Global M2 cannot predict the price of Bitcoin, explains the analyst

SINA – Co -founder of the Hedge Fund 21st Capital – published by the demonstration of a popular Bitcoin Prize model promoted by the CEO of Real Vision, Raoul Pal, calling a case of manual of illiteracy and over -adjustment.

The model in question establishes a close correlation between bitcoin and Global M2—A measurement of the global money supply – by moving the M2 data by a number of weeks, generally from 10 to 12, to “predict” supposedly the movements of future Bitcoin price. Raoul PAL used this graph to assert that the liquidity macro conditions lead the cryptographic cycles and that the current market behavior can be planned using a monetary expansion.

Expert correlation of M2-Bitcoin torches

But Sina, a scientist of the formed data who teaches the analysis of the data at the first cycle and the higher cycles, said This model collapses under control. “This is a terrible failure not to understand over-adjustment,” he said in a video of June 24 published on X. “What I see does not even pass the first month of a first-year data analysis course.”

Related reading

Sina stresses that the apparent correlation between bitcoin and global M2 exists only because the data has been “tortured” to adapt to historical models. “If I am authorized to play with the data and to arbitrarily move things, I can certainly find large correspondences between the data pockets,” he said, warning that this flexibility is exactly what allows analysts to create the illusion of a predictive precision.

The main problem, he explained, is that global data M2 itself is intrinsically defective. It is compiled by multiplying various M2 figures of central banks by exchange rates – mixing rapid report savings like the United States with countries that have delays in weeks or even months. This creates a misleading impression of daily fluctuations in global liquidity. “It seems to evolve daily, but it actually mixes frequent and infrequent updates,” said Sina. “This is not a real signal.”

More importantly, Sina maintains that the model fails when you zoom in zooming from selective diagram slices. While Raoul Pal And others have presented examples of peaks and stockings closely aligned between Bitcoin and Global M2, Sina has demonstrated how minor adjustments in delivery time or the scale can produce considerably different results. “Let’s try an 80-day lead. It doesn’t look good. What about 108? Ah, now the summits align-so zoom in again and do as if it works,” he said sarcastically. “It is not modeling. It’s a game. “

Related reading

He underlined how each model adjustment – passing by an advance from 12 weeks to 10 weeks, to 108 days – exposes his lack of systematic foundation. “If you don’t have an appropriate model, you fail to predict the future,” said Sina. “It is a classic over-adjustment. You force the data to correspond to historical behavior, but you lose all generalization.”

To illustrate the concept, Sina compared it to the adaptation of a curve through a noisy sinusoidal wave. A well -structured model captures the central model and ignores noise. An overput model, on the other hand, tries to match each small fluctuation – resulting in poor predictive performance when new data arrive. “Surre-up seems better, but it models noise. And noise is not repeated,” he said.

Sina also wondered if Bitcoin could really carry out liquidity, not follow him. “If you look at the last cycle, Bitcoin exceeded the first. Liquidity exceeded 145 days later,” he said. This reverses the causality involved by the M2 world model and calls into question its premise as a prospective tool.

His conclusion was frank: “You must be very cautious with the over-adjustment. It seems to be matching, but it is forcefully on historical data. You have no idea of ​​the predictive precision of this thing.”

At the time of the press, Bitcoin exchanged $ 106,952.

Star image created with dall.e, tradingView.com graphic