Behavioral Biases of Cryptocurrency Investors

Abstract

Cryptocurrencies are deemed to be highly influenced and driven by investors' sentiments flowing across social media platforms. Consequently, researchers are attracted to investigate investors' behavioral biases in investing in cryptocurrencies. The existing related research majorly focuses on the investigation with the implementation of questionnaires and surveys. However, to what extent the feedbacks to these questionnaires or surveys truthfully reflect the investors' actual practices in investing in cryptocurrencies is uncertain and dubious. Therefore, in this study, we inspect and appraise the behavioral biases and portfolio properties of cryptocurrency investors by utilizing the on-blockchain (on-chain) information of wallet records directly from the Ethereum network. By retrieving and analyzing the unique wallet addresses and related transactions, we have obtained three behavioral bias proxies of the investors behind the wallets and five different properties of the wallets. Furthermore, we distinguish and analyze the wallets of human investors and trading bots. The results of statistical tests indicate the significant differences between human investors and trading bots on most behavioral biases and wallet properties.