Bitcoin Expectile Model

The Bitcoin Expectile Model maps Bitcoin's long-run trajectory with seven expectile log-log regressions, a squared-error cousin of quantile regression that is cheaper to fit at many levels at once. Low tau values hug historical lows and accumulation territory; high taus track the highs where corrections have started. Inspired by PlanC's Bitcoin Quantile Model, it is built for the Bitcoin all-time-history index on 3-day or weekly log-scale charts, though assets with a long upward bias can be fitted on daily or higher timeframes.

How to Trade the Bitcoin Expectile Model?

• Green zone: the long-term buy and accumulation area defined by the lower expectile fits.
• Purple zone: rare overshoot bottoms, historically anomalous value below even the accumulation fits.
• Red zone: the distribution area where prior cycles saw sustained selling.
• Forward paths: fits extrapolate to your Forecast End date, sketching potential long-term corridors.

Treat it as a slow variable, like any crypto cycle model: context for accumulation and distribution decisions, not a signal generator. On other assets, a linear scale and a later Start Date can repurpose the fits as shorter-term support and resistance mapping.

Bitcoin Expectile Model Settings

• Start Date (default 2010-07-16): where the fit begins, matching common Bitcoin model conventions.
• Correct for Genesis Block: offsets the X-axis for the missing bars before recorded price history.
• Expectiles: enable individual fits and set each tau to sit nearer the highs or the lows.
• Forecast End: the date at which extrapolated projections stop.
• Iterations Number: precision versus speed. More iterations sharpen the fit but compute longer.

Frequently Asked Questions

How do expectiles differ from quantiles?

How does it compare with the Bitcoin Power Law?

Is the Bitcoin Expectile Model free to use?