The Increasing Prominence of Alternative Data in Finance
December 9, 2021
By Acuity Trading
In the last few years, hedge funds and other investment firms have increasingly tapped into alternative data to drive their strategies. Information advantage has always been the holy grail of finance, the primary factor in driving alpha and risk management. Any edge, such as a narrow timing advantage, can lead to a more effective algorithm, trading signal, or investment model. No data is too obscure, as long as firms can use it to meet their clients’ investment needs.
By late 2020, nearly 45% of investment firms were using alternative data to gain a competitive edge. This number is expected to grow as firms invest in new technologies during the pandemic. A 2021 report by AIMA, together with Simmons & Simmons and Seward & Kissel, revealed that 48% of investment firms with AUM above $1 billion and 18% with AUM below $1 billion are investing in alternative data, driven by AI and machine learning. The adoption of alternative data is growing exponentially and will continue to do so in the next 3 to 5 years.
What is Alternative Data?
Alternative data in finance encompasses data culled from non-traditional or non-financial sources to drive an edge in the market. Various digital and non-digital interactions generate such forms of data, including financial transactions, satellites, mobile devices, sensors, internet use, public records, and more. Newsfeeds, online communities, metadata from communications, geo-spatial information are some of the other popular sources.
By using advanced technologies like AI and ML, it is possible to extract valuable information from these data sources. For instance, sentiment analysis can be fuelled by Natural Language Programming, which derives insights from earnings call transcripts or evaluates the language complexity of earnings calls. The same technology can be used to extract information from tweets, blog posts, speeches, press releases and presentations, at a rapid pace to provide easily quantifiable metrics.
Must-Know Types of Alternative Data
Data from multiple additional sources can help fund managers gain a more accurate idea of a company’s overall position. Managers can use this data, along with traditional data sources like SEC filings, financial statements, economic data and more, to decide on where to invest. Some established categories in the industry are:
Consumer Spending Patterns
Credit card transactions on specific apps can give an idea about company sales, without any impact on consumer data privacy. Spending patterns in particular seasons can help investors gain insights into which sectors to invest in during various stages of the economic cycle.
Real-time satellite images are being used to predict retailers’ sales. For instance, the number of cars in the parking lot of a store chain can indicate whether its stock is overvalued or undervalued. A 2019 research revealed that if investors bought shares of a retailer weeks before it reported its quarterly earnings, due to higher parking lot traffic, and then sold the shares when traffic was lower, the return earned could be 4.7% higher than the typical benchmark return. Satellite imagery can also help forecast supply chain disruptions, due to forest fires or other natural disasters, all of which can impact a company’s bottom line.
Wall Street’s interest in GPS data is understandable. Where people are going reveals broader consumer movement trends. This information has become more valuable during the pandemic, as brands want more information about consumer behaviour.
Social Media Sentiment and Product Reviews
Investment firms are using social media data to evaluate stocks or overall market sentiment about an asset or economy. Data that tracks the number of “likes” and check-in counts on Facebook for several companies, for instance, is valuable for investors. Twitter is also a vastly used platform by companies, traders, and investors alike. Combining information derived from Twitter with sophisticated analysis tools has been very helpful for investment managers. Even as far back as 2012, researchers proved that the daily number of tweets can be correlated with S&P 500 stock market indicators. Also, closing prices of stocks can be predicted using data from Twitter.
Similarly, product reviews can be evaluated by investment managers to identify demand levels and predict future performance.
Some other popular sources of alternative data are:
Government contract activity
Ship vessel tracking
Investment choices of politicians
Corporate flight tracking
How is Alternative Data Used?
With the growth in the types of data available from websites and apps, the use cases of alternative data are expanding.
Stock Price Predictions
Data can be gathered from various sources to predict stock price movements. This can include data from public discussion forums, foot traffic in stores, earnings call transcripts, analysts’ forecasts, and more. Similarly, inquiries to credit agencies can reveal supplier payment history, which can provide clues about the financial stability of a company.
Web scraping can be used to track the prices of millions of products online to measure inflation levels. Inflation impacts asset price movements, as well as the direction of monetary policies of central banks. Tracking inflation could be crucial for risk management.
Commodity Price Predictions
Satellite imagery has been used to track the movement of oil tankers and ships carrying essential raw materials for industries like iron ore, copper, and natural gas. These can be used to predict supply and demand gaps, and the future price of commodities.
Downloaded apps can reveal the location data of people. This can be used to evaluate foot traffic in particular areas, which is an indication of possible changes in real estate prices.
Data helps investors stay ahead of the curve on key investment factors like:
Making sense of economic data that impacts forex movements.
Augmenting fundamental analysis with sentiment data to monitor changes in the market over time.
Identifying risks in specific sectors through high-frequency news flows.
Analysis of behavioural flows in the market.
Driving an Edge in ESG Investments
Machine learning is a huge advantage when it comes to ESG assessment. Companies release ESG disclosures frequently on social media, news channels, blogs, forums, and other platforms. These sources reveal events experienced by a company related to environmental and social responsibility, such as shortage of water, deforestation efforts, and racial discrimination. Machine learning models can apply ESG taxonomy (based on ESG standards) to assess companies consistently. It helps managers raise ESG levels in their clients’ portfolios and also focus on risk-adjusted performance.
Alternative data is changing the face of finance and the capital markets. Investment managers who fail to follow this trend and update their processes face strategic risks. They can be outmaneuvered by competitors.