Where Is the 80-Day Cycle Right Now? Cross-Asset May 2026 Snapshot

A correction, and a sharper question

J.M. Hurst's 1970s nominal hierarchy lists an "80-day" component sitting between the 40-day and 20-week scales. Hurst himself measured the average wavelength of that bucket at 68.2 calendar days, and reminded readers that "cycles are constantly varying in wavelength." The label is a bucket, not a literal period. He also formulated a Principle of Harmonicity: each cycle in his hierarchy is roughly 2x the next-shorter one. A market that does not carry a clean rhythm at the 80-day scale itself can still carry the same nested structure at 40 days or at 20 weeks (≈140 to 190 trading bars). So the right cross-asset question is not "does each market have a peak in 70 to 90 trading bars?" It is "where in Hurst's harmonic family does each market carry the structure right now?" This piece runs that scan on May 5, 2026, correcting an earlier version of this post.

What we did

Twenty-two instruments, 600 daily bars per instrument, across seven asset categories: world indices, US mega caps, sector ETFs, currency indices, commodities, government bonds, and crypto. The 600-bar window covers roughly seven to eight repetitions of an 80-bar cycle, comfortably above the rule-of-thumb minimum for spectral identification at that scale.

For each instrument we scanned three bands in Hurst's harmonic family:

• 0.5x band: 35 to 50 trading bars (Hurst's nominal 40-day)
• 1x band: 70 to 90 trading bars (Hurst's nominal 80-day)
• 2x band: 140 to 190 trading bars (Hurst's nominal 20-week)

Inside each band we picked the strongest peak by fit score, the Pearson correlation between the detected sinusoid and the detrended price series. Fit score is what the platform's "highest alignment" indicator (the orange-dot marker in the analysis UI) ranks on. Each peak's fit rank is its position in the full spectrum, where rank #1 is the strongest sinusoidal fit in the resolvable range. Each market's primary harmonic is whichever of the three bands carries the lowest fit rank. Methodological background is in the cycle detection methods guide and the dominant cycle period walkthrough.

Two priors before the results. Granger's 1966 result on the typical spectral shape of an economic variable warns that price-level series tend to carry smooth, low-frequency-dominated spectra, so narrowband peaks at intermediate frequencies are the exception and demand scrutiny. Sullivan, Timmermann and White's 1999 study of nearly 7,800 technical trading rules over a century of Dow data is the universe-of-tests warning: rules that look significant in isolation often lose force once the size of the search space is acknowledged. Three bands across 22 markets is 66 search slots, a real degrees-of-freedom cost. Bartels' 1932 and 1935 significance test is the methodological lineage for that scrutiny. We are detecting and reporting, not declaring. More on the discipline in overfitting in cycle detection and is cycle analysis reliable.

The headline finding

All 22 instruments carry a top-five fit score in Hurst's 80-day harmonic family. Twelve of them carry the orange-dot rank #1 there. Seven sit at rank #2, two at rank #3, and one (XLF) at rank #4. None of the 22 fall to rank #5 or worse. There is no "missing cycle" in this sample.

The split between which band actually holds the strongest fit is the part worth reading slowly. Eight markets are 1x-dominant, with the strongest fit landing inside the 70 to 90 trading-bar band itself. Fourteen markets are 2x-dominant, with the strongest fit landing in the 140 to 190 trading-bar band, which is the neighborhood of Hurst's nominal 20-week. Zero markets are 0.5x-dominant.

That 14-versus-8 split is the editorial centerpiece. The harmonic family of the 80-day is widely present, but for most of these markets in May 2026 the structure is sitting one octave higher than the 80-day band itself. This is exactly the pattern Hurst's Principle of Harmonicity predicts when a longer scale of the nested structure dominates the local window. The 80-day is not absent. It is just not the loudest voice in the room. The cycle hierarchy is intact and detectable; the bucket label "80 day" was the wrong precision target. Deeper version of this nesting argument in multi-timeframe cycle nesting.

Caveat: 66 search slots is wide. Some of the 22 top-five hits would land there by chance from filtered noise alone. The 12-of-22 rate of rank #1 hits is harder to dismiss but is not proof. We are reporting structure consistent with Hurst's hierarchy, not claiming a tradeable law.

Where the rhythm actually lives

Group the 22 markets by which harmonic they carry primarily and the asset-class signal jumps off the page.

Primarily 2x (Hurst 20-week scale, 140-190 bars): Apple, NVIDIA, XLE, US Dollar Index, Japanese Yen Index, Euro Index, gold, crude oil, silver, the 10-year Treasury yield, TLT, the 13-week Treasury yield, Bitcoin, and Nikkei 225. Fourteen instruments, dominated by commodities, currencies, bond markets, and a slice of mega-cap and Asian equity.

Primarily 1x (Hurst 80-day scale, 70-90 bars): Dow Jones Industrial, Nasdaq Composite, Hang Seng, Euronext 100, Microsoft, XLK, XLF, and Ethereum. Eight instruments, weighted toward equity-index aggregates and US tech.

Primarily 0.5x (Hurst 40-day scale, 35-50 bars): none.

The pattern worth flagging without overclaiming: the longer harmonic concentrates among macro-driven instruments (commodities, FX, rates), and the shorter harmonic concentrates among equity aggregates and tech. One observation from one window. Worth watching, not a thesis. For instrument-level reads see our crude oil seasonal structure, gold macro cycle structure, EUR/USD central bank cycles, Treasury bond cycles, Nasdaq cycle structure, and Bitcoin beyond halving write-ups.

Phase positions across the family

Reading phase off each market's primary harmonic, the cross-asset distribution is almost evenly split. Six are Rising, six are Peaking, six are Falling, and four are Bottoming. The naive everything-moves-together story does not survive the corrected lens. Phase here uses the standard four-quadrant convention: Rising climbs toward the next peak, Peaking sits at or just past the top, Falling descends toward the next trough, and Bottoming sits at or just past the bottom. Convention details in cycle phase interpretation and four phases of market cycle.

Three contrasts read cleanly under the harmonic-aware lens.

Gold is bottoming on its 2x harmonic. Crude oil is peaking on its 2x harmonic. Same scale, opposite phases. Gold near $4,560 per ounce sits well off its January 28, 2026 record above $5,600 and reads as a low-phase point on its 169-bar fit (rank #1, fit 0.34). Crude near $104 per barrel is near a local peak on its 161-bar fit (rank #1, fit 0.48). The energy-versus-safe-haven divergence dominating headlines since the Iran war began in February shows up at the cycle level as a phase split rather than a level correlation.

Bitcoin is rising on the 20-week scale. Ethereum is bottoming on the 80-day scale. Crypto is not monolithic on either dimension. BTC near $80,000 reads as climbing on its 176-bar fit (rank #1, fit 0.40). ETH near $2,377 carries its strongest fit on the 1x band, an 81-bar component (rank #2, fit 0.13) reading as a low-phase point. Different harmonic, different phase, same asset class. More in Bitcoin beyond halving.

The currency triangle reads internally consistent on the 2x scale. DXY at 98.48 sits in a Rising phase on its 178-bar fit (rank #1, fit 0.52, the cleanest read in the sample). The Euro Index is Falling on its 176-bar fit; the Japanese Yen Index is Bottoming on its 160-bar fit. Three currencies, one harmonic family, three parts of the wave, all rank #1. One tension worth flagging: DXY is sitting near multi-month lows on the level chart even while the 2x harmonic reads as Rising. Cycle and level diverge here, a useful reminder about the limits of any single lens. Macro context in EUR/USD central bank cycles and Treasury bond cycles.

The backdrop motivating those phases without explaining them: the FOMC held the target range at 3-1/2 to 3-3/4 percent at the April 28-29 meeting with four dissents, the most since late 1992; the statement cites Middle East uncertainty; the Strait of Hormuz blockade has lifted WTI roughly 50 percent since February. Phases are positions on a fitted sinusoid against a real macro tape, not predictions.

The Nikkei and Apple cases

Two instruments most clearly demonstrate why the harmonic family matters.

Nikkei 225: the 1x band carries an 81-bar peak with fit 0.08 at fit rank #11. That is essentially noise. The 2x band carries a 159-bar peak with fit 0.18 at fit rank #2. The cycle structure is present and cleanly identified; it sits at the 20-week scale rather than the 80-day band. A scan that looked only inside 70-90 bars would have flagged Nikkei as cycle-weak.

Apple: the 1x band carries an 83-bar peak with fit 0.00 at fit rank #18, below the noise floor. The 2x band carries a 161-bar peak with fit 0.37 at fit rank #1. Apple has the cleanest read in the entire 22-market sample. A scan that looked only inside 70-90 bars would have flagged Apple as the lone outlier; an earlier version of this post made that exact mistake. The corrected reading: Apple is not an outlier. Apple is a textbook example of a market whose dominant structure sits at the 2x harmonic, and the platform's rank-#1 marker would put the orange dot at 161 bars, not 83.

This is the cleanest illustration of Hurst's per-issue point: "a specific, or current, cyclic model must be tailored to each issue." The hierarchy is the same. Which scale dominates the local window is not.

Fit and regime: still tempered

Fit scores vary widely. The 1x reading on XLE comes in at fit 0.19 (rank #4) and would have looked weak alone; the 2x reading on XLE comes in at fit 0.45 (rank #1), the second-cleanest read in the sample. DXY's 2x reading at 0.52 is the cleanest. ETH's 0.13 is the weakest of the 22 primary-harmonic fits, still rank #2 in its own spectrum. The cleanest reads cluster among the 2x-dominant macro instruments.

Regime is the second filter. The multi-method composite (Hurst exponent, detrended fluctuation analysis, fractal dimension, volatility scaling) classifies almost every market as random-walk on this 600-bar window. Three exceptions: ^IRX and ETH-USD carry trending composites; TLT carries a mean-reverting composite. That is consistent with Granger's 1966 prior: even where a clean cyclic component exists, the broader spectrum mostly carries the smooth declining shape with limited persistent directional memory at the daily timeframe. Our Hurst exponent guide covers what these regime labels mean. To disambiguate: the Hurst exponent is Harold E. Hurst's hydrology statistic from the 1950s, a different person and a different result from J.M. Hurst's 1970s cycle hierarchy that defines the 80-day bucket. The platform uses both.

What this means and what it doesn't

It means: a daily-timeframe cyclic component in Hurst's 80-day harmonic family is widely present across this 22-market cross-section in May 2026, more often as the 2x harmonic than the 1x band itself, and the dominant-harmonic split between macro instruments and equity aggregates is the most interesting observation. Phases are not synchronized; rank-#1 fits are common; the precision of the scan matters because the bucket label is not the literal period.

It does not mean the 80-day bucket is tradeable in isolation. Noise-floor filters cull the most obvious random matches, but the universe-of-tests cost remains real and cycle phase is one input among several. It does not mean the cycle "explains" macro outcomes. The Iran war, four FOMC dissents, near-record gold, and the Hormuz blockade are doing the explanatory work; the cycle is positioning against that tape, not driving it. It does not mean the phase reads will hold out-of-sample. The fitted sinusoid is built on the 600-bar window we measured; Hurst himself reminded readers that wavelengths drift. To repeat the precision point: an 80-day cycle is a nominal label, not an exact period; this is not a 20-week cycle by either calendar or trading-week convention (those land near 100-140 trading bars), and it is not a Kitchin (~40 month) cycle. Background in cyclical analysis explained and understanding market cycles.

Full snapshot, by category

Every row was computed on May 5, 2026 from 600 daily bars. The mini chart marks current phase position with a colored dot. The period column shows each instrument's primary harmonic in Hurst's 80-day family (1x = 70-90 bars, 2x = 140-190 bars). A ★ marks instruments where that primary harmonic is rank #1 by fit score in the full spectrum (the platform's "highest alignment" indicator). Hurst column is the 600-bar Hurst exponent; regime is the multi-method composite.

Methodology pointer and what to do with this

For the broader methodology note the platform produces each month, see the cycle state report for April 2026. To run a regime check against your own watchlist using the same method, the regime check tool is the entry point. The regime indicator stack, the spectral peak detection, and the Bartels-style significance gate are all surfaced there. Snapshot articles like this one are useful as cross-asset context. The per-instrument question is what each market is doing inside its own harmonic family.

FractalCycles is a quantitative analytics platform built on three pillars: J.M. Hurst's cyclic principles, spectral analysis via Goertzel DFT, and Hurst exponent regime detection. Built for market analysts, investors, and traders who want disciplined cycle measurement rather than narrative cycle commentary.