Giovanni Explains why S2F is a Fake Model Based on the Equivalent of a Magic Trick
Вставка
- Опубліковано 20 вер 2024
- Please listen to the video where I respond to a question on X about S2F. It is really based on very trivial case of correlation is not causation that for some reason has fooled thousands of people including Institutional Directors, Bitcoin podcasters and financial experts. It is incredible.
Nobody ever believed the day to day issuance ever had an impact on price. It was the time average of issuance rate over months that mattered.
It still has no impact because you can see the price goes up smoothly independently from issuance. There is no indication the price is doing anything else than what it did before.
Also, issuance is a drop in comparison with typical transaction volume we experience every day.
The model has no way to show this deficit of coins if there was one. It doesn't represent at all what is supposed to represent.
Look at the shape of the mode. It is flat (that means no correlation) and then a sudden jump at the precise timing of the single halving block (that we know didn't change the price either).
Where is the signal in the data that shows a delayed event that is due to this deficit? It is just wishful thinking.
Also the point is that if we can recreate similar looking models, with the stairs, the jumps and so on then it means the model PlanB create are in all and all equivalent to this fake models we created using noise. How can we distinguish between the two?
Simply what the model does is to mark where the halving is. But I do that int the power law model with a star or a colored marker or whatever. You don't need a model for that.
What the model is try to do is to say the price goes up because of this marker. It doesn't. It goes up before the halving, after the halving and so on. And it is clearly independent of issuance at any time.
Pretend there are only 4 data points: time averaged price over each bitcoin epoch for y points and s2f at each epoch for x axis. That’s all s2f is really doing.
First of all that would be a much honest way to present the model. The problem with that is with 4 data points you would not get statistical significance. Also there will not be a way to associate issuance with price simply because price grows before and after.
How S2F explains the price goes up before the halving? How S2F explains why we have a bear market?
Why at the moment when there is the highest demand, so the higher scarcity of coins the market collapses on itself?
Why we had a bubble before the halving?
How a quantity that is 100 times smaller than the average transaction volume can affect the price?Everything together makes it a really nonsensical model.
Because s2f is inherently exponential, everyone knows it must fail eventually. But because bitcoin price goes up over time and s2f goes up over time, s2f can be made to reasonably fit over some time interval. But it can’t fit well forever.
The real question is whether power law model will fit longer than s2f. Because power law is not exponential, it has the potential to fit for longer. But of course neither _has_ to fit tomorrow or a decade from now.
The main power of the power law is the scale invariance property. Scale invariant systems tend to continue to be scale invariant as the name says. Also we did so much work showing the real causative chain (adoption and Metcalfe Law using network differential equations and simulations, we did many scale invariance tests showing that actually the power law is a deep statistical property of Bitcoin and so on). It is a full consistent theory and even a philosophical understanding of Bitcoin. Also we have 6 years out of sample data. Of course, we will have to see if it continues to hold in the future.
For the laymen among us, it would be helpful if you could explain how you *would* see a correlation! In other words, instead of showing a graph that demonstrates there is no correlation, show a graph where there *is* a correlation between scarcity and price, and then compare it to ours.
Ok, I already did it on X but I can do a video about that too.