June 21, 2021 By Katelyn Rothney 4 min read

Asset managers, banks, and other financial institutions employ armies of analysts to sift through spreadsheets and media chatter. But even the biggest players can’t hope to keep up with the volumes of data produced by the modern global economy.

“I was managing north of $30 billion as a portfolio manager for Apple,” says Chris Natividad, now chief investment officer and co-founder of EquBot, an AI platform and Portfolios as a Service (PaaS) company. “It’s impossible for a human portfolio manager or investment manager to look at everything, on top of all the financial statements, all the different news articles, industry reports — it’s just impossible.”

EquBot is a fintech company that makes the AIEQ, billed as “the world’s first AI-powered equity ETF” (Exchange Traded Fund). To launch AIEQ, EquBot partnered with ETF Managers Group, one of the leading ETF providers. The platform was designed to give investors faster access to smarter insights from broader, deeper data sets. It allows users to construct models using machine learning, knowledge graphs and natural language processing (NLP), rigorously test those models, and understand the factors that drive their performance. It’s been live for almost three and a half years.

“We give the analogy that it’s like thousands of research analysts working around the clock, while operating in dozens of different languages — and they know what one another knows, all at the same time,” says Natividad.

AIEQ is powered by several IBM Watson technologies, such as natural language understanding (NLU) in Watson Discovery, and Watson Studio for bias detection and reduction. AIEQ is collecting data on over 6,000 US companies each day, including structured data from third-party data providers. But where AIEQ really shines is collecting and parsing unstructured data stored in formats that are difficult for analysts to inspect quickly.

A clearer picture, every year

Natividad recalls January’s social media frenzy that sent video game store GameStop’s stock price soaring. AIEQ is designed to monitor unconventional data sources like blogs and social media, not just financial statements in standardized formats.

“Each one of these different data points, whether it’s a market price falling, or social media posts by Yellen or the US treasurer or an Elon Musk for that matter, it’s a data point,” says Natividad. “It’s a pixel in a broad, global economic picture, or a US market picture.” The more of these data points EquBot can access, the higher the resolution that picture will be, and investors can make better decisions based on the granularity of that image.

By combining structured and unstructured data, AIEQ enables EquBot to select portfolios that are more likely to have the highest opportunity for market appreciation. Over time, the knowledge graphs that IBM Watson allows EquBot to build are growing, allowing for more predictive accuracy over time. As the assets have grown, the ETF continues to improve. In AIEQ’s first year, it underperformed against the broad US market. The next year, it matched this benchmark, and in the subsequent year, it significantly outperformed the US market.

Under the hood

One of Natividad’s goals is to dispel some industry myths that persist of AI being a “black box” technology. He describes four models that make up the assessment of a given stock in AIEQ. The first is a financial score, which describes the overall financial health of the company as determined by the kind of trading statistics found in financial reports that would traditionally be interpreted by teams of quantitative analysts.

But even up-to-the-minute financial data is not enough context to understand a stock’s current value. To place the stock within a broader context, AIEQ leans most heavily on Watson’s NLU capabilities. A recent case study that demonstrates these capabilities occurred as pharmaceuticals began to consider COVID-19 treatment options. Watson enabled AIEQ to spot COVID-related trends early, allowing the fund to capitalize on the frenzy of market enthusiasm around several healthcare brands. The platform found north of 3,000 clinical trials on FDA.gov relating to COVID-19 treatments, testing and vaccines. Between scouring these and industry reports, some names that are now synonymous with the COVID-19 vaccine effort began to pop up — Moderna, Pfizer, J&J.

Beyond searching for the frequency of keywords, Watson enables AIEQ to perform sentiment analysis, providing additional context around treatments (like whether a given treatment passed a clinical trial) and the trustworthiness of the data source. Is a source traceable to multiple other instances of market movement? Are other respected voices pointing to this source? How much prestige or readership do those sources have? All of these relational data points help to paint the most detailed picture possible.

“Some of those positions have done tremendously well for the AIEQ,” says Natividad.

After taking news data into account, AIEQ applies a score to the company’s management team by looking at industry leadership data, environmental, social and governance scores, and historical performance data for the company’s leadership team.

Finally, the fund considers external factors like the overall financial health of the company’s industry and host country, and the current position of global and local economies within economic cycles.

Leveling the playing field

Natividad explains how working with IBM Watson enables EquBot to punch above its weight. “IBM Watson really allows many of the smaller players, even us initially, to get our foot in the door to say ‘Here’s what you’re missing with your investment approach,’” he says. Nimbler companies like EquBot are able to lead the charge into AI-powered solutions because they are able to take more risks than large financial institutions. However, Natividad observes a transition happening now as his company speaks with sovereign wealth funds, endowments, banks and large asset managers. As AIEQ’s track record extends, the platform has allowed EquBot to have conversations with these larger, more conservative investors. Big financial institutions are now taking notice, but EquBot’s foresight in implementing AI has given them a lead that promises to widen as their fund continues to learn.

Learn how EquBot built AIEQ using Watson

Was this article helpful?

More from Artificial intelligence

Democratizing Large Language Model development with InstructLab support in watsonx.ai

5 min read - There is no doubt that generative AI is changing the game for many industries around the world due to its ability to automate and enhance creative and analytical processes. According to McKinsey, generative AI has a potential to add $4 trillion to the global economy. With the advent of generative AI and, more specifically, Large Language Models (LLMs), driving tremendous opportunities and efficiencies, we’re finding that the path to success for organizations to effectively use and scale their generative AI…

IBM watsonx Code Assistant for Z: accelerate the application lifecycle with generative AI and automation

2 min read - The business outcomes from scaling AI are clear. Companies at the forefront of generative AI and data-led innovation are seeing 72% greater annual net profits and 17% more annual revenue growth than their peers. A key step toward unlocking these gains is the adoption of purpose-built AI assistants that are tailored to your use case. IBM watsonx™ Code Assistant for Z is an AI-based coding assistant designed to accelerate the application lifecycle for mainframe applications, leveraging generative AI and automation…

IBM unveils Data Product Hub to enable organization-wide data sharing and discovery

2 min read - Today, IBM announces Data Product Hub, a data sharing solution which will be generally available in June 2024 to help accelerate enterprises’ data-driven outcomes by streamlining data sharing between internal data producers and data consumers. Often, organizations want to derive value from their data but are hindered by it being inaccessible, sprawled across different sources and tools, and hard to interpret and consume. Current approaches to managing data requests require manual data transformation and delivery, which can be time-consuming and…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters