Are Cryptocurrencies Correlated? ? Binance Finds Out…
A recent study released by Binance Research, the research arm of the largest cryptocurrency exchange by trading volume, has delved into the question of whether the returns of the major cryptocurrencies are correlated with one another. This is a pertinent question for anyone who is in any way exposed to the prices of cryptocurrencies.
The question is especially relevant to those who need to take such associations into account when exposing themselves to the prices of various cryptocurrencies. This post breaks down the Binance study to present the key findings and exceptions.
What even is correlation?
Correlation is a measure of the strength of association between two variables.
If two variables are positively correlated, they will tend to move together. If they are negatively correlated, they will tend to move in opposite directions.
If there are low levels of correlation between the variables, there is little association between them. Values above 0.5 are considered to have moderate to strong correlation and values below 0.5 are considered weak to no correlation.
Positively correlated variables will have a plus sign and negatively correlated variables will have a minus sign.
What the study found?
The study analysed the returns of the top thirty cryptocurrencies from 1st December 2018 to 1st March 2019 in terms of both USD returns and BTC returns. The study also compared these results with correlations from 1st December 2017 to 1st March 2018 and correlations from 1st June 2018 to 1st September 2018.
The study found high levels of correlation in terms of USD returns between bitcoin and other major cryptocurrencies with an average correlation of 0.78. These levels of correlations were actually higher than in previous periods.
In terms of BTC returns, the levels of correlations were lower with correlations decreasing as the periods progressed.
What is driving this correlation?
The high levels of correlation were noted by Binance as underscoring bitcoin as the bellwether of the cryptocurrency market. Bitcoin represents over 50% of the market share of the cryptocurrency market at the time of writing. Furthermore, BTC is tradable against almost all major altcoins.
The price movements of bitcoin have an unquestionable association with those of altcoins. The value of 0.78 for the average correlation between bitcoin and top altcoins is a very high value that indicates a strong level of association.
One of the potential reasons for the correlations increasing across the periods examined is the volatile price movements taking place in late 2017. Periods of high volatility can result in cryptocurrencies moving more independently from one another and thus having lower correlation values.
Binance found that the cryptocurrencies listed to Binance were more correlated among each other than cryptocurrencies that were not. Dubbed the Binance effect, the study note that the increased correlation is likely due to the cryptocurrencies having access to the same pool of liquidity.
Cryptocurrencies that were not listed to Binance such as dogecoin exhibited significantly less correlation with the cryptocurrencies that were listed to Binance.
The possibility that the propagation of stablecoins in 2018 played a role in the correlation was also noted. The amount of volume traded through stablecoins has increased from less than 10% one year ago to over 30% of all trading volume on Binance.
When traders speculated in altcoins prior to the widespread propagation of stablecoins, traders typically transitioned between bitcoin and the altcoin using the BTC trading pair. For example, if a trader speculated in ether, they would typically use the ETH/BTC trading pair.
However, with an increasing amount of stablecoins and an increasing amount of altcoin versus stablecoin trading pairs, this is a likely factor which is driving the increased amount of correlation being seen in USD returns for major cryptocurrencies.
Are there exceptions?
There were also a number of exceptions of cryptocurrencies where less correlation was seen. The Binance effect has already been noted and dogecoin is one of the cryptocurrencies experiencing low correlations with others due to this.
However, other cryptocurrencies such as tron, bitcoin cash SV, and waves also experienced lower correlations due to what Binance has labelled as idiosyncratic events. Idiosyncratic events can be a range of factors which drive the price to be less correlated with other major cryptocurrencies.
For example, factors such as hard forks which was the case for bitcoin cash SV can play into price. Idiosyncratic events essentially drive the price of a cryptocurrency to react differently from how the rest of the market is moving.
Binance also found evidence that the consensus mechanism may be having an impact on the correlations among cryptocurrencies with cryptocurrencies that share the same consensus mechanism having higher correlations than those that do not. NEM, which utilizes a proof-of-authority consensus mechanism, exhibited significantly lower levels of correlation than the majority of other cryptocurrencies analysed.
Practical takeaways?
Why does this matter? There are a number of different ways correlation can factor into the decision making for different entities. When trading and investing, correlations will often be taken into account when considering risk exposure and hedging. By investing in assets which are positively correlated, a portfolio is more exposed to the price movements of assets whereas investing in negatively correlated assets serves to somewhat hedge the amount of risk. A number of other businesses often factor correlations into decision making including miners, merchants, and vendors.
