What does Flywheel Ignition and Replication Losses conclude, in numbers, about “Delay and Oscillation: a Gain above One Is Not Stability”?

Answer: Chapter 4 uses the Jury test to show that with data four weeks old the weekly correction can be at most 0.3473.

Scope of the numbers: Its parameters are illustrative and all calculations come from its programs. It is not individual business advice, promises no revenue, growth or payback, and discusses only single-tier referral rewards paid to the direct referrer.

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: 0.3473

A. largest stable correction gain k by reporting delay d
d k_max (Jury)2 sin(pi/(2(2d+1)))
02.0000 2.0000
11.0000 1.0000
20.6180 0.6180
30.4450 0.4450
40.3473 0.3473
50.2846 0.2846
60.2411 0.2411
B. dominant root of z^(d+1) - z^d + k: decay or growth, and swing
d k |z| period(weeks) verdict
20.2 0.7249 20.3 damped
20.5 0.9406 10.7 damped
20.8 1.0780 9.4 growing swings
40.2 0.9169 22.2 damped
40.5 1.0601 16.4 growing swings
40.8 1.1445 15.1 growing swings
Show the full program
# Chapter 4. p41: delay shrinks the gain a correction loop can bear
# A: an owner corrects spend from d-week-old data, x(t+1) = x(t) -
# k x(t-d); characteristic polynomial z^(d+1) - z^d + k. The Jury
# table (Schur-Cohn reduction) says if all roots lie inside the unit
# circle; bisection finds the largest stable k, compared with the
# closed form 2 sin(pi / (2 (2d + 1))).
# B: the dominant root (Durand-Kerner, cmath): its modulus says damped
# or growing, its angle gives the period of the swing.
import cmath
import math


def jury_stable(c):
    c = [float(x) for x in c]
    while len(c) > 1:
        kappa = c[-1] / c[0]
        if abs(kappa) >= 1.0:
            return False
        n = len(c) - 1
        c = [c[i] - kappa * c[n - i] for i in range(n)]
    return True


def poly_of(d, k):
    if d == 0:
        return [1.0, k - 1.0]
    return [1.0, -1.0] + [0.0] * (d - 1) + [k]


def k_max(d):
    lo, hi = 0.0, 2.5
    for _ in range(60):
        mid = (lo + hi) / 2
        poly = poly_of(d, mid)
        if jury_stable(poly):
            lo = mid
        else:
            hi = mid
    return lo


def roots(c):
    n = len(c) - 1
    c = [x / c[0] for x in c]
    z = [complex(0.4, 0.9) ** i for i in range(n)]
    for _ in range(2000):
        new = []
        for i, zi in enumerate(z):
            num = complex(0.0)
            for co in c:
                num = num * zi + co
            den = complex(1.0)
            for j, zj in enumerate(z):
                if j != i:
                    den *= zi - zj
            new.append(zi - num / den)
        z = new
    return z


print("A. largest stable correction gain k by reporting delay d")
print("  d   k_max (Jury)   2 sin(pi/(2(2d+1)))")
for d in range(0, 7):
    print("  {}   {:.4f}         {:.4f}".format(
        d, k_max(d), 2 * math.sin(math.pi / (2 * (2 * d + 1)))))
print()
print("B. dominant root of z^(d+1) - z^d + k: decay or growth, and swing")
print("  d   k     |z|      period(weeks)   verdict")
for d in (2, 4):
    for k in (0.2, 0.5, 0.8):
        zs = roots(poly_of(d, k))
        top = max(zs, key=lambda z: (round(abs(z), 9), z.imag))
        ang = abs(math.degrees(cmath.phase(top)))
        per = "none" if ang < 1e-6 else "{:.1f}".format(360.0 / ang)
        verdict = "damped" if abs(top) < 1.0 else "growing swings"
        print("  {}   {:.1f}   {:.4f}   {:>8}        {}".format(
            d, k, abs(top), per, verdict))

Single source and citation

Fan, Y. (2026). Critical Ignition Conditions for Business Flywheels, Loop-Gain Friction Accounting and Cross-Vertical Fractal Replication (Version 1.0.0), Chapter 4: Delay and Oscillation: a Gain above One Is Not Stability. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-413 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026413,
  author    = {Fan, Yafu},
  title     = {Critical Ignition Conditions for Business Flywheels, Loop-Gain Friction Accounting and Cross-Vertical Fractal Replication},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book