ABYSS Risk: Bitcoin's cycle risk (0–1)
ABYSS Risk sums up, in a single 0-to-1 number, how expensive or cheap Bitcoin is within its cycle. It starts from how far price sits from its 2-year moving average and scales that by elapsed time to offset each cycle's diminishing returns; the result is mapped to a 0–1 range. It's the idea of a 0–1 risk metric, a format popularized in crypto cycle analysis, but computed by us with the formula and constants published. Near 0 = historic floor (accumulation zone); near 1 = top (maximum risk).
Formula: 0–1 min/max-normalized: ln(price ÷ 2-year MA) × (days since 2012-08-16)^0.395.
Full history 2011→today, weekly points, computed by us from price.
Method & caveat: Reading the history: the 0–1 scale is calibrated on the min and max of the whole 2010–2026 series, so a past point is drawn using information from after that date, and the extremes —the day of the max (near the 2017 top) and of the min (2022 bottom)— define the 1 and the 0 by construction. The table further down samples the first weekly point of each month, so there you'll see near-extreme values (0.94 · 0.09), not exactly 1 and 0. It's a fit over the full history, not a signal you'd have seen identically in real time, and it isn't stationary (for the same valuation, risk tends to drift up over the years). Constants: exponent 0.395; min −22.75; max 45.51; reference 2012-08-16.
How to read it
Below 0.2 risk is very low (the 2022 bottom fell to ~0.1 and the 2018-19 one was around ~0.2; 2015 sat a bit higher, ~0.3); 0.4 to 0.6 is the mid phase; above 0.6 risk is high. In hindsight the 2013, 2017 and April 2021 tops all sat at or above 0.8; the November 2021 all-time high came in softer (below 0.7), a reminder of the cycle's diminishing returns.
| Range | Zone |
|---|---|
| < 0.2 | very low risk (floor) |
| 0.2 – 0.4 | low risk (accumulation) |
| 0.4 – 0.6 | moderate risk |
| 0.6 – 0.8 | high risk |
| ≥ 0.8 | extreme risk (top) |
At past tops and bottoms
Where the calibrated 0–1 scale places each major turning point. Note that the 2017 top and 2022 bottom define the scale's ends by construction (see the caveat above); the other rows are out-of-anchor.
| Turning point | ABYSS Risk |
|---|---|
| Top · Dec 2013 | 0.80 |
| Top · Dec 2017 | 0.94 |
| Top · Nov 2021 | 0.66 |
| Bottom · Jan 2015 | 0.31 |
| Bottom · Dec 2018 | 0.20 |
| Bottom · Nov 2022 | 0.09 |
See all indicators side by side in the cycle turning points table.
Does it work? Point-in-time backtest
The honest test: at each day the risk is computed using only data available up to that day (no look-ahead), then we measure the return one year later — versus the unconditional base rate. Over 2013→2025.
| Bucket | Median 1y | Up 1y later |
|---|---|---|
| Base rate (any day) | +66% | 69% |
| Low risk (<0.4) | +89% | 81% |
| High risk (≥0.6) | +8% | 53% |
Point-in-time (no look-ahead), low risk preceded returns above the base rate (81% higher a year later vs 69%), and high risk badly underperformed it (53%). The pattern holds without look-ahead.
Median 1-year return, overlapping daily windows over ~3 cycles (so few independent episodes — treat the exact figures as indicative). Buckets use the causal scale (min/max from past data only), which differs slightly from the fixed 0–1 shown above. And since the signal is essentially price's distance to its 2-year moving average, the forward test inherits mean-reversion; a secular bull market also lifts the base rate — read the DIFFERENCE, not the absolute. Historical context, not a promise — past performance doesn't guarantee future results.
See the full dashboard → Cycle overviewMethodology
Computed by ABYSS Index from our own Bitcoin node and reference price — freely redistributable. For educational purposes only. Not financial advice. Español