City Research Online

Behavioural Finance and Cryptocurrencies: Studies of Behavioural Finance in Cryptocurrency Markets

Marinoff, E. A. (2024). Behavioural Finance and Cryptocurrencies: Studies of Behavioural Finance in Cryptocurrency Markets. (Unpublished Doctoral thesis, City, University of London)


Interest in cryptocurrencies among researchers has been on the rise as this novel asset class continues to attract significant fund flows from retail and as well as institutional investors due to the promising returns it has offered historically. While the behavioural biases and performance of traders have been studied extensively in traditional financial markets, the literature on these topics within the crypto space remains lacking.

This thesis aims to provide insight into the behavioural characteristics of cryptocurrency traders by investigating two behavioural biases that have been widely explored in traditional financial markets, namely, the disposition effect and the gambler’s fallacy. Moreover, I investigate the impact of market sentiment on trader performance and activity in the crypto space using an alternative on-chain measure of sentiment.

Regarding the study on the disposition effect, I apply the popular disposition
spread metric of Odean (1998) on a unique data set of individual cryptocurrency traders from an anonymous exchange and find significant evidence of an antidisposition effect. In analysing the disposition effect across market conditions, trader experience, and age groups in the cryptocurrency market, the study finds that neither market trends nor average trade size significantly alter traders’ biases in realising gains or losses. Younger traders, especially those aged 18-30, exhibit a unique positive disposition effect, suggesting quicker gains realisation. Conversely, older traders display a reduced anti-disposition effect, indicating that the tendency to hold losing investments decreases with age. The study aligns with existing
literature in suggesting that experience mitigates behavioural biases, evidenced by the consistent patterns observed in both the cryptocurrency and traditional financial markets.

In the second study, I investigate whether traders exhibit trend-chasing behaviour by examining the relation between traders’ past performance and their future trade size. Specifically, those who exhibit the gambler’s fallacy are likely to increase their trade size after experiencing poor past performance as they double down on future investments to make up for poor past performance. Alternatively, those who exhibit the hot-hand fallacy are likely to increase their trade after experiencing positive past performance as they believe that good performance will persist into the future. My results show that crypto traders exhibit the gambler’s fallacy, such that they are likely to increase their position size after exhibiting poor past performance, suggesting that they expect a trend reversal.

In the final study, I investigate the impact of market sentiment on trader performance and activity. While the literature has mainly focused on text-based models to gauge market sentiment, I employ an alternative on-chain metric called the Net Unrealised Profit Loss (NUPL), which is a measure of accounting of the overall state of profitability of a blockchain network. A positive (negative) NUPL suggests that the blockchain network is in a state of profit (loss) and thus nearing a market top (bottom). Hence, this metric offers insight into the general degree of market sentiment based on fundamental on-chain data. My findings show that changes in sentiment positively impact the total return experienced by traders. Moreover, traders experience the highest levels of total returns when market sentiment is very high. Second, traders who react immediately to market sentiment, especially when sentiment is very high, are likely to realise higher positive returns. Third, higher levels of market sentiment lead to larger future trade sizes; hence, traders increase their exposure when market sentiment is high. Finally, I report weak evidence supporting the notion that higher levels of market sentiment result in traders modifying their trade size. This suggests that a change in trade size is agnostic to market sentiment. For robustness, I also adopt the VIX, a common equity market volatility index, to measure sentiment in the cryptocurrency market; however, the results showed no consistent impact, highlighting the need for developing a sentiment measure specifically designed for the unique characteristics of the cryptocurrency market.

The concluding chapter reviews the main findings of this thesis and discusses avenues for future research.

Publication Type: Thesis (Doctoral)
Subjects: H Social Sciences > HG Finance
Departments: Bayes Business School > Bayes Business School Doctoral Theses
Bayes Business School > Finance
Doctoral Theses
[thumbnail of Marinoff thesis 2024.pdf]
Text - Accepted Version
Download (2MB) | Preview


Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email


Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login