Long range dependence and sentiment as drivers of volatility in bitcoin’s market
It is inarguable that bitcoin is one of the world’s most volatile currencies, especially when compared to fiat currencies such as the USD, JPY, and GBP. Even though high volatility is threatening bitcoin’s chance to become a widely adopted currency, academic researchers are interested in the reasons underlying this volatility. Most research studies have stressed that less rational factors, such as speculations and/or attention of the society, represent plausible elements driving bitcoin’s volatility. Moreover, long range dependence have been used to predict bitcoin’s volatility and to formulate profitable trading strategies.
Researchers have proven that there had been a link between sentiment and the price of bitcoin during one of the most volatile periods in bitcoin’s history: during the end of the year 2013 and the beginning of the year 2014. Throughout this article, we will take a look at sentiment and long range dependence as factors influencing volatility of bitcoin’s markets.
Sentiment and bitcoin’s volatility:
In behavioral finance, investor sentiment refers to a group of beliefs regarding the return of investment and associated risks which cannot be proven by the facts in hand. Sentiment can also be defined as misperceptions that can result in mispricing of an asset. Bitcoin is one of the financial assets which may be highly prone to be influenced by sentiment due to a number of reasons including:
1- Bitcoin is still relatively new with non-abundant information sources.
2- Bitcoin’s price formation is still poorly understood.
3- Bitcoin is a complex currency, whose understanding requires extensive technical knowledge of programming and cryptography.
4- Bitcoin is mysterious as it had been created by an unknown person aliased as Satoshi Nakamoto. The mystery is augmented even more by the story of people who made millions with very small investments in the early days of bitcoin.
5- Bitcoin operates on a novel platform, the blockchain, which is unknown to conventional economists.
6- Big financial institutions are still not part of the bitcoin economy, which is yet to take place to increase the public’s trust in the currency.
Research shows that Reddit can reflect the sentiment of the bitcoin market. It has also been proven that during periods of high volatility, an explanatory sentiment value rises as well, particularly positive sentiment.
Long range dependence across bitcoin’s markets:
During the past few years, researchers were able to recognize repeated cycles of increased volatility during the same months each year. For example, during the past couple of years, bitcoin experienced a pronounced bullish run during the months of November, December, and January of the following year. Some has correlated this rise in bitcoin price to the Christmas holidays. Those who support this postulation explain that a large number of individuals use their bitcoins to buy Christmas gifts which is responsible, at least partially, to the rise in bitcoin price.
A recently published paper used high frequency data of bitcoin to examine the long memory features of bitcoin’s conditional and unconditional volatilities throughout different time scales using the ARMAFIAPARCH model, the local Whittle (LW) estimator, and the exact local Whittle (ELW) estimator. Results of the study prove that the long memory parameter is quite stable and significant for conditional, as well as, unconditional volatility measures throughout different time scales. The paper also investigated the long memory features of the conditional (realized) and unconditional volatilities of bitcoin across different time scales via means of the ARFIMA model, local Whittle (LW) estimator, and exact local Whittle (ELW) estimator.
Long memory has been shown to be stable and significant also when conditional (realized) and unconditional volatilities are considered. The study also undertook quarterly non-overlapping rolling window analysis to better understand the evolution of the long memory parameter over different time scales. The results proved that there exists high persistence across the bitcoin market. This study has multiple useful implications for market participants and investors with different exposures in bitcoin’s market as per their trading horizons. The findings of this study can help them in predicting bitcoin’s volatility, and thus developing and deploying successful trading strategies, during different times of the year.