SHLL Model BTC Prediction

9 godzin temu

This indicator provides the model predictions for the Seasonal Halving Log-Log (SHLL) model as described here.

As an overview, the SHLL model is a power law model for Bitcoin price with a couple notable differences from the standard power law model:

1. Its exogenous variable is progression through halvings rather than days from the genesis block. Progression through halvings is defined as number of blocks mined through a given day divided by 210,000 (the number of blocks in each cycle)
   
2. Future predictions are based on the current rate at which blocks are mined
   
3. It is a mixed effects model where the fixed effects are a power law using progression through halving (also called "fair value") and the random effects are a seasonality component
   
4. The seasonality component is created by taking the percent progress through halving and breaking it into 48 periods. These allow cycle variability to be calculated versus the fair value
   
5. Variance is also hierarchical; a base variance is calculated and then individual, added variance is calculated for each seasonal period
   
6. It also incorporates IRLS like elements to trim outliers for more robust predictions
   

R^2 of the model versus the standard power law model:

• Log scale: 0.976 vs. 0.956
  
• Linear scale: 0.857 vs. 0.745
  

8 godzin temu

Informacje o Wersji

removing unneeded half range input and lines