For a long time, financial economists have been arguing about market efficiency—the extent to which the prices of financial assets contain all relevant information. If one believes in the strong form of the efficiency hypothesis, i.e., that market prices contain all relevant information, then all assets would be fairly priced and there would be no way to generate risk-adjusted excess returns on the market. Only passive investment strategies would make sense in that case.
Yet, we know that some market participants earn excess returns for a protracted period of time, which casts doubt on the strong form of market efficiency hypothesis. Active investment strategies clearly pay off. However, they require good data.
Traditionally, investors have been using both individual financial and economy-wide data to value stocks and predict their returns. Metrics such as price-earnings, price-cash flow, EV/EBITDA, dividend yield, or price-to-book value have been envisaged to forecast prices using individual company accounting data. Furthermore, a wealth of macroeconomic and relevant sector-level data have been brought to bear on forecasting the stock prices. Sophisticated models – whether based on fundamental or technical analysis – were developed to make use of this data, helping some investors gain abnormal returns. However, it seems that even though new models and techniques are being developed every day, it is increasingly difficult to beat the market while using only the traditional sources of data. These data come with a time lag, sometimes misrepresent the true state of the company, often are subject to problematic revisions and do not necessarily capture the reality in a complex way. Hence, investors are increasingly turning to alternatives.
With the advent of machine learning and artificial intelligence (AI), the investment community has begun exploiting alternative data, which the new technology allows to process very efficiently and at a recognition level increasingly on par with humans.
Which alternative data have been gaining traction with investors recently?
Since about the mid-2000s, various social media has become widespread and transformed many aspects of our lives. With a penetration rate of some of the social media reaching sky-high levels, its influence and the information value of its content might be enormous. Prevailing sentiments not only about the individual companies but also about the general economic and political environment are revealed daily by the contributions found on the social media. It seems that there is a value (with regard to asset pricing) to be exploited in scrutinizing how connection networks are formed as well as in analyzing individual contribution posted on social media. Further attesting to this, several academic studies suggest that using data from social media can help predict stock returns or market events. (1,2)
On a separate front, many hedge funds and other investors have started using satellite-based information on shipping containers, mines, ports, parking lots, plantations, or farmlands, which can capture economic activity in real time. From this, they can derive insight into the overall economic activity and/or even expected sales of public companies involved. One particular study indicates that the number of containers in ports can significantly predict stock index returns in 27 out of 33 countries at a daily frequency for the 2019–2021 period. An investor making use of satellite data on marine ports can, on average, receive an annualized return of 16.4%. (3)
Increasingly, investors, especially those exposed to stocks in the consumer goods sector, have been trying to gain additional insight into consumer behavior. Transaction data from debit and credit cards clearly contain valuable information that can be exploited to forecast stock returns. One recent study shows that debit/credit transaction data positively predict various measures of a company’s future earnings surprises up to three quarters in the future (4). The study found that a portfolio constructed using insights from card transactions generates returns of 16% per annum net of transaction costs. Clearly, consumer transaction data garnered from various sources can be enormously valuable to investors.
The market for mobile phone applications has been booming in recent years. With the increasing number of application users, the mobile application market provides a new fresh source of data that could help predict a firm’s future financial performance. One particular study investigates the relationship between mobile app usage and future stock returns utilizing a panel data set of 326 public firms that have released mobile applications on the Apple iOS App Store or Google Play Store (5). The results of the study show that monthly abnormal app downloads and abnormal daily users are positively associated with the next-month abnormal stock returns. This suggests that mobile app usage can help predict firm-level future stock returns.
Above were just some of the examples of alternative data frequently used by the investor community nowadays. Other examples include data on bank loans announcements, product review data, electricity consumption data, or data on shipping trackers. Increasingly, investors are also using geolocation and climate data to predict economic activity or a specific company’s performance. Sophisticated investors also increasingly use AI technology to deeply analyze transcripts of the earnings calls by employing tools like ChatGPT for detection of tone, hidden clues, and nuances in corporate statements. They also scrutinize video earnings calls for both tone of the CEOs’ voices as well as their body language. Interestingly, some investigators go as far as using biometric data of CEOs or data on their spending patterns to predict their personality types and accordingly stock returns or a likelihood of potential misconduct.
In a similar vein, according to Deloitte’s report, one of the largest Japanese investment managers used data from job portals and recruitment websites to infer the strength of organizational culture in individual firms to help generate investment picks (6). Clearly, there are no bounds to one’s imagination and inquiry, and we will probably see a significant expansion of using alternative data going forward.
Deloitte’s report predicts astounding growth in the size of the market for alternative data (6). While revenues for alternative data providers reached an estimated USD 11 bn in 2024, the market is likely to grow by more than 50% annually (CAGR) until the end of the decade. By then, the total revenues for alternative data providers are predicted to surpass those of traditional financial data providers.
What are the implications for investment managers?
Given the current market dynamic, it seems that investment firms that shy away from using alternative data and fail to incorporate them into their business model might not only forgo alpha but also be exposed to additional risks due to suboptimal decisions.
This is because alternative data can create an information advantage for the investment managers via (6):
- Unique insight – alternative data can provide insight that is not available from traditional financial or economic data
- Timeliness – these can be real-time or nearly real-time data, allowing investors to make much quicker decisions than relying on traditional financial data
- Predictive power – alternative data can improve the predictive power of forecasting models, helping anticipate the market movements or the company performance
Rush or perish?
Unique and timely insights into asset pricing, together with forecasting models’ higher predictive power, can contribute to the higher alpha of investment firms that use alternative data. In this very competitive investment landscape, it seems that in the longer run, investment managers that intensively use these data will earn abnormal returns and clearly outcompete those that do not. Furthermore, alternative data can improve risk assessments and lead to more effective detection of early warning signals about potential risks. All of this means that the firms on the frontier of adopting alternative data will gain a competitive edge over firms that have been relying only on traditional data. This will also likely lead to a higher customer satisfaction and retention in the technologically most advanced firms, the ultimate objective of any business undertaking.
In a nutshell, rushing to adopt the most recent technology and embrace alternative sources of data might be the only way to remain competitive in this market.
Vladimír Zlacký, LookingEast.eu
References:
- Cookson, J. Anthony; Lu, Runjing; Mullins, William; and Niessner, Marina. The Social Signal. Journal of Financial Economics, volume 158, 2024
- Cookson, J. Anthony; Niessner, Marina; Schiller,Christoph. Can Social Media Inform Corporate Decisions? Evidence from Merger Withdrawal, working paper, 2025
- Honghai Yu; Xianfeng Hao; Liangyu Wu; Yuqi Zhao; Yudong Wang. Eye in outer space: satellite imageries of container ports can predict world stock returns. Humanities and Social Science Communications, 2023
- Tarun Gupta; Edward Leung; Viorel Roscavan: Consumer Spending and the Cross-Section of the Stock Returns. The Journal of Portfolio Management July 2022
- Ziqing Yuan; Hailiang Chen; Can Mobile App Usage Help Predict Firm Level Stock Returns? Fortieth International Conference on Information Systems, Munich 2019
- Deloitte Insights. Alternative data at investment management firms: From discovery to integration, 2024