Big data and mobile geolocation can predict a firm’s future fundamentals, providing a new and potentially powerful investment tool.

Fundamental analysis has long been both the foundation and the calling card for mutual funds, wealth management and institutional investment. Celebrity investors such as Warren Buffet have long espoused the importance of investing in the profitability, growth and operational efficiency of companies, as opposed to treating their share prices like random numbers to be gambled on as if driven by animal spirits and whims of the capital market.

Academics and practitioners alike have spent decades building models for stock prices using financial statements, forecasted future earnings and even comparable company analysis, seeking to capture the true nature of firm performance. But today a new source of potential revelation comes from technology: big data.

Studies from the Fintech Research Center at Fudan University’s International School of Finance utilise mobile big data, that is, the data gleaned from location information captured from mobile phones and other devices, to predict both key accounting variables and stock prices. This data allows analysts to determine the actual number of employees working at any given time and the number of shipments to and from manufacturing and other facilities. This captures labour as well as key manufacturing inputs and outputs critical to production.

There are several conclusions we can reach from using this data. Firstly, big data is far more timely than traditional measures of fundamentals. Big data is available in real-time, though researchers have focused on aggregate daily data, whereas accounting statements offer delayed information at a quarterly or annual frequency. It is also more objective and not subject to accounting manipulations or “earnings smoothing” techniques.

In a study of more than 4400 listed companies over five years, big data accurately forecasts future revenues and earnings, well in advance of the release of official reports, and allows investors to act in advance of that reporting.

Even a naïve strategy that buys or sells the best or worst firms as measured by big data yields a substantial return and an annual alpha of nearly 7 per cent, even after including consideration of the most important traditional measures, such as CAPM alpha and other factors. This outpaces nearly all mutual and hedge funds over the span of study.

By looking at discrepancies between big data fundamentals and reported accounting numbers, one may even be able to identify instances of potential accounting fraud in advance of the market and even of regulatory bodies.

New wealth management

It bears asking how these findings potentially impact the wealth management space. As part of our research, we investigated a new fund management firm formed as a joint venture between a big data tech firm and a traditional securities company. The former uses AI to manage and auto-label real-time big data, generating thousands of profiles. The use of profiles as opposed to personal information, along with appropriate governance and regulation, ensures data privacy.

These profiles coupled with geolocation information can be used to determine not only how many people are entering and exiting a particular location but also who those people are likely to be, for example: employees, short-term visitors, or shipping/logistics providers. This, of course, can be done simply by stationing a person at the entrance to a company as well. However, with thousands of firms and many multiples of that in terms of locations and facilities, doing this manually becomes untenable.

Financial firms armed with this new data can use it to make a variety of key decisions. These may include tracking the amount of product being shipped to forecast revenue. It can likewise be used to look at both downstream cost fluctuations or upstream supply chain reliability. Creditors can use this data to monitor the health of operations for borrowers.

In and around the Covid-19 pandemic, this data was used to track slowdowns in production and logistics, as well as to monitor return-to-service. One can use outputs from upstream providers to forecast demand from downstream finished products, all potentially weeks before sales are actually generated for these products. Scale is key. The high level of automation in big data coupled with AI means we can track 1000 companies almost as easily as following one, vastly increasing work efficiency and potential coverage.

Blurred lines

Indeed, we see a truly organic synergy generated that blurs the line between tech and finance, a true “fintech” revolution on the horizon. Will the future be finance firms adopting new technology or technology firms expanding into financial services? Evidence suggests both. Indeed, one can easily foresee an ecosystem of regulators, financial service providers, individual investors, as well as data and AI technology firms, all working with the same data to create a more efficient, lower cost, more stable capital market in virtually all asset classes.

We should not get carried away. It is important to keep in mind that results reported describe the aggregate. Big data applied to individual stocks or over limited time periods will not necessarily yield the same results. Indeed, investors would be encouraged to view big and traditional data as complementary and to use all available information to make investment decisions. The impact of data on returns relies upon this data being used by a substantial portion of the investment population to deliver returns. So big data in the immediate future is more likely to augment existing investment strategies rather than act a stand-alone investment signal.

However, it appears we are on our way to something a bit special. With these new technologies and capabilities improving every day, a truly next gen investment advisory or wealth management service provider is one that is forward looking, open-minded, and exceptionally agile. It seeks out new skill sets and provides new employment opportunities to a tech-savvy and big data-literate constituency. In short, it has innovation and rapid evolution as a core competency.

Charles Chang is director of the Fintech Research Center and professor of finance and Chloe Yang is assistant professor of finance at Fudan University’s International School of Finance