In one sentence — From a high-low table labeled “Z Corp.”, Hurst formulates a buy forecast; the security is later revealed as Gruen Industries, and the book reports a highly favorable trade.
Why the case matters
Gruen is the narrative thread of Chapter 4. Hurst says he received weekly data for an unidentified security, without news or fundamentals, and analyzed it as Z Corp.. This setup reduces one form of hindsight: the identity of the security does not explicitly guide the described analysis.
It is not a blind experiment in the modern sense. The book gives no preregistered protocol, universe-selection rule, excluded cases, or independent replication. Its documentary value is that it shows the full procedure and the author’s claims in sequence.
The stated decomposition
On the narrowest envelope Hurst counts seven valid intervals averaging about 6.0 weeks, discarding an anomalous three-week interval. A wider envelope suggests a component of about 19 weeks, which he judges to be roughly 16 weeks old. He also reads the sum of higher components as already pointing upward.
From this structure he forecasts a convergence of lows about three weeks later around 7¼–7½, followed by a move toward roughly 11¼ by October. These are case values dependent on envelopes drawn from the chart available.
Confirmation and reported outcome
The book describes a valid-downtrend-line break in the third week and a purchase around 7¼. Later updates show price moving first outside the short envelope and then the wider one, which Hurst interprets as raising the projection.
The reported close produces about +62% net in eight weeks. Hurst also converts the result into an exceptionally high compounded annualized rate. Cyclepedia retains the trade change as a reported historical datum but does not use the annualization as a performance measure: extending an eight-week move over a year assumes unproven repeatability.
What is actually verified?
Observed price lies within the broad direction and interval of the described forecast. This supports the internal coherence of the account: envelopes, status table, turn zone, and VTL are applied in a recognizable sequence.
Critical questions remain open. We do not know how many similar analyses went unpublished, how results would differ with another analyst, whether all data were available without revision, or how the method performed after every cost and every signal were included.
How it should be replicated today
A replication would predefine market, period, envelope and VTL algorithm, outlier tolerance, entry and exit rules, costs, and benchmark. It would then apply the same procedure to a specified universe and a period not used to choose parameters.
Harvey, Liu, and Zhu show that financial research involving many hypotheses requires adjustments for multiple testing. Their work is not about Gruen or Hurst; it helps place one surprising result against the evidence required for a general rule.
Evidence boundary — The case documents a forecast and a favorable outcome as presented by Hurst. It supplies no success frequency, loss distribution, independent comparison, or proof that the process is replicable.
Sources
- J. M. Hurst, The Profit Magic of Stock Transaction Timing, Prentice-Hall, 1970, Ch. 4, pp. 68–85, Figs. IV-1–IV-10 — Z Corp. data, envelopes, forecast, signals, and Gruen Industries reveal.
- Campbell R. Harvey, Yan Liu, and Heqing Zhu, … and the Cross-Section of Expected Returns, NBER Working Paper 20592, 2014 — multiple testing and evidentiary standards for financial hypotheses.
Related entries
- Envelope construction and cyclic state — decomposition
- Graphic buy timing — procedure
- Edge-band and mid-band — the case’s two entries