CYCLELAB · TURNING POINT RESEARCH

Recurrence rate × Strength: how to read a Turning Zone

Turning Point (TP) is a historical statistic. It describes how often historical highs and lows have repeated at the same cycle phase, and where that repetition is most concentrated. The most important habit is to keep “how often it happens” and “how tightly it clusters” as two separate questions.

Turning Point is an advanced CycleLab research feature. Sections 01–03 on the homepage (Macro Seasonal Summary, Market Seasonal Structure, Stock Seasonal Ranking) are easier starting points. Section 04 — Turning Point — is for researchers who want to go deeper into cycle phase, historical turning structure, and HIGH/LOW clustering.

1. Three questions, three different metrics

REPEATABILITY

Recent recurrence rate

4 / 5 · 80%

Of the latest 5 complete cycles, how many contained at least one same-direction TP inside this Turning Zone. A cycle with two TPs in the zone still counts once.

Answers: has this phase pattern kept repeating recently?

LONG HISTORY

Long-term recurrence rate

14 / 20 · 70%

Of every eligible complete historical cycle, how many hit this same Turning Zone.

Answers: does this phase still hold up across the full history?

PHASE DENSITY

Strength

2.80×

The smoothed TP density at this phase, expressed as a multiple of the cycle's uniform-distribution baseline.

Answers: within the zone, which phase is the TP cluster most concentrated at?

2. How the Strength waveform forms a Turning Zone

The chart below is a schematic Strength waveform for one direction (e.g. LOW) across cycle phase.

1.0×Peak P322.80×LOW Zone P30–P33A small bump only slightly above 1× may just be weak local structure

Existing definition: a local peak with Strength above 1 becomes a candidate Turning Zone, extended outward on both sides until density returns to ≤1. The peak position is the phase where TPs are most concentrated within that zone.

Important: 1.0× is a mathematical threshold for “above the uniform background,” not a quality threshold for “research-worthy.” Whether weak peaks such as 1.2×, 1.4× or 1.5× should be de-emphasized at the product layer needs a separate, real-distribution calibration — the detector itself is never changed to do this.

3. Why Recurrence rate and Strength can't replace each other

Two real patterns can produce the same Recent recurrence rate story very differently. Compare:

A: high Recurrence rate, a wider waveform

Recent recurrence rate: 5/5 = 100%

All five cycles had a LOW, but the positions are more spread out, so the peak isn't necessarily very sharp.

B: lower Recurrence rate, a sharp narrow peak

Recent recurrence rate: 3/5 = 60%

Only three cycles hit, but those three TPs land almost on the exact same phase — so Strength can be very high.

So: Recurrence rate answers “how many cycles took part,” while Strength answers “how tightly the participating TPs stack on the phase axis.” The two are related, but they are not the same thing.

4. Reading Local and Global together

Recent recurrence rateLong-term recurrence rateResearch interpretation
HighHighRepeats both recently and across long history — a relatively consistent pattern.
HighLowThe recent phase pattern has strengthened — possibly a newer structure.
LowHighOnce stable over the long run, but weakening in the most recent cycles.
LowLowLimited historical support — read any local Strength peak conservatively even if it looks high.

5. TP describes historical position, not return

Turning Point: answers “which cycle phase have historical highs/lows tended to cluster at, and is the current position close to that zone?”

Seasonal Outlook (01/02): answers “from the current seasonal position, what have the historical 1W / 1M / 3M up share and average return been?”

TP shows the historical position → 01/02 show historical returns and up share → Historical Paths verify it

TP itself should never be read as “LOW means price must rise” or “HIGH means price must fall.”

5b. Two production details worth knowing

The card's Price Confirmation status has been removed in favor of one compact “Historical adverse excursion” statistic (Typical / Elevated, i.e. median / P75, direction-signed) for the current zone — how far price historically moved against a hypothetical position after entering this zone, before the next matching Turning Point actually occurred. The underlying MA20 confirmation research (V4) remains valid and is preserved as background evidence; it simply no longer drives a live status widget on the card.

When the current phase falls inside both a HIGH and a LOW Turning Zone at once, CycleLab shows this as a single, compact “Turning Zones Overlap” card labeled Historical directional divergence, preserving both zones' evidence side by side rather than picking one as a winner — historical outcomes for overlapping zones have not shown a reliable direction either way.

6. A suggested per-stock TP summary

Illustrative example only — not live data for any specific instrument.

CURRENT TURNING ZONE

LOW · P30–P33

Recent recurrence rate:4/5 · 80%

Long-term recurrence rate:14/20 · 70%

Peak:P32

Peak Strength:2.80×

Currently inside a LOW Turning Zone.

NEXT TURNING ZONE

HIGH · P37–P42

Recent recurrence rate:3/5 · 60%

Long-term recurrence rate:11/20 · 55%

Peak:P38

Peak Strength:1.76×

Distance:≈ 6 months (illustrative)

7. Not yet decided: should weak Strength peaks be hidden?

It may be reasonable to de-emphasize very weak Turning Zones at the presentation layer, but “Strength < 1.5× gets removed” is not currently used as a new detector rule.

The right next step is to first study the distribution of zone-peak Strength across the full instrument universe, cross-checked against recent and long-term recurrence rate. Only if 1.0–1.5× zones mostly also show low recurrence rate and low repeatability would there be evidence to treat them as lower-priority or hidden by default.

Next step: do display-quality calibration first, without changing the TP detector. Even if weak zones are hidden by default in the future, Advanced / Detailed Stats views should still keep the full data available to cycle researchers.

← Research Guide