Milton Friedman was spot on

Quantile Model Major Breakthrough Embedding Dynamic Decay Without Guesswork I have discovered a way to dynamically embed decay into the quantile model without manually fitting or guessing the decay function. The Problem: Static Quantile Fits Miss Residual Evolution Traditional quantile regression assumes that the relationship between variables remains static over time. This leads to a critical issue—residuals often show systematic drift, which means the quantile fit must also evolve. Instead of manually selecting a decay factor like exponential decay, I found that the residual slope itself provides a natural mechanism for dynamically adjusting the quantile fit over time. By embedding decay directly into the quantile model rather than treating it as an external adjustment, we allow the model to evolve naturally over time. Why This Matters: Smarter Forecasting Without Overfitting This breakthrough addresses a major challenge in long-term projections. Traditional models either ignore decay or rely on manual adjustments, making them unreliable. This approach eliminates guesswork by letting the data define its own decay. The result is a self-updating quantile regression that adapts over time instead of remaining static. In Simple Terms I embed the decay rate directly into each quantile regression band, aligning it with the historical evolution of decay for that band. This eliminates the need for hybrid models or arbitrary decay function assumptions, leading to more accurate and adaptive forecasting.

JUST IN: $5 billion asset manager Bitwise says governments and companies who want to buy bitcoin will have to buy it from individuals willing to sell it 👀

I have to try this 🤔