What does Pruning a Domain Portfolio conclude, in numbers, about “The Price Evidence: Repeat-Sales Indices, Survivorship Bias and the Cost of Holding”?

Answer: Chapter 2 examines the price evidence: in the illustrative model a repeat-sales estimate is biased upwards by 0.1284, and applying the index's growth of 0.0723 a year leads to systematic under-pruning.

Scope of the numbers: Chapter 2 examines the price evidence: in the illustrative model a repeat-sales estimate is biased upwards by 0.1284, and applying the index's growth of 0.0723 a year leads to systematic under-pruning. Marked boxes separate published evidence, runnable Python calculations (20 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: 0.1284

A. true drift vs repeat-sales estimate, sigma 0.40
   true mu   estimate   bias    pairs   names traded
     -0.10     0.0332  0.1332     160    1,202
     -0.05     0.0784  0.1284     261    1,585
      0.00     0.1225  0.1225     440    2,095
      0.05     0.1576  0.1076     773    2,816
B. bias grows with volatility, true mu = -0.05
   sigma 0.20: estimate -0.0078, bias 0.0422
   sigma 0.40: estimate 0.0784, bias 0.1284
   sigma 0.60: estimate 0.1386, bias 0.1886
Show the full program
# Chapter 2. p21: a repeat-sales index sees only names that trade.
# Model: 20,000 synthetic names; log value drifts by mu per year with
# volatility s. A name trades in a year with a chance that rises with
# its value (buyers bid on names that look valuable). The repeat-sales
# estimate is the mean annual log change between two sales of the same
# name; the truth is mu for every name, traded or not. Seeded and
# illustrative: it shows the direction of the selection effect only.
import math
import random


def run(mu, s, years=12, n=20000):
    rng = random.Random(20261120)
    pair_rates, traded = [], 0
    for _ in range(n):
        lv = math.log(300) + rng.gauss(0, 1.0)
        last = None
        sold = False
        for t in range(1, years + 1):
            lv += mu + rng.gauss(0, s)
            chance = min(0.25, 0.004 * math.exp(lv - math.log(300)))
            if rng.random() < chance:
                sold = True
                if last is not None:
                    pair_rates.append((lv - last[1]) / (t - last[0]))
                last = (t, lv)
        traded += sold
    return math.fsum(pair_rates) / len(pair_rates), len(pair_rates), traded


if __name__ == "__main__":
    print("A. true drift vs repeat-sales estimate, sigma 0.40")
    print("   true mu   estimate   bias    pairs   names traded")
    for mu in (-0.10, -0.05, 0.00, 0.05):
        est, pairs, traded = run(mu, 0.40)
        print("   {:>7.2f}   {:>8.4f}  {:>6.4f}  {:>6,}   {:>6,}".format(
            mu, est, est - mu, pairs, traded))
    print("B. bias grows with volatility, true mu = -0.05")
    for s in (0.20, 0.40, 0.60):
        est, pairs, _ = run(-0.05, s)
        print("   sigma {:.2f}: estimate {:.4f}, bias {:.4f}".format(
            s, est, est + 0.05))

Single source and citation

Fan, Y. (2026). Real-Options Valuation of Geo-Exact Domain Portfolios, Optimal Renewal Pruning and the Moats of Organic Search (Version 1.0.0), Chapter 2: The Price Evidence: Repeat-Sales Indices, Survivorship Bias and the Cost of Holding. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-341 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026341,
  author    = {Fan, Yafu},
  title     = {Real-Options Valuation of Geo-Exact Domain Portfolios, Optimal Renewal Pruning and the Moats of Organic Search},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book