DataLight interview for CoinPost Japan
Danila Chestnykh, CEO and Founder, and Vsevolod Lukovsky, Co-Founder and Chief of Analytics at DataLight, on the unique metrics, tools for collecting data and DataLight’s opportunities for users.
What is DataLight? What made this project possible?
Danila: There was a portfolio headed by me and Vsevolod. We were developing an investment strategy and what we lacked for the picture to be aggregate was the possibility to store all of the projects’ data in one place. It was necessary to compare projects with each other by objective metrics. We could not go gavel to gavel with the miscellaneous information from the websites and other sources. Similarly, analyzing the existing projects was comparing apples and oranges. The data has always been crucial in our own decision-making processes: Vsevolod had previously worked as an analyst for a bank. Consequently, we started to aggregate the data. We first used Google Sheets and filled it in manually. The problem was that the data kept becoming obsolete while we were working with it. As we felt the need to authorize the processes, we have written scripts for Google Sheets. At some point these tables became humongous and the computers ceased to proceed it, warming up and buzzing instead. We deployed the database and started collecting data into it. It has then dawned on us that what we did for ourselves was unique for the crypto-market. And we decided to create a product based on this data.
How do you collect the data for your infographics?
Danila: DataLight is, in fact, a platform that enables you to analyze the crypto assets by different parameters, then make comparisons and inflationary and speculative decisions. The data can be used for research purposes and technical analysis as well. In summary, our infographics are mostly based on the data that is already available on the platform or the data that we have collected but have not yet uploaded.
Please explain what some of your unique data sets tell us about the Crypto market.
Danila: We mostly collect the data from Twitter and Telegram. We will also add the statistics of Google information requests in the nearest future, along with the metrics commonly applied for media: sentiment analysis, number of mentions, etc. But what is the point? It is not a stretch to say that most of us have learned about Bitcoin and other cryptocurrencies from the word of the mouth: either from our friends and acquaintances, colleagues, or from our own social networks or our own media experience. Most likely, before delving into this topic, you have come across certain buzzwords. Only when you started reading more on the topic did you become inspired by the myriad ways to get benefits from these currencies. You then bought it on the stock exchange, made your very first transactions, and eventually became part of the community and shaped your opinion on risks and perspectives. And it is crucial to understand how new people become involved in the industry. They are the actual reason behind the tremendous success of crypto we saw in 2017. A lot of people came to the market, and we saw an organic growth. Crypto assets are undoubtedly of a different nature compared to its conventional сounterparts. Unlike classical assets with a clear business structure (including big companies standing behind them, revenue, profit, etc), Bitcoin and Ethereum are community-driven, do not have an owner, and its’ value is proportional to the community’s size and the degree of the users’ involvement. That’s why bitcoin is #1. There are obviously currencies that are less time-consuming in terms of operations. There are options with more technical sophistication, but Bitcoin still remains unrivaled – thanks to the community around it. We believe that this can be attributed to fundamental data, social metrics, and the benefits you can get are both pure metrics and synthetic metrics, such as capitalization, separative mentions. The division by parameters for various projects and currencies is nearly the same, except for a number of details. Thus, we can determine the currencies that are overbought or resold by a particular parameter (we aggregate stock data as well). All these aspects are very much correlated with price and are often the leading factors, beefed up with social media. And this range of data allows you to see the big picture and make accurate decisions.
It is clearly not the holy grail: speculating effectively and making decisions with one metric only is fairly impossible, but a comprehensive approach allows you to enhance the quality of trading substantially.
Now to the “Total Mood”: what information is used when calculating it?
Vsevolod: We first collect the data from social media, such as Twitter and Telegram using the keywords. For example, the XLM keywords would be ‘stellar’, ‘xlm’, ‘interstellar’, ‘lumen’. We then remove all the messages that are not related to crypto: for instance, ‘stellar’ can be mentioned in a totally different context. The next step is to lower the weight of spam and advertising. After that we use machine-learning and linguistic approaches to determine the tone of messages – which we eventually see in Total Mood indicators.
MarketCap is often criticized for its inaccuracy when it comes to measuring cryptocurrencies. What is the “best” way to calculate the performance of a particular coin?
Danila: MarketCap is one of the most accurate from the market’s perspective. And market evaluation is the most accurate at the point of time. We aim to estimate capitalization in the context of other metrics. We display the cryptocurrencies rating (the global estimate of currencies based on market capitalization, volatility, profitability, social hype and mood, the blockchain characteristics such as the number of transactions, and synthetic, like the usefulness of the network). For example, diving Capitalization/Hype makes perfect sense, given the nature of Hype. It can be explained as the number of references made to the project. The higher Cap/Hype, the more overbought the project is regarding its social virality. Consequently, this will affect our rating negatively.
What are your thoughts on the recent report from BitWise? How do you take “wash trading” and other nefarious actions into account with your calculations?
Danila: We are observing what is happening at the crypto-market. Indeed, we found the report really interesting. In fact, we did something similar six months ago – but with different parameters.
And we largely agree with the results of this report, as we see a tremendous difference in the audience of active exchanges, in their communities, references and the trading volumes. And many exchanges that are claimed to be the leaders of the Top 10, in reality, have a website audience of 50 to 500 thousand users per month. Either they all are millionaires and trade millions of dollars per transaction, or these volumes are too far from the truth. We lean toward the second option – it is unlikely that millionaires will go for these exchanges.
Capitalization is not a perfect indicator, but it is the most accurate from the point of view of the market. And despite the fact that this market is not very young and not very big, and is clearly being manipulated, we still proceed from the fact that the market valuation it is the most accurate at the moment. It can change, the coin can at any time be either overbought or oversold, but the price is still the price and the market is still the market.
How/Why is DataLight better than CoinMarketCap?How does DataLight add value to an industry that is theoretically open to all who have a device to research the blockchain?
Danila: CoinMarketCap is not our direct competitor: they focus on the aggregation of market data. The information itself is useless – what it is important is how to analyze and collect it. We provide a tool for analysis and a variety of synthetic metrics for fundamental analysis. Our product is unique, and this is how we add value to the market. With DataLight you are now able to compare projects with each other and get real-time information with a convenient and user-friendly interface.
Last year Facebook was under fire for their involvement with Cambridge Analytica and having their data misused.
Data transparency has also been a major issue (especially in Japan), with the government tampering with census data to skew results in favor of their preferred conclusions.
How does DataLight provide transparency and clarity?
Danila: We do not collect personal data. We provide the data directly from the exchanges. And we are confident in its quality. As for the data from the blockchain, anyone can check the figures that are shown as it’s fairly available. What we do is we use a specific algorithm to ‘clean’ this data, as it can be hard to comprehend. We do a lot of work, relying on various parameters and seeking out for better trading options.
Using your proprietary data sets/measurements, can you teach us how to practically use it as an example?
Vsevolod: I personally look at the growth of blockchain activity, hype and volumes. Lately, all of these characteristics have increased, especially the trading value. Conversely, hype has not grown so much. Taken together, these factors allow us to be excited over the future. The accurate prognosis is yet to appear, but the fundamental indicators say that we will not see new bottom in the nearest future