Updated October 2023
The volume of data has grown about 30,000% over the past 20 years. Wealth management firms must embrace artificial intelligence and other new tools. Data science will no longer remain the preserve of large quantitative managers.
Our world is changing – and very rapidly so. Advances in technology have brought about an explosion in the amount of data that humanity is producing daily. To put it in context, according to Sergey Brin, co-founder of Google, from the beginning of humanity to the year 2003, only about 5 exabytes of information or 0.5% of a zettabyte (1021 bytes) was ever [1]. By contrast, by 2024, the volume of data is estimated to grow to 147 zettabytes (Chart 1). This means that from 2003 to 2024, the total amount of data will have grown a whopping 30,000%.
Historically, investors have relied on what data scientists call structured data, which comes in a standardized format and can be efficiently accessed by software. Leading the astounding growth in data, however, is the rise of unstructured, non-numerical data – text, video, audio, etc. With unstructured data representing approximately 80% of the total, financial advisors must learn how to put it to use in their portfolios.
Chart 1: Growth in Global Data Volume
Source: Statista, 2023 Additional notes: The data was taken from various publications released over several years: Forecast for the years 2018 and 2019 as of 2018; Forecast for 2020 as of May 2021; Forecast for 2021 to 2025 as of March 2021 based on figure for 2020 provided by the source. Figures were rounded to provide a better understanding of the statistic. The figures from 2021 to 2025 were calculated by Statista based on the 2020 forecast figure and the five-year compound annual growth rate (CAGR) of 23 percent provided by the source. The figures prior to 2020 are based on IDC's forecast from late 2018.
Investment firms lag in data-analysis capabilities
As the world evolves, most investment management processes remain largely unchanged – or at least, have advanced at a much slower pace than digital technology. Many investors continue to rely on traditional ways of processing information, which primarily consists of formulating their views based on news, research reports, or conversations with experts.
But investors are exposed to far more information than they can process organically. Considering a human can likely retain just five to nine digits in its short-term memory bank for no more than 15-30 seconds[i], processing all the useful data we come across daily is impossible. Great investors can focus on a handful of valuable signals and disregard the rest as noise. Unfortunately, this is a skill that few possess, and this is where data science comes in – it can provide crucial tools to help financial advisors make sense of the onslaught of information.
While many quantitative tools have been around for decades, most financial advisors have been slow to embrace them. Until recently, tools that analyze big data have been out of reach of all but the largest firms with extensive quantitative expertise. It required specialized skills in statistical and data science analysis to acquire, clean and make sense of data, making the barriers to using big data high. Additionally, the perception was that “quants” were a special breed of investor who spoke a different language and looked at the world through different lenses (which often were as thick as Coke bottles.) For many institutional investors, let alone retail investors, the divide between quants and other investors appeared too wide to bridge.
Bridging the Divide
The proliferation of large language models (LLMs) and other forms of AI in recent years has been breathtaking. AI and many data-science methods can now be used by a layperson to extract valuable insights from unstructured data, such as articles, podcasts, etc. This data can be used alongside other inputs to inform investment decisions in an efficient, transparent, and testable fashion. For example, advisors can use LLMs to synthesize views of multiple strategists efficiently and combine them with more traditional models to generate market outlooks.
Investors who ignore the growth of big data run the risk of being made obsolete by those who embrace it. With this in mind, our company built a system that makes data science accessible to all financial advisors in a way that is compatible with how traditional investment management businesses are run. Our approach creates systems that are not only quantitatively robust but also consistent with current practices and behaviors. In doing so, we aim to substantially lower the barriers to using quantitative analysis and make this new frontier available to all investors.
The potential benefits of utilizing data science and artificial intelligence are manifold. Financial advisors can achieve greater efficiencies and deliver better investment outcomes for their clients.
[2] Miller, G. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. The psychological review, 63, 81-97. Atkinson, R. C., & Shiffrin, R. M. (1971). The control processes of short-term memory. Institute for Mathematical Studies in the Social Sciences, Stanford University.
Comments