FRAKTAL·ICIOUS
Historical simulation

The 2013–2015 cycle

Bitcoin's first two-sided cycle on this site's data: a blow-off to $1,135 on 4 December 2013, a crash into the Mt. Gox collapse, and a grinding bear into early 2015. The model's round trips through this cycle were mixed — one cost BTC, one gained it.

Max drawdown, model vs. buy-and-hold
-71.1% vs -84.5%
Peak-to-trough decline over the Jan 1, 2013–Oct 1, 2015 window. Shallower is better.
Return vs. buy-and-hold
2.03×
A multiple OF buy-and-hold's return over the same window, not of starting capital — beat buy-and-hold.

Round trips this cycle

Each row is one exit signal, the low the model sat out through, and the re-entry signal that followed. "BTC change" is the exit price divided by the re-entry price: above 1.00× means the round trip bought back cheaper than it sold (a gain in BTC terms); below 1.00× means it bought back higher (a cost), even in cycles the model won overall.

# Exit signal Low reached Re-entry signal Days out BTC change
1 Nov 9, 2013
$331
Nov 10, 2013
$328 (-1.1%)
Apr 10, 2014
$367
152 0.90× ▼ cost BTC
2 Aug 10, 2014
$589
Aug 18, 2014
$469 (-20.4%)
Aug 31, 2014
$479
21 1.23× ▲ gained BTC
Dates and prices are the model's signal dates, not intraday extremes. "Low reached" is the lowest close during the time the model was out, which is not always the same day the re-entry signal fires.

In this cycle

The model exited this cycle far too EARLY, not too late: the first exit fired on 9 November 2013 at $331, and price then ran another 3.4x to its 4 December peak of $1,135. That is why the first round trip above shows a loss in BTC terms — buying back cost more than simply holding would have. It is the clearest instance of a failure the model still has: it cannot reliably tell a parabola that keeps running from a top that is about to break, and it sells into both. The second round trip, later in the same grinding decline, won some of that back.

The cycle's max drawdown (above) reflects Bitcoin's roughest early stretch, before on-chain data existed to inform the model at all — most of this cycle's signal came from price and momentum alone.

Other cycles

2013–2015 · this page
2.03×
2015–2018
1.08×
2019–2022
2.23×
2023–2026
1.79× (open)

Figures are the model's return as a multiple of buy-and-hold over the same window — not a multiple of starting capital. 1.00× is a tie.

Fraktalicious Research · Call ledger · Privacy
All figures on this page come from a historical SIMULATION of the model's rules against past price and on-chain data, not a live trading track record. Nothing on this site is financial advice, personalised, or a price prediction.