The short version
The Economic Confidence Model — ECM for short — is a cycle framework built by the forecaster Martin A. Armstrong. Its central claim is that the movement of capital, and the confidence that drives it, are not random. Capital concentrates, peaks, breaks and resettles on a recurring interval of roughly 8.6 years.
The part that makes the model unusual is not the existence of a cycle — almost every school of economics accepts that activity moves in waves. It is the claim that the interval is fixed, and that because it is fixed the turning points can be written on a calendar years before the event rather than identified afterwards.
8.6 × 365.25 = 3,141.15 days. π × 1,000 = 3,141.59. The correspondence between the two is the origin of the model’s nickname, the pi cycle.
What the model actually measures
Confidence — not output. This distinction does most of the work in the book, and it is the thing most summaries of the ECM get wrong. Armstrong’s framing is that panics and recessions are, underneath the statistics, events in what people believe about institutions: banks, currencies, courts, governments. Belief is what moves capital, and belief is what the cycle is supposed to track.
The practical consequence is that an ECM turn does not have to coincide with a GDP print or an official recession date. A turn might surface as a currency crisis, a sovereign default, a bank failure, a political rupture, or simply a change in where the world’s money wants to sit. Armstrong’s argument is that these are the same event wearing different clothes.
Where the 8.6 years came from
By Armstrong’s own account he was not looking for pi. He was testing whether the crash of 19 October 1987 sat at a precise distance from earlier panics or had simply happened to land where it did. Counting the days inside an 8.6-year business cycle returned 3,141 — and, as he put it in a 1999 essay, “suddenly, there was clearly more at work than mere coincidence.”
Whether that is a discovery or a numerical accident is the argument the model has been having with its critics ever since. The arithmetic is worth doing yourself, and it is more interesting than either side usually admits. We walk through it here →
The cycles above and below
The 8.6-year wave is not the whole model. Armstrong nests it inside a set of longer and shorter intervals, each derived from the base by multiplying by four or by six.
| Level | Length | Days | Composed of |
|---|---|---|---|
| Quarter cycle | 2.15 years | ~785 | one quarter of the base wave |
| Base cycle (ECM) | 8.6 years | ~3,141 | four quarter cycles |
| Wave | 51.6 years | ~18,847 | six base cycles |
| Major wave | 309.6 years | — | six waves |
The nesting is the reason practitioners of the model talk about turns of different magnitude. A quarter-cycle turn is expected to register as a wobble; a turn that lands where a base cycle, a wave and a major wave coincide is expected to register as history. In the model’s own terms, the size of the event should scale with how many levels line up at the same date.
How a single cycle is supposed to behave
- Rising confidence. Capital concentrates into whichever market or nation is winning; optimism compounds and valuations stretch.
- The turn. The model’s signature moment — the point at which confidence in the concentration breaks.
- Unwind. Money does not evaporate; it leaves. Capital flees toward whatever is perceived as safe, which is rarely where it just was.
- Reset. Capital resettles, a new concentration begins, and the count starts again.
This is why the second of the book’s five truths is about migration rather than destruction. In the ECM’s account, a crisis is a relocation of capital that looks like a disappearance only if you are standing in the place it left.
Read the objections before you read the record.
Mainstream business-cycle research does not recognise a pi-derived cycle. It treats expansions and contractions as variable in length, driven by credit conditions, shocks and policy, and it would regard a fixed constant as the wrong shape of explanation.
The 8.6-year-to-pi correspondence rests on a day count, and the day count depends on how you define a year. It is a numerical coincidence rather than a result derived from any economic first principle.
Critics also argue that the model’s apparent hit rate benefits from hindsight: a fixed grid laid across four decades of eventful markets will land near something most times, and the turning points are defined loosely enough to absorb a miss.
Armstrong’s own history is part of the picture too, including a United States conviction for investment fraud and eleven years served. That record is set out here →
None of this settles the question, and none of it is a reason to look away. It is a reason to read the record yourself rather than take any summary — including this one — on trust.
Where the book comes in
The Armstrong Economic Code is Kerry Lutz’s attempt to put the whole framework in front of a general reader: how the model is constructed, which turns Armstrong called and when, how the post-2020 record extends the same chart, and what a reader might reasonably do about the next turn. It runs 302 pages and carries a foreword by Armstrong himself.
If you want the model’s claims in their own words, with the charts, the book is the place. If you want the arithmetic first, start with the 8.6-year cycle.