Statistically obvious? Traders and analysts sometimes use – often ? – statistics on cryptocurrency exchanges to try to explain a rise or fall in the price of a cryptocurrency. Sometimes even to predict its trajectory over a given period. Do the hypotheses that might seem logical describe the reality of the markets?
>> Play it safe, register on the reference of FTX crypto exchanges (affiliate link) <<
Fewer visitors on exchanges: crypto investors necessarily in the accumulation phase?
According to data from The Block Research, web traffic to cryptocurrency exchanges has recorded a decrease from the previous month for the third consecutive month in January 2022. This traffic reached 427.6 million visitors, a figure down by 8.3% compared to the traffic of December 2021, which itself is 14.7% lower than that of November 2021.
How can we interpret this drop? Besides, is it possible to draw another conclusion ? A fall in the balance of an asset on the exchanges is often interpreted as a sign of accumulation of this asset, their holders would keep their assets in cold portfolios without therefore thinking of parting with them. This accumulation should theoretically be followed by a rise in the price of the asset thereafter.
Is a drop in traffic on cryptocurrency exchanges also accompanied by a drop in crypto balances on these exchanges? Thus indicating a phase of accumulation theoretically heralding a rise.
These claims are actually theoretically attractive but, only a statistical study of the possible relationships between the relevant variables would allow to validate them or notonce confronted that they are confronted with empirical data.
Relationship between exchange web traffic variation and bitcoin price variation
We also had “fun” to study possible relationships between falls or rises in web traffic on crypto exchanges compared to previous periods since January 2018 based on data provided by The Block Research, AND changes in the price of bitcoin (BTC) based on data provided by Bitfinex.
Why did we choose bitcoin price as the second variable? Bitcoin remains the dominant cryptocurrency in the cryptosphere. “we” could therefore to be tempted to correlate its course with exchange traffic.
We studied the potential relations Ri (i varying from 1 to 4) between the following variables (in capital letters):
- R1: INCREASE or DECREASE in exchange traffic for month M compared to the previous month and monthly variation (closing in RED or GREEN) in the price of bitcoin during this month M;
- R2: INCREASE or DECREASE in exchange traffic during quarter T compared to the previous quarter and quarterly variation (closing in RED or GREEN) in the price of bitcoin during this quarter T;
- R3: INCREASE or DECREASE in exchange traffic for month M compared to the previous month and monthly variation (closing in RED or GREEN) in the price of bitcoin during month M+1: in other words, a decrease in traffic during month M compared to the previous month (theoretical accumulation), will it be followed by an increase the following month?
- R4: INCREASE or DECREASE in exchange traffic in quarter T compared to the previous quarter and monthly variation (closing in RED or GREEN) in the price of bitcoin during quarter T+1: in other words, a drop in the traffic during quarter T compared to the previous quarter (theoretical accumulation), will it be followed by an increase the following quarter?
The analysis was therefore carried out over two periods, monthly and quarterly, in order to determine whether or not the temporal difference will have an impact on the possible relationships between the variables studied.
Given the time periods chosen, the choice of data source for the bitcoin price – exchange selected – is almost irrelevant. Historical prices were mostly needed only to know if bitcoin closed the period in the red or in the green – regardless of the magnitude of the variation.
After having reconstituted the table of frequencies observed between the variables INCREASE, DECREASE on the one hand, and RED, GREEN on the other hand, we calculated the value of Chi-square, an indicator which makes it possible to appreciate the existence or not of a statistically significant relationship between two qualitative variables. The value noted r for each relation Ri tested amounts to:
- R1: r = 14.57%
- R2: r = 13.80%
- R3: r = 82.73%
- R4: r = 10.69%
The value of the Chi-squares being all well above a significance level of 5% – these values are even greater than 10% except for that of R4 which is close to it – a relationship between variables is therefore not statistically significant.for each of the four hypotheses formulated above.
A few essential points (or reminders) before concluding:
- This study did not investigate the relationship between exchange bitcoin reserves – but between the variation of the web traffic of these exchanges – and the change in the bitcoin price over the chosen periods;
- Is the change in web traffic on cryptocurrency exchanges correlated to the price of a major altcoin like ether (ETH) or to the change in market capitalization over certain time periods: maybe, but that would require d other statistical analyzes to be able to decide.
These statistical studies will not have revealed much in the end, the crypto markets not being limited to bitcoin, the results were therefore obvious and predictable? Perhaps, but the evidence would therefore lead us to say that the earth is very flat. As for bitcoin, the only certainty at the time of writing this article would be that it currently makes the markets doubt about a sustainable bullish recovery.
To take your first steps in bitcoin and cryptocurrencies, start by making sure you evolve in a secure environment, and which will allow you to organize a solid investment strategy: Register without delay on the FTX reference crypto exchange platform and benefit from a lifetime discount on your trading fees (affiliate link, see conditions on official website).