Professor Charlie Cai’s research confirms the unsophisticated and risk seeking behaviour of specialised crypto investors, often driven by the fear of missing out (FOMO) rather than objective return valuations.
In a joint study titled 'Salience theory and cryptocurrency returns' and co-authored with Fowler College of Business’ colleague, Dr Ran Zhao, Charlie introduces a new salient effect measure that offers a close description of the current price dynamics in the crypto market.
They document how crypto investors tend to follow the crowd, putting money into cryptocurrencies which stand out due to price peaks that hit the headlines and trigger social media buzz.
This salience effect driven by the fear of missing out on cryptos with significant upward returns, causes overvaluation that results in profit loss.
The analysis reveals this investment strategy yields substantially lower returns over the following month than investing in cryptocurrencies with salient price drops.
The investigation also confirms the salience effect is more pronounced in the crypto market than in equity markets, making it a significant risk factor for this developing asset class.
Price volatility in the crypto market: a story of peaks and troughs
Offering a groundbreaking blend of financial freedom, growth potential and technological intrigue, the crypto market has attracted a diverse range of individuals and institutions, reaching £2 trillion in value in just over 15 years.
This period has also seen an unprecedented surge in popularity among novice crypto-curious investors with little experience in dealing with extreme price volatility.
A glance through historical price charts in the crypto market shows a rollercoaster scenario with extreme spikes and slumps which occur quicker and more often than in mainstream markets.
Extremely sensitive to signs of mainstream acceptance and government bans, recurrent fraud scandals and even social media posts, the value of cryptos is difficult to gauge due to high uncertainty and limited fundamental information.
Most cryptocurrencies don’t have a clear, tangible use, or their use is less obvious compared to companies in the stock market.
Unlike fiat money, cryptos lack economic fundamentals, such as being backed by governments, centralised, legally recognised and widely accepted.
Compared to traditional financial assets, they don’t generate cash flows, and unlike precious metals, crypto coins don’t have a long history of trust or cultural preferences as a means of storing value.
Despite the growing number of cryptos in the market, they are one of the most speculative assets to invest in, which gives rise to the question: How do investors decide which coins to invest in or whether to invest at all?
How concentrating on price spikes translates into profit loss
Investment decisions are not necessarily driven by rational risk-return trade-off, with evidence suggesting subjective estimations often prevail over objective valuations in the crypto market.
For example, investors may try to maximise future returns by favouring assets with salient payoffs that capture their attention, against average alternatives.
This behaviour closely matches that of the crypto market, which due to its emerging and non-mainstream nature creates an environment more likely to attract those with a salience bias.
Using market value data of over 4,000 coins1 from coinmarketcap.com from January 2014 to June 2021, Charlie and Ran’s study explores the link between salient thinking and crypto pricing.
To do so, they first constructed a salience measure (ST) that quantifies how much salience bias distorts expectations of future returns compared to objectively realised past returns.
The analysis revealed coins with salient upsides yield lower returns over the following month than those with salient downsides.
Specifically, the average return for an investment strategy that involves buying high and selling low ST cryptocurrencies results in:
- 26% monthly loss in profit for an equally weighted portfolio, in which the same percentage of funds is allocated to each coin regardless of its size or market capitalisation
- 32% monthly loss in profit for a value weighted portfolio that favours larger cryptos with a higher percentage of funds, reflecting their greater market value.
Conversely, purchasing downward salience and selling upward salience cryptos, would be a lucrative strategy that yields the opposite (positive) returns.
These figures are over 20 times greater than those reported in the US equity market and considerably larger than the microcap stocks2 results.
This finding implies that there is a disproportionately large group of salience-driven investors who struggle to fully understand what crypto assets represent, underscoring the importance of regulations and investor protection.
Risk-seeking behaviour dominates the crypto market
When compared to other investment opportunities, the results show a strong salience effect is not observable in traditional markets, where fundamental information is essential for investors’ decisions.
In other words, allocating capital among assets in the global financial market is more efficient and less influenced by salience bias.
Zooming closer into the behaviour of crypto investors, the analysis indicates crypto investors are more active during less volatile periods in the stock market and the economy, and when attention to the crypto market is high.
This means investors influenced by salience bias are likely to be risk-seekers, drawn to cryptocurrencies when other asset markets are calm and attention-grabbing cryptocurrency fluctuations occur.
So, the next question is: does this type of behaviour influence the price of crypto coins?
The salience effect as a mispricing phenomenon
Anomalies occur when an asset or a group of assets perform contrary to the notion of efficient markets. This can be caused by, for example, behavioural biases.
If the salience effect captures one of the key trading behavioural biases of crypto investors, it may be behind other return anomalies in this emerging asset class.
If so, this would mean the salience effect is an important risk factor, influencing mispricing.
To investigate this, the authors tested to what extent the salience effect was able to explain several pricing anomalies in the crypto market 3.
The analysis indicates the salience effect is on par with the strongest risk factors documented so far, and therefore can be considered a distinct measure or crypto risk.
This means, the salience effect can not only predict future returns, but most importantly, it influences prices and expected returns.
Although the salience effect is more prominent in the crypto market, its impact on prices varies depending on investor sophistication.
The largest cryptos are more likely to attract institutional investors’ attention given their size and liquidity.
With more experienced investors putting money into big cryptos, the salience bias effect on pricing reduces.
This finding suggests that, as cryptocurrencies develop and become more regulated and integrated with mainstream finance, sophisticated investors will reduce the influence of salient thinking on this market.
Mispricing is more relevant to micro cryptos. While they account for just 3% of the market capitalisation, these assets represent approximately 87% of the number of cryptos.
This result is still economically important, as the market efficiency of the small but numerous cryptos will be essential for cryptocurrencies to consolidate as an asset class.
The study confirms salience thinking is more relevant for emerging assets with high uncertainties, but it’s possible that once the market becomes more mature other pricing mechanisms may dominate.
Before then, Charlie and Ran’s study offers a solid explanation of crypto market return dynamics.
1 The sample included crypto coins with a market value greater than $1 million (£760K) and didn’t contain stablecoins - cryptocurrencies with value pegged to another asset, eg fiat currency or gold.
2 Microcap stocks are small publicly traded companies with a market capitalisation between $50-$300 million (£38-230 million).
3 This includes significant and insignificant anomalies identified in Liu et al. (2022), plus new anomalies relevant to the prospect theory and skewness.
 |
Professor of Finance |