The adage that “data is the new oil” is frequently tossed around in the discourse surrounding financial markets, and for good reason–after all, having the ability to develop niche products, identify potential customers, and essentially predict the future are incredible assets to any company or venture.
But as the market for data grows, so too does competition between data providers. It is perhaps for this reason that the data science and analytics space has grown and evolved so much over the last decade.
Analytics firms are constantly searching for new ways to process the world and create narratives that can help us to understand it better–and in a world that is increasingly online, they are uniquely positioned to have access to practically infinite amounts of information.
This is also true within the world of finance. With the amount of information that is available, there is also an opportunity to examine more deeply how things that may not be typically associated with financial markets can have a powerful impact on the ways that these markets move.
Recently, Finance Magnates spoke with Zak Selbert, CEO and co-founder of LA-based data analytics firm Indexica, about how artificial intelligence is empowering the financial and cryptocurrency industries with new kinds of alternative data, as well as the ways that his own firm is analyzing things like the way people talk about assets in order to predict how they will move in the future.
What is Indexica?
Indexica, which describes itself as an “interactive web-intelligence platform, leveraged to transform global events into quantified metrics, predictive indexes, and strategy signals,” was founded by Selbert, along with Miguel Enguica and Tobin Hanspal, in 2012.
Selbert explained that while his background is in finance, the other two co-founders of Indexica “have completely different skill sets, ranging from econometrics to statistics to natural language processing and other types of data-based skills.”
“The premise of Indexica is that what’s going on in the world around us–and the opinions on what’s going on in the world around us, most importantly–impact markets in important ways. Whether it’s rain in Brazil impacting the price of coffee, or the opinion of Jeff Bezos impacting Amazon stock price.”
“So, if you multiply that by everything that’s happening in the world, and everything that trades in the world–it starts to become this unlimited amount of data to crunch. And so, our methodology for making this practical is to say that the news is generally narrating what’s happening in the world around us. It’s narrating events really, really well,” he said, adding that this is especially true if you have the ability to “crawl” news from everywhere in the world.
“It’s also very good at picking up emotions or opinions about what’s happening around the world–not just sentiment [analysis],” which simply measures positive and negative statements, “but anger, fear, uncertainty, probability, opportunity–all these different ways that you can quantify various feelings. So, we crawl the news and we quantify [it] into this portfolio of metrics.”
These metrics are then compared with market data, which the company either does itself as part of a product offering, “or our clients do it, using models that are somewhat blind to us.”
Alternative data & atypical metrics: for example, “Futurity” can be used to compare verb tense with stock performance
What do these metrics look like? Selbert explained that “futurity,” for example, “reads language and analyzes whether or not the tense [of the language] is in the future or the past…it also extracts times and dates.”
“So,” for example, “if [it was] mentioned in a news article that ‘Amazon was going to be releasing a fleet of drones next year’,” the company would quantify that statement “as being ‘future’”, Selbert explained, because of the use of the phrase ‘next year.’”
“Whereas, if we read an article that said ‘Boeing grounded its fleet’ of 737s’, we would ground that as ‘last year.’”
FBS Gives Away Signed FC Barcelona Jerseys for Playing Penalty SimulationGo to article >>
“So, imagine multiplying that type of verbiage by millions of sentences from all the text that we read, and then mapping it to publicly traded securities,” as well as certain cryptocurrency assets, like Bitcoin. “That’s what this metric does–and interestingly, if you look at it, and you look at market performance, there tends to be a very strong relationship between how often and to what extent a company is spoken about in the future [tense], and then the performance of their stock.”
In other words, the more that a company is mentioned in the future tense, “the better it tends to do.”
Selbert explained that Indexica believes that this is most likely because “if you’re mentioned in the future [tense], it’s because you’re innovating, you’re attracting investments, you’re releasing new products; good things are happening. Whereas, if you’re mentioned generally in the past [tense]–although it could be something good that happened in the past–the news cycle tends to be that it is referencing something bad that’s happened.”
The moral of the story here? “Companies should aim to be spoken about in the future. The future has to be better than the past for markets to do well.”
Indexica found data implying that “Bitcoin is part of a much larger ecosystem”
Several times throughout this year, as part of its “Predictive Indexing” product, Selbert explained that Indexica used futurity and other metrics to discover some interesting things about Bitcoin markets. The findings have been published in Bloomberg and a number of other publications.
— Bloomberg Markets (@markets) October 11, 2019
Selbert explained that the findings varied based on the time period that the metrics were being used to study. Earlier in the year, the futurity of Bitcoin was found to have “predictive power for the price of Bitcoin.”
Selbert added that this “led us to think that Bitcoin was moving like a stock would to some extent because futurity tends to have a lot of predictive power when we’re talking about equities–but we’ve never seen that before with Bitcoin, so that was a real revelation.”
Since then, “we’ve been continuing to follow what is moving Bitcoin,” Selbert continued. “And lately, it hasn’t necessarily been the futurity of Bitcoin [itself].”
Instead, Indexica has found that the things that may be moving Bitcoin are not necessarily directly related to the coin (at least, not in the way you might think): “it has been other metrics applied to other entities, like ‘quotability’ for [fiat] currencies,” Selbert said, referring to another of Indexica’s metrics that essentially counts the number of unique quotations in the news that mention a security or another type of tradable asset, like Bitcoin.
Other things that could be influencing the price of Bitcoin include “sentiment for Mastercard, futurity for gold, volume for ETFs, and buzz sentiment for blockchain.”
Selbert explained that “this really implies–if you look at the entity [itself]–that Bitcoin is part of a much larger ecosystem, and what moves it could be anything that happens in that ecosystem,” adding that Indexica has the ability to nail down more precisely what that “anything” could be.
“Rather than making a blanket statement like this, Predictive Indexing was able to nail down exactly which entities one needs to pay attention to, and exactly which metrics should be used to measure what’s going on with those entities.”
“…This is a way to create a narrative out of a lot of quantitative data around Bitcoin–so, this is kind of the human element of the artificial intelligence that Predictive Indexing employs.”