Hurst nominal cycles: duration and variation

In the book’s historical sample, Hurst associates a mean duration of about 21 weeks with the component called “26 weeks.” This association depends on analysis of the surrounding oscillations, not just numerical proximity. The name remains 26; an observed duration can be shorter or longer. This distinction between label and actual measurement is central to the nominality principle presented by J. M. Hurst in 1970.

What the ladder shows

The rows organise long, intermediate, and short components as one family. Many steps are close to a 2:1 relationship, but the book's table does not apply one multiplier to every row. It is therefore more precise to describe the ladder as near-harmonic.

In the model, a longer component also has greater nominal magnitude: this is the proportionality principle. Variation explains why an individual oscillation need not last exactly as long as its label.

Table II-1 from the 1970 book

This is a transcription of the table on p. 33. The columns express the same duration in years, months, or weeks where the source supplies the equivalent.

* The original note says that the 26- and 13-week components often appear in the data as a combined effect of about 18 weeks.

The example that makes it clear: “26 weeks” measures 21.4

In Chapter 2, Hurst draws an envelope, a band containing prices, around the weekly Dow Jones Industrial Average (DJIA) from 1965 to 1969 and measures the distance between consecutive lows. After excluding the intervals of 14, 9 and 12 weeks, treated as variants, he retains seven samples: 23, 21, 22, 24, 17, 20, and 23 weeks.

The mean is 150 / 7 = 21.428… weeks. The book treats it as the current expression of the nominal 26-week component. For ±3.5, the book reports extreme deviations of +2.572 and −4.428 weeks and averages their absolute sizes: (2.572 + 4.428) / 2 = 3.5 weeks. This is a descriptive summary, not a standard deviation or confidence interval; the symmetric band does not even contain the 17-week observation.

The selection of samples depends on the envelope analysis. This is therefore a historical reconstruction of Hurst's procedure, not independent proof of its predictive effectiveness.

How to use the ladder on a chart

1. Choose the time scale and identify a coherent sequence of lows.
2. Measure the time between consecutive lows; Hurst generally regards lows as better defined than highs.
3. Calculate the sample's mean duration and spread, recording which observations were excluded and why.
4. Compare the estimate with the most plausible nominal row and its neighbours. A value near 18 weeks may also reflect the combined effect described in the table note.
5. Update the estimate when a new low can be identified with the data then available. To resolve a component graphically, the book requires at least 6–7 data points within a single cycle: this does not mean observing 6–7 cycles.

The envelope and nesting help separate oscillations embedded within one another. A monthly view makes a longer history readable but, over the same historical interval, it adds no observed cycles. Shorter cycles require more closely spaced observations.

From the short view to the long view

Moving down the ladder, Hurst starts from the 21.4-week estimate, divides it by three, and then refines a nominal 6.5-week component to the reported measurement of 6.766 weeks across 15 samples.

Moving up the ladder, intervals of 67 and 75 weeks produce an estimate of 71 ± 4 weeks, compared with the 18-month nominal. On the monthly DJIA from 1949 to 1969, the book also measures four cycles of 52 ± 1 months, close to the 4.5-year label.

These examples show what the ladder is for: giving common names to different measurements without turning historical measurements into universal constants.

What a nominal duration does not tell you

A row in the table does not, by itself, locate the next low, confirm that an oscillation will continue, or provide a buy or sell signal. A close fit in past data can also depend on the chosen window, selected turning points, and other components present in price.

A testable use requires selection rules, tolerances, and available data to be fixed first, followed by checks outside the sample used for estimation. The nominal duration is a working hypothesis; risk must be managed separately.

Intraday extension: Hickson, after the book

David Hickson attributes a ladder down to 5 days to Hurst and claims the extension down to 3 minutes as his own. This later intraday implementation is distinct from the 1970 Table II-1 and is not attributed here to the Cycles Course.

Sources

• J. M. Hurst, The Profit Magic of Stock Transaction Timing, Prentice-Hall, 1970, Chapter 2, Table II-1 p. 33 and measurements pp. 37–41 — printed page numbers; example reconstructed from the scan.
• Google Books record for the work — author, publisher, year, contents, and terminology.
• David Hickson, 10 Core Concepts of Hurst Cycles, concept 5, PDF pp. 6–7 — attribution of the intraday extension to its author.

Links

• Five principles of the cyclic model — summation, variation, nominality, and proportionality
• Curvilinear envelope · Nesting envelope — measurement tools
• Hurst cyclic moving averages — the relationship between span and nominal duration
• Hurst tradition — complete historical path