Asset Class Participation and the Effect of Peer Performance on Allocation

Topic and Objectives

This project consists of two related papers. The first attempts to answer the research question “who invests in cryptocurrencies?” while the second focuses on how various products and asymmetric returns in the portfolios of peers can affect individual portfolio allocations.

While cryptocurrencies have existed for the last decade, these assets have recently begun to gain widespread attention from the media, individual households, financial institutions, and regulatory agencies. Despite growing interest, cryptocurrencies are still under-researched academically. Most notably, information on the characteristics and behaviors of individual investors who chose to buy and trade these assets remains largely undocumented. One reason for the lack of empirical research in this area is the inherently anonymous nature of cryptocurrencies.

We provide an initial examination of who invests in cryptocurrencies by exploiting indirect investments into cryptocurrency-based structured retail products. While relatively new to the marketplace, these assets provide an indirect avenue into cryptocurrency investments. These structured products, like participation or index certificates, allow shareholders to invest indirectly in a particular cryptocurrency via their existing investment account. Our paper considers which types of investors are drawn to these assets and quantify their relative performance.

In a second paper, we use a novel setting to examine peer effects in investment decisions. We leverage administrative retail bank data and measure the peer relationships between customers that recommended the bank to an acquaintance and those who got recommended. Our data allows us to precisely measure the connection between the individuals in our sample and therefore limited assumptions need to be made about the composition of peer groups. Moreover, because we leveraged detailed trade-level data, we can shed light on how peer effects affect the quality of portfolios across several comparable metrics. Previous literature focuses on how peers affect participation in risky asset markets, with much less attention focused on if peer effects result in ‘good’or ‘bad’ financial decision-making.

Key Findings

• We find that cryptocurrency investors are active traders who are prone to investment biases and hold risky portfolios.
• Cryptocurrency investors are more likely to invest in stocks with high media sentiment and more likely to employ heuristics from technical analysis.
• In line with attention effects and anticipatory utility, we find that the average cryptocurrency investor substantially increases account logins and trading activity after his or her first cryptocurrency purchase.
• Furthermore, cryptocurrency investors tend to tilt their portfolios toward even more risky securities after cryptocurrency adoption.    
• Using unique data on peer relationships and portfolio composition, we show considerable overlap in investment portfolios when an investor recommends their brokerage to a peer. We argue that this is strong evidence of peer effects and show that peer effects lead to better portfolio quality.
• Peers become more likely to invest in funds when their recommenders also invest, improving portfolio diversification compared to the average investor and various placebo counterfactuals.

Policy Implications

• Some investors are more likely to adopt new financial products and various investors have different levels of vulnerability to cryptocurrency investments.
• Investors seem to take additional risk in other settings after investing in cryptocurrencies, this does not bode well for the risk-adjusted performance of their portfolios. This is important for regulators to be aware of.
• Social networks can provide good advice in settings where individuals are personally connected.