Fraktalicious

Live  ·  Phase 6 of 6  ·  3 September 2026

You are
here

The model says Wait in Stables

Fraktalicious is a full-cycle navigation system for Bitcoin. Six phases, from accumulation through rotation to exit and back to cash. It tells you which one you're in, and what to do next.

Nothing to do. Sit in stables until the model sees a confirmed turn, not just a cheap price.

See today's read Every call, including the wrong ones
Accumulate Hold Eth Alts Exit Stables
Bitcoin
$77,189
From all-time high
−38.2%
Cycle score
−0.022
Market state
Early recovery
Target in stables
100%

Model call as of 2026-09-03, on data through 2026-09-02. Written down the day it was made and hash-chained to the one before it — read the ledger.

The sequence

Six phases.
The cycle runs them in order.

Bitcoin's cycle rhymes. The model's job isn't to guess where it ends — it's to tell you which phase you're standing in today, and what that phase asks you to hold. When the evidence changes, the phase changes, and so does the answer.

  1. 01 Accumulate BTC Bitcoin is cheap relative to the cycle. Buy on a schedule and don't try to time the exact low. 90% BTC
  2. 02 Hold BTC The trend is working. Hold what you have and stop adding — the easy accumulation is behind you. 75% BTC
  3. 03 Rotate to ETH Bitcoin's run is maturing and ether typically moves next. Shift some weight across. 30% ETH
  4. 04 Rotate to Alts The short, violent part of the cycle where the broad market outruns bitcoin. Weeks, not months. 40% Alts
  5. 05 Distribute / Exit Sell into strength while there are still buyers. This is an allocation decision, not a top call. 55% → 20% crypto
  6. 06 Wait in Stables Nothing to do. Sit in stables until the model sees a confirmed turn, not just a cheap price. 100% Cash

Distribute is the one phase without a fixed weight: it walks down from 55% to 20% crypto over 30 days, so the exit is staged rather than a single day's decision.

and back to the start — the cycle does not end, it repeats

Percentages are the model's target allocation for each phase, read live from the engine — not written into this page. The legend: BTC ETH Alts Stables.

Today's order

What the model is asking for, right now.

Stay in stables — wait for a confirmed turn, not just deep value.

Target weights for Wait in Stables, the phase the model has us in as of 2026-09-03. It recommends moving toward these, not trading to them in a day.

The evidence behind it

Cycle score (−1 to +1)
−0.022
Confidence
80%
Market state
Early recovery
Conditions
Bull
Data through
2026-09-02
Unrealised profit (of market value)
31.7%
From all-time high
−38.2%
Action
Hold

Every field above is the same record published to the call ledger and hash-chained to yesterday's. Nothing on this page is typed by hand — if the model changes its mind, this changes with it.

The record

Every cycle turn since 2012, and what it said.

The model has a call for 5,047 days — every day back to 2012-10-27, scored using only the data available on each of those days. Here is what it said at each cycle top and bottom.

Read the last row first. The two tops it handles best, 2013 and 2017, were both violent blow-offs, and the rule that catches them only triggers in a market going near-vertical — it has not fired once since 2017. Into the 2025 peak, the most recent turn and the one most like today, it was defensive on just 3 of the 30 days before. That is the number to judge it on.

Days scored
5,047
Cycle turns
7
Live calls published
50
Since
2026-07-10
Model version
76ca61c03cf46519
What the model said at each cycle turn — tested against history, misses included
DateTurnPriceSaid that dayApproach
2013-12-04 2013 Top $1,135 De-risk 5/30 defensive on 17% of the approach
2015-01-14 2015 Bottom $176 Heavy DCA 17/30 buying on 57% of the approach
2017-12-17 2017 Top $19,250 De-risk 30/30 defensive on 100% of the approach
2018-12-15 2018 Bottom $3,185 Heavy DCA 24/30 buying on 80% of the approach
2021-11-10 2021 Top $64,756 De-risk 24/30 defensive on 80% of the approach
2022-11-21 2022 Bottom $15,778 Light DCA 14/30 buying on 47% of the approach
2025-10-06 2025 Peak* $124,824 De-risk 3/30 defensive on 10% of the approach

Approach counts the 30 days leading into each turn, and how many of them the model already had you on the right side — defensive going into a top, buying going into a bottom. 30 out of 30 means it was positioned the whole month before. 0 out of 30 means it missed.

This is the model as it stands today, run back over the past. On every one of those days it could only see what was actually knowable that day — we test for that specifically, because we have twice found and fixed places where it was accidentally peeking at the future. It is a test, not a live result.

And the live ledger, which is young

The last calls, exactly as published. Highlighted rows are days the call changed.
DateCallActionScoreBTCHash
2026-09-03 Wait in Stables Hold −0.022 $77,189 148fedca2b
2026-08-29 Wait in Stables Hold −0.080 $77,717 e7c9012a15
2026-08-28 Wait in Stables Hold −0.016 $80,297 f69c2f1617
2026-08-27 Wait in Stables Hold +0.044 $78,955 aaa7b42ccd
2026-08-26 Wait in Stables Hold −0.006 $78,601 db61588bb4
2026-08-25 Wait in Stables Hold +0.022 $78,894 3a21f1f218
2026-08-24 Wait in Stables Hold −0.010 $77,593 8ed6d37c61
2026-08-23 Wait in Stables Hold −0.042 $77,020 fc7b82cf9a

Underneath, the model changed its mind 12 times in this window. We published 4. A call that flips every other day is noise, so a new one has to hold for three days before we print it — and we publish the unfiltered version too, so you can see exactly what we held back. Each hash covers the date, call and score of that row plus the hash before it, so a changed past entry breaks every link after it. Read the full ledger — it includes the days the model was wrong, and a verifier you can run yourself.

And the part that is a backtest

Every performance figure we publish is a historical simulation, not a live result. Across three completed Bitcoin cycles the model returned 2.03×, 1.08× and 2.23× against buy-and-hold. The current cycle is still open: it reads 1.79× as of 2026-09-03, and an open cycle's multiple moves with the price — treat it as a live reading, not a result. We also test it on 18 separate stretches of history it was never tuned on. It lost less in the crashes in 18/18 of them — a consistency check, not proof of timing skill on its own, since a portfolio with no timing at the model's own market exposure scores the same — but out-earned holding in only 11/18. Over a short stretch it is not reliably better than simply holding bitcoin. Over a whole cycle it has been.

Each cycle has its own page — what the model did, turn by turn, including where it was wrong: 2013–15, 2015–18, 2019–22, and the open 2023–26.

Method

How the call gets made.

No single indicator decides anything. Dozens of signals are scored, grouped into families, weighted by how much each has actually earned, and collapsed into one number. That number picks a phase; the phase picks an allocation. Same chain every day, whether or not we like where it lands.

  1. 01SignalsOn-chain, macro, technical, positioning, sentiment — each scored against its own history, not a fixed threshold.
  2. 02CompositeWeighted into one score between −1 and +1. Weights are regime-aware: what matters in a bull is not what matters in a bear.
  3. 03PhaseThe score, the market state and the cycle's age select one of six phases.
  4. 04AllocationEach phase carries a target mix. That's the answer — a position, not a prediction.

What it can't do