Historical analysis of Bitcoin social mentions and sentiment index

When comparing cryptocurrency market to the traditional stock market one of the things that stand out is that many of the traditional tools for evaluation that are available for stocks are not present for the cryptocurrencies.

E.g. when considering the valuation, the company value is the value of discounted cash flows that the company expects to realize in the future. This concept was first introduced by John Burr Williams in his 1938 book “The theory of investment value” (an excellent book btw, for me on par with Security Analysis of B. Graham, bought a reprint of Williams’s book and read it in full in my days as the finance quant).

No comparable and useful concept exists for crypto coins. In absence of financial statements and traditional financial analysis there are other metrics that have arisen in the crypto world. But they do not have the same power as the DCF and/or financial ratios valuation in the stock market.

This has led to more reliance on opinions as one way to evaluate the value of coins. And in our era of social media, there are literally millions of opinions in form of social media posts each day, about cryptocurrencies.

This is what platform BittsAnalytics uses as the source of their opinion or sentiment analysis. Platform collects more than 1 million tweets each day.

To determine sentiment of each post, one needs a scalable approach of machine learning. A text classification model was trained and is deployed on this mass of posts each day to determine each tweets sentiment polarity (sentiment or negative). To determine sentiment of news, it is useful to utilize article extractors to determine the relevant texts from websites.

After all posts have their sentiment computed, average sentiment is determined for each coin in coverage, on both hourly basis and daily basis.

Here is an example of Bitcoin sentiment chart on hourly basis:

This hourly sentiment is computed from typically 10,000 to 15,000 tweets that are published on Bitcoin each hour.

Hourly mentions of Bitcoin can be seen in the following chart:

Hourly mentions often spike during corrections or major crashes in price.

Another chart that is also interesting is the daily number of Bitcoin and BTC mentions but from longer historical perspective. BittsAnalytics has been collecting this data since August 2017 and here is the chart for the Bitcoin:

There has been 3x rise in the daily social media interest for Bitcoin in the recent period when compared to earlier years.

Similar patterns can be observed for coins like Dogecoin:

Solana:

and others.

In our next article we will take a closer look at the metrics and indices which are computed by taking all coins into account and which give us information on the whole crypto market.

Among these indices are Crypto Fear Greed Index which measures the Fear and Greed in the crypto market.

Another interesting set of BittsAnalyics indices are those that are based only on those social media posts that were published by investors / traders / hedge funds / other professional investors. One is Investor Sentiment Index measuring the social sentiment of investors and the other is Investor Social Buzz Index which measures the social buzz of investors.

An interesting application is also exploring the usage of technologies that are used in specific field.

A popular one is e.g. Slider revolution, and if wants to find more information about it, like which technologies are used with Slider revolution, from our analysis:

One can examine other interesting things, e.g. which verticals like to use this technology: