What Is the 80-Day Cycle in the Stock Market?

The 80-day cycle is a repeating rhythm in stock prices: a meaningful low, a rise to a peak, and a decline into the next meaningful low, averaging about 68 calendar days from trough to trough. It is one rung of the nominal cycle model that J.M. Hurst extracted from decades of market data in the 1960s, and it is the cycle I build my own trading around — long enough to matter, short enough that the market hands you a fresh one about five times a year.

That is the whole answer in two sentences. The rest of this page is the detail: where the cycle comes from, what sits inside it, what it looks like on a real chart, how I use it day to day, and — just as important — what it cannot do. I will say the limits out loud as we go, because a tool you only know the good side of is a tool you don't actually know.

Where does the 80-day cycle come from?

In the 1960s an aerospace engineer named J.M. Hurst put early computers to work on decades of market price data — he was the first analyst to do it — and found that prices move in a nested set of cycles, each roughly twice the length of the one below it. He published the resulting table as the nominal model, built a 1,200-page correspondence course around the method (long out of print), and the work was then largely forgotten. Not because it was wrong. Because almost nobody went back and read it.

The 80-day cycle is one row of that table. The idea underneath it is what Hurst called the principle of cyclicality: price is not one wave, it is a combination of many waves of different lengths, all running at once. I like the symphony picture. Every cycle is an instrument. None of them plays alone, and price is simply the sound of all of them at once.

I found Hurst's work around 2011, and about two years later I found the FLD research that finally made it click as a trading approach. Since then, the 80-day cycle has been the frame I read every chart through.

Why is it called 80 days when it averages 68?

The names in the nominal model are labels, not stopwatches. The measured average length of the "80-day" cycle, across many years of data, is about 68.2 calendar days from trough to trough. Real cycles come in anywhere from roughly 60 to 80 days. That slop is not a flaw in the model — Hurst named it too: the principle of variation.

And the variation itself carries information. Shorter 80-day cycles — the ones that finish under the 68-day average — have tended to run bullish. Longer ones, up in the 75-to-80-day range, have tended to run bearish. So the day count is never just bookkeeping. The count is part of the read.

If you remember one number from this page, make it 68. When I count days forward from a trough, 68 is the number I am counting toward.

What sits inside an 80-day cycle?

Cycles nest, almost always by a factor of two. Hurst called this harmonicity. Two 10-day cycles make a 20. Two 20s make a 40. Two 40s make an 80. And two 80-day cycles make the 20-week cycle above them.

The full structure runs much higher — 20 weeks, 40 weeks, on up into cycles measured in years. I keep a separate page on the whole thing: the nominal cycle ladder.

Two more of Hurst's principles turn this nesting from trivia into something useful. First, synchronicity: the cycles phase themselves so their troughs coincide. When an 80-day cycle bottoms, the 40, the 20, and the 10 inside it are all bottoming on the same day. That is why real cycle lows stand out on a chart — you are looking at the day several instruments hit their low note together. Peaks do not line up the same way, which is why tops are rounded and messy while bottoms are sharp. Second, proportionality: bigger cycles make bigger moves. An 80-day cycle moves price further than a 20-day cycle does, and a 20-week trough launches a move that is generally faster, cleaner, and longer-lasting than an ordinary 80-day trough — because on that day, even more cycles are pushing together.

On a chart, each 80-day cycle draws an "M" shape. The first leg of the M is the first 40-day cycle; the second leg is the second. The 40-day cycle turns prices over midway through the 80, then — in bullish cycles — carries them back up to a higher peak. Whether the second half of the M runs bullish, bearish, or neutral depends on the pull of everything larger: the sum of all cycles longer than the chart shows, which a cycle trader calls sigma-L. Sigma-L is just a precise word for trend.

What does the 80-day cycle look like on a daily chart?

Start with the trough. An 80-day cycle low is a prominent daily low — the kind price rises away from and does not revisit for weeks. Because the smaller cycles bottom on the same day, the real one usually looks obvious once you know what you are looking for. On my charts, cycle troughs are marked with diamonds along the bottom — a phasing analysis, one row per cycle. Finding and confirming a trough is its own skill, and I wrote it up separately: how to identify a cycle low.

Then there is the line that tracks the cycle: the FLD, the Future Line of Demarcation. It is an average of past prices displaced forward in time by half a cycle — that forward displacement is the entire trick. When price crosses up through the 20-day FLD, a 20-day cycle trough has just formed behind you. When price crosses down through it, a 20-day peak has. You are not predicting the turn; you are reading confirmation that it happened. In the first half of an 80-day cycle the line tends to act as support. In the second half it tends to act as resistance. Same line, different half, different job. The full explanation lives at what is an FLD.

Price and the 20-day FLD meet eight times in every 80-day cycle, and the method letters those meetings A through H. Each letter has its own character, its own typical timing, and its own typical way of failing — and the same eight-letter sequence repeats, cycle after cycle. That alphabet gets its own page: the eight FLD interactions.

How do traders actually use the 80-day cycle?

My daily use of it compresses to five steps:

1. Find the most recent 80-day trough. The prominent low, two to three months back, that price rose away from and never revisited.
2. Count calendar days from that trough to today. Weekends count. The number is the cycle's age: "day N of the 80-day cycle."
3. Check price against the FLD. Which side of the line is price on, and which way did the last cross go?
4. Name the position. Day count plus FLD side gives you the current lettered interaction.
5. Form the expectation — and the invalidation. Each position carries a typical next behavior, and a clear line on the chart that would prove the read wrong. Both get written down before the market opens.

That is it. Notice what the output is: not a signal — a position. A red day early in a young cycle is a different event from a red day late in an old one, even when the two candles look identical on the screen. Once you can tell them apart, most of the noise goes quiet, because you finally know which things are yours to act on and which are just the cycle breathing.

And because there are only about five 80-day cycles in a year, patience is built into the method rather than bolted on. The market hands you the same repeating structure again and again. You do not have to chase any single pass through it.

What does the 80-day cycle look like in a scored record?

Reading about a cycle is one thing. Watching a read resolve is another. Every call I make on air is logged the day it is made and scored later against what price actually did — hits and misses both. Here is an 80-day cycle read from my own ledger, resolved:

July 9, 2026 — S&P 500, 80-day cycle peak read — HIT. On the 4PM show I said the 80-day cycle was "now probably peaking," with long-term buyers still trying to hold the market up late in the cycle. What happened: the index printed a marginal higher peak within a session, held near that area for a few more days, then rolled over — by July 17 it was trading below where the call was made. A peak formed within days of the read, and the late buyers were underwater a week later. Scored a hit.

The same ledger holds misses from the very same week — including a timing call that ran one to two sessions late, and an index-futures read whose stated invalidation triggered. I keep them all, because a record with the losers removed is not a record. How calls are logged and scored — what counts as a call, what counts as resolved — is written up on the methodology page.

What can't the 80-day cycle do?

Honest limits, stated plainly:

• It does not run on a schedule. The 60-to-80-day spread is real. A count gives you a window, never an appointment.
• It does not forecast news. In cycle terms, events change the size of moves more than the timing structure — and size still matters when you are holding a position through one.
• It does not remove judgment. Choosing the trough is a skill. Two honest analysts can disagree on one by a few days, and the method has to survive that. It does — but only if you hold your count loosely and recount when the evidence changes.
• It can fail outright. Most declines respect the cycle map. Crashes and melt-ups are what it looks like when a cycle fails instead — rare, real, and worth understanding before you ever lean on a count.
• It promises nothing. A cycle read is a structured expectation with a stated invalidation. That is everything I will ever claim for it.

Questions traders ask

Do I count calendar days or trading days?

Calendar days, weekends included. The 68-day average — and the 60-to-80 range around it — are measured in calendar time, so a count made in trading days will run short and place you in the wrong part of the cycle.

Is the 80-day cycle only in the S&P 500?

No. Hurst's principle of commonality says the big markets tend to make their important lows and highs together, so one good chart reads for nearly all of them. I do my daily cycle work on the S&P 500 daily chart, and the concepts carry anywhere charts are used to trade — indexes, sectors, individual stocks.

How many 80-day cycles happen in a year?

About five, at the 68-day average. Five fresh troughs, five M-shapes, five passes through the full A-to-H sequence. The same structure, handed to you again and again — which is why this method rewards patience over activity.

What happens when my count turns out to be wrong?

You recount, and you move on. If the count runs past 80 days, the usual culprit is a trough one cycle too far back — look for a newer prominent low and count from there. Variation is a principle of the model, not an exception to it. Meeting it early is part of learning the method, not a failure of it.

Keep reading