In one sentence — Nesting treats an existing envelope as the new object of study: it wraps that band to look for a longer cycle, or studies the swings inside it to look for a shorter one.
What does it mean to nest an envelope?
Imagine a small wave travelling inside a larger wave. The first curvilinear envelope describes the smaller movement; a second band may reveal the slower movement that contains it. Hurst called this extension nesting.
The procedure has two directions. Nesting up moves toward a longer component. Nesting down moves toward shorter swings. It is a historical graphical technique, not an algorithm that automatically discovers true cycles.
Nesting up: looking for the higher cycle
Hurst draws the second envelope around the first envelope, not directly around prices. The inner band moves between the boundaries of the outer one; elapsed time between comparable contacts is read as an estimate of the higher component.
On the weekly Dow Jones Industrial Average chart for 1965–1969, the book presents only two intervals: 67 and 75 weeks. Hurst reports their mean as 71 ± 4 weeks and associates it with the nominal cycle of roughly 18 months. The intermediate 9- and 12-month cycles are described as observationally insignificant in that sample.
These figures document what the author saw on that chart. They do not establish that every market has a stable 71-week component.
Nesting down: looking for the sub-cycle
Moving down the scale, Hurst observes the smaller “gallops” between two lows of the longer component. In one example he divides 21.428 weeks by three swings and obtains about 7.14 weeks.
He then refines the reading by joining highs and lows with short segments. In Figures II-5 and II-6 he counts 15 intervals and reports a mean of 6.766 ± 0.8 weeks. For Standard Packaging he instead compares 18.75 and 5.71 weeks, a ratio of 3.28. These too are results from the displayed samples, not mandatory ratios for every series.
How the historical procedure was applied
- Draw a first, nearly constant-width envelope and mark useful contacts.
- To move up the scale, draw a second band around the first and measure the new intervals.
- To move down, count the swings inside one span and seek confirmation with the segmented envelope.
- Change data resolution when weekly observations provide too few points: daily for shorter cycles, monthly for longer ones.
The reading remains provisional at the right edge of the chart. A new observation may alter the envelope and therefore the interpretation of the latest swing.
The statistical limit that must remain visible
The notation 71 ± 4 should not be presented as a modern confidence interval. It comes from the mean of two observations and their elementary spread. NIST explains that the precision of a mean estimate also depends on sample size: additional observations allow uncertainty to be assessed more meaningfully.
A contemporary test would need rules fixed before analysis, out-of-sample data, costs, alternative benchmarks, and stability checks across instruments and periods. Cyclepedia preserves the historical method while separating a source description from modern evidence.
Sources
- J. M. Hurst, The Profit Magic of Stock Transaction Timing, Prentice-Hall, 1970, Ch. 2, pp. 38–41, Figs. II-4, II-5 and II-6 — nesting up, nesting down, and the numerical examples.
- NIST/SEMATECH, Confidence Limits for the Mean — relationship between sample size and the precision of a mean estimate.
Related entries
- Curvilinear envelope — the starting band
- Hurst nominal cycles — the historical classification scale
- Five principles of the cyclic model — variation, nominality, and proportionality
- Hurst tradition — complete path and reading order