We can't predict the price. So here's an honest band instead.
Every Bitcoin forecast you've seen is a single number — a target. A target is the one thing the data cannot honestly give you. So we draw the opposite: a band. From today, near $62,000, we project a floor, a middle, and a ceiling 12 to 24 months forward — built so that about 8 of every 10 outcomes should land inside it.
And it leans up. Because Bitcoin sits below its long-run trend line, the middle of the band drifts higher from here. The band stretches from a deep-bear floor to a new-high ceiling — both tails are real, and we show both. The honest work, in the parts that follow, is proving the band actually holds.
price_cone
When we say 80%, do we mean it?
A band is only honest if it catches what it claims to. So we tested it the hard way. Stand at each month in Bitcoin's past, draw the band using only what was known then, jump forward 12 or 24 months, and check whether the real price actually landed inside it. No peeking at the future — ever.
We promised an 80% band. Out of sample, across all of history, it delivered about 86% a year out and 81% two years out — a touch wide, which is the honest direction to miss.
coverage_curve
calibration
Better than a coin-flip — and far tighter than “anything can happen.”
A calibrated band is only worth drawing if it's tighter than the lazy alternative — the "anything can happen" band you'd get by assuming Bitcoin's wild past swings just repeat. So we compared the two head-to-head, on the same promise to catch 8 in 10.
On both counts the band earns its place. Its middle line beats a coin-flip "the price stays put" guess by a wide margin, and its band is dramatically tighter than the everything-is-possible version — about four times tighter two years out.
band_width
skill_bars
The four-year cycle doesn't actually help the forecast.
Here is the part most Bitcoin forecasts won't tell you. We built the famous four-year halving cycle into this model — the recurring "boom a year after the halving, bust two years on" shape. Then we measured, out of sample, whether adding it made the forecast any better than a plain trend line alone.
It didn't. The cycle tilt buys essentially no extra skill — and this round it slightly hurts. The model noticed: left to choose its own cycle weight month by month, it has dialed the cycle down from a heavy 0.6 to a near-ignore 0.25.
cycle_skill
An honest band you can trust — as long as you trust its width, not its middle.
This is not advice to buy or sell, and it is not a price target. It's an honest band around Bitcoin's next one-to-two years — calibrated, balanced on both tails, and tighter than "anything can happen" — together with the honest admission of exactly how much of it survives a stress test, and how much doesn't.
From ~$62K today, a 1-year band of about $25K–$213K (middle ~$77K) and a 2-year band of ~$40K–$362K (middle ~$130K). It leans up — Bitcoin sits below its long-run trend — and spans a deep-bear floor to a new-high ceiling. A calibrated range, never a target.
We promised the band would catch about 8 in 10; out of sample, across all of history, it caught ~86% a year out and ~81% two years out. Honest, leaning slightly wide — the band's width is the part you can trust.
Anchoring to a sixteen-year trend line makes the band about 4× tighter at two years than an 'anything can happen' band built from Bitcoin's past swings. That tightening — not the exact middle — is the real product.
The middle's edge comes from the long-run TREND, not the four-year cycle. Adding the cycle tilt buys ~zero out-of-sample skill, and the model has dialed its cycle weight down from 0.6 to 0.25 as this cycle flattened. Trust the width, not the middle.
- —The band's width is calibrated; the middle line is only weakly informative — it beats a coin-flip but not a plain trend line. Read the floor and the spread, not a target.
- —The four-year-cycle tilt adds essentially no out-of-sample skill and slightly hurts this cycle. We keep it only at a heavily-shrunk weight, and the model is still dialing it down.
- —Every skill number rests on a handful of events: roughly ten independent one-year windows and only about four independent two-year windows, spanning just three completed cycles.
- —The whole picture assumes the long-run trend line keeps holding. A line that has held for sixteen years carries no guarantee for the next two; regular patterns break.
- —Descriptive research about uncertainty, not a price prediction. Bitcoin can lose most of its value in a year, and nothing here is financial advice.
Descriptive research, not financial advice. Bitcoin is volatile and can lose most of its value in a year — as it has four times in this dataset. A trend line that has held for sixteen years is an assumption about the next two, not a fact.
Appendix — all 6 charts & the method →