Sentiment, Best Predictor of Bitcoin Prices: Interview with The TIE’s Joshua Frank
More than anything else, sentiment moves cryptocurrency prices.
Share this article
Twitter Sentiment is one of the most powerful predictors of Bitcoin and cryptocurrency prices, said Joshua Frank in an interview with Crypto Briefing. Leveraging this information can give traders an edge in profiting from the crypto markets.
Technical traders rely on various indexes to interpret the price action of a given cryptocurrency while fundamentalists put emphasis on specific events to identify trading opportunities. The TIE, a data provider for digital assets, uses a different methodology that it defines as “wisdom of the crowd.”
In a Crypto Briefing exclusive interview at CoinDesk’s “Invest: NYC,” Joshua Frank, the CEO of The TIE, said “in the digital assets market, sentiment is one of, if not the largest mover of asset prices.” Frank shared his insights about how sentiment can be used as a predictive instrument as well as other metrics that his company employs to understand the health of different networks.
The Power of News in Sentiment Analytics
In any given market, investors usually respond to breaking news in real time. Thus, news is considered to be a useful tool in anticipating trends and movements in the market. In the cryptocurrency industry in particular, news tends to break first on Twitter, making this platform one of the most reliable sources to analyze the psychology behind the position of the current market cycle.
Under this premise, The TIE scored an exclusive partnership with Social Market Analytics (SMA), a provider of predictive sentiment analytics. The idea behind the collaboration was to leverage access to the Twitter ‘firehose’ — the full real-time stream of every tweet sent over Twitter.
Since then, the New York-based company has been able to scan over 850 million daily tweets using machine learning and natural language processing technology to determine the sentiment behind major cryptocurrencies. Such a technical tool is common in traditional markets, but is one of the many mechanisms missing in the cryptocurrency industry, according to Frank.
“The data that we have on digital assets is exactly the same data that Fidelity uses to service their equity clients. So, if you look up the sentiment of Apple on Fidelity, that is the exact same sentiment data that we have for cryptos,” said Frank.
Over 3.2 million unique accounts have tweeted about Bitcoin this year. Due to the relevancy of this data, the firm has been able to pick up on instances in which extremely positive or negative conversations are occurring around the flagship cryptocurrency.
An example of this, Frank explained, was the news of the Bitcoin ETF rejection on July 26, 2018. As soon as the ETF was rejected the news hit Twitter, and The TIE’s algorithm picked up on this sentiment.
“Sentiment quickly dropped extremely negative and downward price movement trailed for 15–30 minutes following the negative sentiment spike. As sentiment began to neutralize and turn positive, upwards price movement followed.”
From its eight years of experience and millions in invested dollars, the firm concluded coins with positive sentiment and price momentum substantially outperform the market, while those with negative sentiment and price momentum underperform. Given this knowledge as a trader, now it’s a matter of picking up these signals and trading against that information.
Other Sources for Sentiment Data
Although Twitter is the main source for The TIE’s crypto sentiment and price momentum analysis, it also takes into consideration over 2,000 English and Chinese news sources.
Frank revealed that everything from Crypto Briefing to CoinDesk, from the New York Times to the Wall Street Journal, from the Medium page of IOTA and Ethereum Classic to a Chinese publication, is processed by a computer models that assesses the relevancy of each news publication and its effect on the underlying asset.
“In real-time, we can tell whether an event is happening or is going to happen at a specific time in the future. With this information, we build up a historical database of all of the different types of events regardless of the source where it is coming from, backtest it, and trade on that. Basically, we can determine when X, Y, and Z happen how it affects the price of a digital asset,” said Frank.
Additionally, the firm looks through other data sets, such as every job posting that a crypto company has to measure the health of the network. By looking at the number of jobs that are posted, The TIE can look at a company and predict which departments are products are growing.
The State of the Market
Despite the vast amount of valuable data that The TIE manages, Frank told Crypto Briefing that most of his customers are traditional institutions. These types of investors are familiar with the services provided by different brokerages, such as Charles Schwab, Fidelity, and TD Ameritrade, which also rely on sentiment analytics.
In contrast, the crypto market itself and different funds within it are too small to utilize The TIE’s services. Moreover, most crypto funds aren’t at the scale where it makes sense to spend huge sums of money on data.
Frank estimates that 50 million people have touched crypto. But, in terms of people who are actually involved in the market and actively trading, he would estimate that number is under 250,000. A similar trend can be seen on Twitter since the actual number of unique users discussing crypto is around 40,000 people on a daily basis.
Due to the size of the market and the lack of knowledge about sentiment analytics in the cryptocurrency industry, The TIE developed a free-to-use Beta platform. TheTIE.io enables traders to leverage sentiment analysis to profit from the intra-day volatility in digital assets. Given that most of the market seems driven by sentiment, as Frank argues, then these tools could be some of the most powerful in a trader’s arsenal.