Skip to main content
Track Spectral analysis
10 / 21
Search the archiveLibrary
Library

2007issue C041-3

When cycle mode fails: regime gates for spectral trades

A measured dominant cycle is only a candidate setup. Cycle mode waits until signal-to-noise, trend slope versus cycle amplitude, and half-cycle trendline persistence all stay inside their workable bands.

  • A measured dominant cycle can change over time and is not treated as a single fixed period of the series.
  • When bar-range noise equals cycle peak amplitude, a theoretically perfect valley-to-crest trade can still finish at breakeven if the entry fills at the bar high and the exit fills at the bar low.
  • Cycle mode is intended only when measured trend slope stays inside the plus-and-minus cycle-amplitude band and measured signal-to-noise exceeds a 6 dB threshold.
  • Knowing that a cycle exists, including its period and current phase, is not treated as sufficient if cycle amplitude is too small relative to noise or trend slope.
Entries in this reading3 entries
Current trackSpectral analysis63 readingsDominant cycle detection180 readingsTrend filter184 readings

Cycle work as a permission problem

The historical workflow starts from a measured dominant cycle: the time-varying cycle period extracted from ordered market observations and treated as the signal a cycle-mode trade would try to ride.

Editorial: TradersWeek reads the same workflow as a permission problem, not a period-finding problem. The measured cycle remains a candidate setup until three independent regime gates stay inside their workable bands. Those gates are signal-to-noise, trend slope versus cycle amplitude, and half-cycle trendline persistence.

The period itself can move

Spectral analysis is used as a time-synchronized spectral display of how cycle energy shifts across periods. That display is there to show that a dominant cycle is not a fixed property of the series.

A measured dominant cycle in a market series can change over time rather than remaining a single fixed period.

Regime dependence of the same cycle

The trend filter is a set of regime checks that compare cycle amplitude with bar-range noise, trend slope, and how long price stays on one side of an instantaneous trendline, then flag cycle mode versus trend mode.

Regime dependence means the same measured cycle can be tradable or not depending on noise, trend strength, and trend persistence at that moment.

When noise matches cycle amplitude

Signal-to-noise ratio compares cycle peak amplitude with bar-range noise and expresses that comparison on a decibel scale so a threshold can mark when cycle-mode entries are theoretically workable.

When bar-range noise amplitude equals cycle peak amplitude, a theoretically perfect valley-to-crest trade can still finish at breakeven if the entry fills at the bar high and the exit fills at the bar low.

A 6 dB signal-to-noise case is defined as cycle peak amplitude twice the noise amplitude, and expected profit in that noisy case is half the profit of the noise-free case.

When trend slope swamps the cycle

If a cycle component and a linear trend are added, and the trend slope across one cycle period equals twice the cycle peak amplitude, a short-at-crest and cover-at-valley trade realizes about half the profit of the same trade with no trend.

Trend swamping is the additive case in which that slope is large enough to cancel most of the profit that a perfect crest-to-valley cycle trade would have shown without the trend. Trend swamping is treated as workable when the trend slope across the cycle period exceeds twice the cycle amplitude.

When price stays on one side of the trendline

An instantaneous trendline is a local trend component that a cycle is expected to cross about every half period. A long stay on one side is treated as a reason to avoid cycle mode.

Price is expected to cross that line about every half cycle, and cycle-mode trading is avoided when price has not crossed it within the last half cycle.

Cycle mode only inside the workable bands

Cycle mode is intended only when measured trend slope stays inside the plus-and-minus cycle-amplitude band and measured signal-to-noise exceeds a 6 dB threshold.

A trading mode indicator is a consolidated flag that is high when the trend-mode regime is favored and low when the cycle-mode regime is favored after the amplitude, noise, and persistence checks. A consolidated mode indicator can be high for trend mode and low for cycle mode after those amplitude and noise conditions are combined.

Knowing that a cycle exists, and even knowing its period and current phase, is not treated as sufficient for cycle-mode trading if cycle amplitude is too small relative to noise or trend slope.

T-bond signal-to-noise versus the 6 dB cycle-mode gate

On the daily 30-year Treasury continuous contract, measured SNR only stays above Ehlers’ 6 dB cycle-mode floor in two long stretches — roughly October 2004 through February 2005, then again in April. When the cyan series is below that line, bar range is too large relative to the cycle to take the cyclic trade. Points were read from the published MESA8 SNR subgraph; the last print on the figure is 4.23 dB and the gate is the 6 dB threshold stated in the article.
On the daily 30-year Treasury continuous contract, measured SNR only stays above Ehlers’ 6 dB cycle-mode floor in two long stretches — roughly October 2004 through February 2005, then again in April. When the cyan series is below that line, bar range is too large relative to the cycle to take the cyclic trade. Points were read from the published MESA8 SNR subgraph; the last print on the figure is 4.23 dB and the gate is the 6 dB threshold stated in the article.30-year Treasury bond continuous futures (CBOT US) · Daily · 2004-05-01T00:00:00.000Z to 2005-05-31T00:00:00.000Z

Digitized from the raster of the SNR pane. Vertical labels sit 4 dB apart, so intermediate values are good to about half a decibel except the closing print 4.23, which is taken from the on-chart readout. Dates follow the printed May 2004–May 2005 month axis.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
10 of 21 in the Spectral analysis track
20081-4 pp.Next on Spectral analysisConstructing singular-spectrum forecast baselinesAn observed market series can be treated as a slow trend, an oscillation whose period and amplitude may change, and a residual random component, and singular spectrum analysis forms those parts from the series itself.
All readings on this track · 21 readings
  1. 1982Spectrum-first construction of moving-average trend filters
  2. 1990Crowd questions that switch cycle regimes
  3. 1993Antithetic twins to cancel origin-linked forecast bias
  4. 1994Constructing frequency response from a three-stage filter cascade
  5. 1994Bench-testing smoothers with cycle phase and power
  6. 1995Constructing minimum tradable cycles from spectral noise
  7. 1995Spectral whipsaw limits for a price-crossing average
  8. 1996Constructing market phase from dominant-cycle components
  9. 2000Constructing a Hilbert squelch dominant-cycle gate
  10. 2007When cycle mode fails: regime gates for spectral trades
  11. 2008Constructing singular-spectrum forecast baselines
  12. 2008Stacking a midband cycle gate on a short-memory trend filter
  13. 2011Engineering spectral cycle detection and moving-average harmony
  14. 2011Evaluating fixed oscillators against a moving dominant cycle
  15. 2011A construction checklist for a dominant-cycle sinewave
  16. 2012Assembling a market-mode heatmap from successive band-pass filters
  17. 2012Adaptive trend detection from dominant-cycle spectra
  18. 2015Constructing price paths from cycles and support
  19. 2018Oscillator distribution shape for mean-reversion filters
  20. 2020Constructing a cosine-correlation cycle and phasor trend filter
  21. 2020Constructing a correlation-cycle market-state engine
By fieldCycles and spectral analysis129 readingsBy year2007144 readings
All 63 readings tagged Spectral analysis
Also on Spectral analysis5 readings
2008Stacking a midband cycle gate on a short-memory trend filter2020Constructing a cosine-correlation cycle and phasor trend filter2020Constructing a correlation-cycle market-state engine1984How to construct a maximum-entropy cycle model1984Building moving-window spectral cycle forecasts