Hurst Cycle Analysis

Today’s read posts after the close.

J.M. Hurst published his work on price cycles in the early 1970s. It is a public body of ideas, more than fifty years old, and it is the framework I have built my own reading around. This page is the honest advanced overview: what the framework asserts, what follows from it, and where it stops.

If you have not read market cycles 101, start there. This page assumes the basic picture.

The principles, and what each one buys you

Hurst set out a small number of propositions. They are worth separating, because they carry very different weights.

Commonality: price movements in different instruments share cyclic components. This is why cycle work is transferable across markets rather than needing to be rediscovered for each one, and it is why a component visible in an index often shows up in its constituents.

Summation: what you see is the sum of the components. This is the one that explains chop, explains disagreement between analysts, and explains why a chart looks nothing like a wave despite being made of them.

Harmonicity and synchronicity: the components sit in an approximate ladder of related durations, and they tend to reach their troughs together. Together these two are the load-bearing ideas. Without them you have "prices wiggle at various speeds," which is not actionable. With them you get a structure that says a low in a short component is more significant when it coincides with a low in a longer one, and that is the core of the whole practical method.

Proportionality: larger components produce larger price movements. This is what makes the framework an amplitude tool as much as a timing one.

Variation: none of it is exact. Durations vary around a nominal length. This is not a footnote, it is a first-class part of the theory, and everything downstream has to be built to tolerate it.

What follows practically

Because troughs align, the significant moments are alignments. A short-cycle low that lands where a longer-cycle low is also due is a structurally different event from one that does not, even though a price chart shows the same shape. Nests of lows and cycle low confluence are about exactly this.

Because durations vary, you work in windows and never in dates. A trough is expected within a range. If someone gives you a specific day, they have thrown away the variation principle to sound more confident, and the confidence is manufactured.

Because larger components dominate, position within the larger structure constrains what the smaller one can achieve. A short-cycle upturn inside a falling longer component is a weaker proposition than the same upturn with the longer component rising. This is where a great deal of the practical value sits.

Because peaks translate, you count from lows. Covered on the 101 page and in cycle translation.

Where the analysis actually goes wrong

I want to be specific here rather than gestural, because the failure modes are more instructive than the successes.

The count can be wrong. Identifying which low was which component is a judgement made with incomplete information, and a wrong assignment propagates through everything downstream. You do not usually discover this at the time. You discover it when the next expected trough does not appear.

Variation makes the framework hard to falsify, and this is the deepest problem with it. If a trough is late, was the count wrong, or is this variation? Both answers are available, and the temptation to always choose the one that preserves your view is enormous. A framework that can absorb any outcome has stopped being a tool and become a belief.

The only honest defence I know is to write the read down before the outcome, with a date and an invalidation, and then score it including the misses. That is why the record is published with the losses in it and why scoring your own calls is not an afterthought here. Without that, cycle analysis is unfalsifiable in practice regardless of how rigorous it looks.

The framework also degrades in thin instruments, and it can be genuinely disrupted by events large enough to reset the structure. A count can simply stop being valid, and continuing to defend it is the most expensive mistake available.

What I do not publish, and why

I teach the framework openly because it is public, fifty years old, and not mine to gate. What I do not publish is the live computed output: today's specific reads, the exact parameter choices, the scan conditions. Not because they are magic, but because that output is the working product, and a framework plus somebody else's live answers is not education, it is a feed.

The distinction I try to hold is between the map and the position. The map is public and I will explain any part of it. Where I currently think price sits is the work.

Is it worth learning

Honestly, it depends on temperament. It demands tolerance for ambiguity, willingness to be wrong on a count, and discipline about windows rather than dates. It rewards patience and punishes the desire for a mechanical answer.

What it gave me is a way to know where I am inside a move rather than only what the move is doing, and an amplitude expectation, which is arguably more valuable than direction. It also gave me a structure that can be written down in advance and scored afterwards, which is the property that separates a method from a story.

The starting practice is counting a cycle trough, and the line I use most to read turns is on FLD crossings.

Educational market analysis only. Nothing here forecasts any market turn or recommends any position.

Questions traders ask

What is this Hurst Cycle Analysis page?

A dated, running read on hurst cycle analysis from a trading desk that scores its own calls publicly — hits and misses both.

How often is it updated?

After the market close on trading days.

Is this investment advice?

No — educational market analysis only. Nothing here is a recommendation.