drgo on Nostr: lol…no. Flat out no. All models are wrong. A wide variety of models could be used ...
lol…no. Flat out no.
All models are wrong. A wide variety of models could be used and fit bitcoin price over a period of time. That’s math. But what type of model makes sense? It’s true S2F and power law models fit historical price data well; both fit far better than any straight line.
But inherently exponential models can’t both fit for a long time and have a high slope today. Power law models can because the rate of change of their slopes is constant. Exponential models have an ever increasing rate of change of slope. One can’t expect eternal acceleration, especially if slope is already high.
For bitcoin to grow in dollar price as an exponential and for an exponential to fit the price for a long time, bitcoin price would have to be growing far more _slowly_ than power law model — otherwise a predicted price of infinite dollars happens way way way sooner than you’d expect.
All models are wrong. A wide variety of models could be used and fit bitcoin price over a period of time. That’s math. But what type of model makes sense? It’s true S2F and power law models fit historical price data well; both fit far better than any straight line.
But inherently exponential models can’t both fit for a long time and have a high slope today. Power law models can because the rate of change of their slopes is constant. Exponential models have an ever increasing rate of change of slope. One can’t expect eternal acceleration, especially if slope is already high.
For bitcoin to grow in dollar price as an exponential and for an exponential to fit the price for a long time, bitcoin price would have to be growing far more _slowly_ than power law model — otherwise a predicted price of infinite dollars happens way way way sooner than you’d expect.