Can Individual Human Financial Behaviour Be Mathematically Modelled? A Case Study of Elon Musk’s Dogecoin Tweets

Juraj Medzihorský, Peter Krištofík

Can Individual Human Financial Behaviour Be Mathematically Modelled? A Case Study of Elon Musk’s Dogecoin Tweets

Číslo: 2/2022
Periodikum: Statistika
DOI: 10.54694/stat.2022.9

Klíčová slova: Behavioral economics, cryptocurrencies, pump-and-dump scheme, linear regression, time series analysis

Pro získání musíte mít účet v Citace PRO.

Přečíst po přihlášení

Anotace: The price of Dogecoin has been influenced by Elon Musk’s tweets on several occasions. Moreover, there are repeating patterns in the Dogecoin prices. However, is there also a pattern to the timing of the tweets? Applying linear regression, we have been able to make the reverse analysis – to use hard financial data (prices) to analyse the human behaviour (tweets) that preceded and influenced the financial data. Selected tweets could be paired thanks to the projections of their timing on the regression line that had been created over the prices. Our model exhibits inaccuracies only in the order of the days. That is surprising, as pump schemes do not usually require such a high level of long-term deterministic timing.