How to Identify a Cycle Low on a Daily Chart

A cycle low is a prominent trough on the daily chart — a low that price rose away from and did not come back to — and you identify one with three checks: prominence (it stands out from every low around it), spacing (the distance back to the previous trough fits the cycle's average length), and confirmation (price crosses up through the FLD shortly after it). Pass all three and you have your trough. Count forward from it and you know the cycle's age, which is the single most useful fact in this whole method.

This page is the practical walk-through: why lows are the thing we count, what a valid one looks like, the step-by-step count, and — because I would rather you meet them here than on a live chart — the specific mistakes that break a count.

Why do cycle traders count lows instead of highs?

Because the structure of the market makes lows datable and highs blurry, and that is not a matter of taste — it falls straight out of two of Hurst's principles.

Cycles nest by a factor of two (harmonicity), and they phase themselves so their troughs coincide (synchronicity). Put those together: when an 80-day cycle bottoms, the two 40-day cycles, four 20-day cycles, and eight 10-day cycles inside it are all bottoming on the same day. Every cycle in the stack plunges and reverses in unison, and the chart prints a sharp, deep, well-timed V. Peaks get no such treatment — the rungs crest at different times, so tops come in rounded, drawn out, and staggered, and the trend drags a cycle's visible high off-center besides. You can time a bottom precisely; you can only bracket a top loosely. The full story of that asymmetry is here: why bottoms form faster than tops.

So the entire bookkeeping of cycle analysis runs on troughs. Get the troughs right and everything downstream — the day count, the expectations, the letters — has a foundation. Get one wrong and everything downstream is wrong with it. That is why this page exists.

What does a valid cycle low look like?

Before any counting, you should be able to recognize the animal on sight. A valid cycle low of the 80-day class has a short list of features:

• It is the deepest low in its neighborhood. Not one low among many — the one that stands out at a glance, two to three months back.
• Price rose away from it and did not revisit it for at least a couple of weeks. A low that got retested and undercut a few days later was not the trough; the undercut was.
• It looks obvious, not faint. Because every cycle in the stack bottoms together, you are looking for the day several instruments hit their low note at once. If you are squinting and talking yourself into a candidate, it is probably the wrong one.
• On a phased chart, the diamonds stack there. My charts carry a phasing analysis — rows of diamonds along the bottom, one row per cycle, marking troughs. An 80-day trough is also a 40, a 20, and a 10-day trough, so the diamonds pile up under the same bar.

How do I count a cycle trough, step by step?

Here is the exercise exactly as I would give it to a new trader. It takes about fifteen minutes on any free daily chart. I use the S&P 500 index, daily bars, because the big markets tend to make their important lows together, so one good chart reads for nearly all of them.

1. Open the chart. S&P 500, daily bars, zoomed out to about six months of price.
2. Draw the FLD. The Future Line of Demarcation is price's own track displaced forward in time by half the cycle's length — the concept is the whole recipe. The line should extend past the last price bar into empty space on the right. If it stops at the current bar, the displacement didn't take, and the displacement is the whole point.
3. Find the candidate trough. Scan back two to three months for the deepest daily low in the window — the one price rose away from and did not revisit. That prominent low is your candidate 80-day cycle trough.
4. Count the days. Count calendar days — weekends included — from that trough to today. The number is the cycle's age: "day N of the 80-day cycle." Write it down.
5. Check the spacing. If your count is already past 80, your trough is almost certainly one cycle too far back. Find a newer prominent low and recount.
6. Read the line. Note which side of the FLD price closed on today, and which direction the last cross went. A cross up through the line says a 20-day trough — possibly larger — formed just behind you.

Those two coordinates — the cycle's age in days and price's position against the FLD — are what the whole method runs on. Everything I do during market hours starts from exactly those two facts.

How does spacing confirm the count?

Each cycle has a measured average length, and the distance from your candidate trough back to the previous trough of the same class should sit near it. This is the ruler you check every candidate against:

The spread is real — Hurst called it the principle of variation — and it even carries information: 80-day cycles that finish short of the 68-day average have tended to run bullish, and the long 75-to-80-day ones have tended bearish. But if your candidate implies a spacing that is wildly off on every rung at once, the problem is your count, not the market. The full structure of these lengths lives on its own page: the nominal cycle ladder.

How does the FLD confirm a trough?

The FLD is past price displaced forward by half a cycle, which gives it one very clean property: when price crosses up through it, the cycle it tracks has already troughed behind you. You are not predicting a turn. You are reading confirmation that one happened. That makes the line the third check on any candidate trough — a prominent low with the right spacing, followed shortly by an upward FLD cross, is about as confirmed as a trough gets in this method.

The line has a second use worth knowing even at the counting stage: in the first half of an 80-day cycle it tends to act as support, and in the second half as resistance. So price's behavior around the line is itself a rough age check on your count. The complete treatment is here: what is an FLD.

What mistakes break a count?

Every one of these is common, and every one of them I have made myself:

That last row is the one I want you to keep. A count is a hypothesis. The chart gets the final grade, and you should know before the fact what a failing grade looks like.

What does a broken count look like in the real record?

Every call made on my show gets logged the day it is made and scored later against what price did — hits and misses both. Here is a miss, because a page about counting troughs owes you one:

July 9, 2026 — S&P futures, rising into a 40-day peak, built on a trough count — MISS. The desk laid out the phasing on air: 20-day troughs formed, an early-June 80-day upswing underway, price rising into the next 40-day peak. The invalidation was stated in the same breath: a close below that day's session low would mean the phasing was wrong. What happened: by July 17 the S&P cash index had closed below that stated invalidation level. The count was wrong, and the pre-stated line caught it. One caveat, logged in the ledger at scoring time: the read was made on the futures and scored against the cash index as a proxy, so the outcome is very likely but not perfectly certain. It goes in the book as a miss either way.

That is what an invalidation is for. The count failed; the process didn't. The read carried its own tripwire, the tripwire fired, and the ledger recorded it — which is the whole difference between a method and a mood. How every call gets logged and scored is written up on the methodology page.

Questions traders ask

Do I count calendar days or trading days?

Calendar days, weekends included. All the averages — 68 for the 80-day cycle, 34 for the 40, 17 for the 20 — are calendar-time measurements. A trading-day count runs about 30% short and will misplace you inside the cycle.

What if my count differs from another analyst's by a few days?

Then you have met the principle of variation, and it is better to meet it early. Two honest analysts can disagree on a trough by a few days, and the method survives that — the disagreement is almost always the trough choice, not the arithmetic. Look at which low the other count implies, recount from there, and see which one the spacing and the FLD cross support better.

Can software find the trough for me?

Software can make the cycles much easier to see — drawing the FLD, marking candidate lows, keeping the day count. But phasing is an inexact art, not a mechanical procedure, and the trough call is still a judgment. I treat any automated phasing as a draft to be checked against prominence, spacing, and the FLD, not as an answer.

How do I know a low is the 80-day trough and not just a 20-day one?

Depth and spacing, together. An 80-day trough is simultaneously a 40, 20, and 10-day trough — every cycle in the stack bottoms there — so it prints deeper and more obviously than a lone 20-day dip. Then check the ruler: the candidate should sit roughly 60 to 80 calendar days after the previous 80-day trough. A prominent low only 17 days after the last confirmed 80-day trough is a 20-day low, no matter how dramatic it looked that afternoon.

Keep reading