Cryptocurrencies, though a reasonably recent phenomena, are starting to attract the attention to academics. A recent paper from researchers at the University of Vaasa in Finland looks at the chances of a cryptocurrency defaulting. This is important because over 2,000 cryptocurrencies have launched. However, only a small portion of those have succeeded. In fact, last year alone 743 coins defaulted. This high level of defaults harms the credibility of the industry and is costly for investors.
Default levels do not inspire confidence. Of all cryptocurrencies launched before 2014, almost 6 out of 10 were in default by the end of 2018. Or looking at cryptocurrencies between 2009 and 2019 the default rate was higher still at 8 out of 10. Of course, the results for cryptocurrency investors, in aggregate, are far better than these figures suggest this because Bitcoin alone represents half of the market cap value of all cryptocurrencies and has performed relatively well. Nonetheless, can academics able to sift the wheat from the chaff when it comes to these digital assets? Though the researchers caution that this is an early model, the results are encouraging at predicting defaults. So, what are the main indicators of cryptocurrency default?
The Importance Of A Strong Day One
Day one performance of a cryptocurrency had a significant impact on the chances of default over the following 4 years. The coins that lasted had far stronger day one performance and associated higher volatility over that first day of trading. However, over the first week, returns and volatility then mattered less for long-term default rates. Interestingly, by the end of the first month, the impact reversed. It was the coins more likely to default showing slightly higher returns and elevated volatility, whereas the coins that lasted, actually had slightly more stable, and in fact lower, returns over the first month. So a coin with a big day one pop and then more stability over the first month may be less likely to default over the coming years. It is interesting how this one day of data, can provide a reasonably informative signal for the next four years of the cryptocurrency’s life.
High Pre-Mining As A Negative Signal
Pre-mining occurs where developer pre-mine coins, creating a larger share of the currency for themselves. This pre-mining can be a negative signal, with those coins headed for default having an average of 1.5% of the total coins pre-mined, whereas the level was under 0.5% for the default group. Therefore, though the presence of pre-mining isn’t necessarily a negative signal, a high level of pre-mining may be. The authors suggest that high levels of pre-mining suggest the potential for a get rich quick scheme on the part of the developers, rather than looking to set up the coin for long-term success.
Following Satoshi Nakamoto of Bitcoin, 58% of founders of cryptocurrencies chose to remain anonymous. However, interestingly and despite Bitcoin, this also is an indicator of bankruptcy, since 79% of defaulting cryptocurrencies have anonymous founders.
The research also suggests that coins that last have lower minimum rewards per block, both in absolute terms and relative to their total coins. As a result those coins with lower controlled supply seem to have greater longevity.
A First Step
As the researchers highlight, this is not a definitive study in cryptocurrencies. In fact, the dataset includes only proof-of-work cryptocurrencies, not proof-of stake. Plus, many variables that could be useful in determining cryptocurrency success weren’t available consistently across the dataset and so could not be used robustly. Furthermore, the project attempts to model the probability of the cryptocurrency defaulting within 4 years, which is slightly different to the chance of a coin truly succeeding.
Nonetheless, when analyzing any new cryptocurrency coming to market there are a few things to note. Perhaps most strikingly, the odds are not in your favor if investing in a new cryptocurrency. The majority of them (59%) will not last 4 years if history is any guide. However, you may be able to put yourself in a better position by looking for a strong day one performance, but a calmer first month, and avoiding high levels of pre-mining and relatively large rewards. Of course, the cryptocurrency space is changing relatively rapidly, with new coins today using different algorithms to earlier versions. Still the researchers find that their model predicts 3 out of 4 cryptocurrency defaults, even out of sample. These early results from a relatively basic model seem encouraging first step.