000)]
print(" put price: {:.3f} of spot".format(put_frac()))
print(" rule month 6 month 12 month 24 month 36"
" loss at 12")
for rule, name in (("now", "start at once"), ("wait", "wait for -20%")):
m = [run(p, rule) for p in paths]
med = lambda k: sorted(x[k] for x in m)[1000]
neg = sum(1 for x in m if x[12] < 0) / 2000
print(" {:<13} {:7.3f} {:8.3f} {:8.3f} {:8.3f} {:10.3f}"
.format(name, med(6), med(12), med(24), med(36), neg))
# put price: 0.067 of spot
# rule month 6 month 12 month 24 month 36 loss at 12
# start at once -0.011 -0.007 0.008 0.002 0.624
# wait for -20% 0.000 0.000 0.005 0.000 0.289每份平值看跌期权的价格约为标的的 0.067。立刻开始的看空者在第 12 个月的累计盈亏中位数是 -0.007,有 0.624 的路径处于亏损;到第 36 个月,在指数已经跌掉一半的情况下,累计盈亏的中位数只有 0.002。原因是期权价格已经包含了很高的波动率,而每次注资带来的反弹都让一整季的期权作废。等到下跌 20% 才开始的看空者亏损的机会小得多(0.289),但赚得同样不多。看对一次慢崩盘,不等于能从中赚钱;时间损耗会把大部分方向上的正确吃掉。
被迫卖出的人什么时候出现
对想买入的一方,更重要的问题是:被迫卖出的人什么时候最多?p04 设定一千位以贷款买入资产的持有人,贷款成数在 0.40 至 0.70 之间,各自在第 1 至第 36 个月之间的某个月到期;到期时,只有贷款不超过资产现值的 0.75,贷方才肯续贷,否则必须卖出。
推导
# Chapter 2. p04: When forced sellers appear. 1,000 owners hold assets
# bought with loans at a loan-to-value L0 between 0.40 and 0.70; each
# loan falls due once, in a month spread evenly over months 1 to 36.
# At maturity the lender rolls the loan over only if the loan is at most
# 0.75 of the asset's value at today's price; otherwise the owner must
# sell. Counted by half-year on the 2,000 paths of p02 (seed 88).
import importlib.util
import random
def load(name):
spec = importlib.util.spec_from_file_location(name, name + ".py")
m = importlib.util.module_from_spec(spec)
spec.loader.exec_module(m)
return m
p02 = load("p02_slow_crash")
if __name__ == "__main__":
rng = random.Random(88)
paths = [p02.path(rng) for _ in range(2000)]
# loan-to-value and maturity spread evenly and independently
loans = [(0.40 + 0.30 * (i + 0.5) / 1000, 1 + (13 * i) % 36)
for i in range(1000)]
buckets = [0.0] * 6
for p in paths:
for l0, m in loans:
if l0 > 0.75 * p[m]:
buckets[(m - 1) // 6] += 1 / 2000
total = sum(buckets)
print(" months forced sales share of all")
for b in range(6):
print(" {:2d} to {:2d} {:12.1f} {:12.3f}".format(
6 * b + 1, 6 * b + 6, buckets[b], buckets[b] / total))
print(" forced sales per 1,000 owners: {:.1f}".format(total))
# months forced sales share of all
# 1 to 6 5.7 0.012
# 7 to 12 24.5 0.051
# 13 to 18 61.9 0.128
# 19 to 24 102.6 0.211
# 25 to 30 138.9 0.286
# 31 to 36 151.7 0.313
# forced sales per 1,000 owners: 485.2在第一章的路径上,前六个月几乎没有被迫卖出(0.012),第 19 至 24 个月占 0.211,第 25 至 30 个月占 0.286,第 31 至 36 个月占 0.313。被迫卖出的高峰出现在下跌的后半段,因为价格要跌得够深,又要恰好碰上贷款到期。每一千位持有人中约有 485.2 位最终被迫卖出。
假说 · 尚未验证
作者的诠释(尚未验证): 在由债务驱动的慢崩盘中,被迫卖出(法拍、清算与违约后的处置)的数量在下跌开始后的第二年下半年到第三年达到高峰,而不是在下跌之初。什么会证明它错: 在至少两次慢崩盘的商业地产或企业资产处置数据中,处置数量的高峰出现在下跌开始后十二个月内。
小结
慢崩盘里的反弹与期权的时间价值会吃掉看空者大部分的正确;被迫卖出则集中在下跌的后半段,也就是价格已经很低、而到期的贷款又无法续借的时候。对持有现金的买方,真正的机会窗口在后面,而不是在第一波暴跌。
Fan, Y. (2026). Slow-Crash Liquidity Reserves, Real-Asset Option Valuation and Intergenerational Wealth Preservation: Thirty-Month Drawdown Kinetics, Distressed Resource Acquisition and Ergodic Dynasty Trusts [慢崩盘流动性储备、实物资产期权估值与跨代财富保全动力学] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090231 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).