In the first article of this three part series, Humans v. Machines, I explored how global economies have entered a digital arms race where data has supplanted oil as the most important commodity. Because the impact of technology manifests in improved productivity, immense power accrues to companies that can improve economies of scale. Consequently, there is legitimate concern regarding the development of “data monopolies”. Yet, these “data monopolies” are highly dependent on the users of their products and services to share the data created by their activity. In effect, the information economy has a circular dynamic.
Prior to the Internet, information was centralized and controlled by monopolistic media providers. Before the proliferation of cable television, for example, there were only three broadcast networks that enjoyed monopoly power and control. Local information was primarily delivered by multiple daily newspapers, with enough demand to warrant both morning and evening formats. Information was essentially a “push” model, in which teams of editors determined what the public was entitled to learn.
The internet transformed the information economy into a distributed “pull” model. The consumer of information acquired leverage and control over how information was received and what they chose to learn. In an application of a well known set of business principles, the buyer of information in the information economy has strong bargaining power while there is low bargaining power for the suppliers of information. As the information economy evolved and social media became its most prominent manifestation, the barriers to entry of information suppliers collapsed.
Leveraging homophilic traits of most societal groups, social media platforms benefit from the tendency of people to place greater trust in information from people with whom they share common traits or beliefs than centralized sources of information subject to journalistic standards. Given this dynamic, should we be surprised by the proliferation of propaganda and “fake news” over platforms like Facebook? Should we surprised that social media could be corrupted by actors with malicious intent to disrupt the democratic process?
Additionally, the signal to noise ratio has dropped sharply as the amount of available information has grown exponentially. One exabyte of data is equivalent to 900 trillion pages of plain text. It is estimated that the sum total of human-produced information through the turn of the 21st century was 12 exabytes of data. We now produce 10 times that amount every single month. Data has been effectively commoditized, and like all commodities, the value of data is derived from volume, not price. Each incremental data point becomes less valuable, but the entirety of a data base becomes more valuable. Does our system encourage or facilitate “data monopolies”? Certainly, we do not discourage their formation. Why has the entire world gravitated toward a singular search engine, for example? A simple explanation might be that Google’s service is superior in terms of efficiency and convenience.
Information providers are essentially data purveyors. In order to be a successful data purveyor, a company must create or distribute content that encourages greater engagement. User engagement becomes a source of data that information providers then utilize to enhance their own value to advertisers and sustain their business models. In order for the information economy to grow, flourish, and avoid disruption, information providers (i.e., data purveyors) must ensure integrity of content, and users must maintain trust in these platforms. The exploitation of Facebook by malicious actors in the 2016 election and the antitrust suit in Europe against Google regarding search results represent direct challenges to these business models.
To date, no Internet-based business model has been developed that does not rely to a certain extent on advertising. Advertising attempts to identify and reach those who could be interested in a particular product. Cigarette companies, for example, ultimately do not care if smokers are male or female, young or old. As with cigarette ads, companies hawking their wares over the Internet are indifferent to the nature of the engagement. Their only objective is to encourage more people to develop the addiction. They care only about increasing the consumption volume of their product.
In other words, internet advertisers are not necessarily concerned if the “user” of a social media platform, App, or web site becomes addicted to the information content available on a particular platform. The business model simply requires user engagement. Often, content is distributed free of charge. Because the platform or service is considered “free” (e.g., Google’s search engine), the “user” supplies additional data through his or her activity that is then, in turn, shared by the owner of the platform with advertisers. Individual engagement allows for better analysis which, in turn, affords advertisers the opportunity to develop more “targeted” ads. This improves productivity and profitability for their clients in the process.
The technology that supports the information economy has, in effect, transformed individual behavior into billions, trillions of bits and bytes. Those packets of data are then analyzed in an attempt to derive predictive value. Ultimately, data purveyors develop applications to predict individual consumer behavior or performance, which increases the value of any future data that is collected, provided the methodology is sound. Such transformations have not escaped the financial markets. There is a school of thought that individuals cannot and should not be trusted with the management of their own money. Humans are generally prone to biases with a tendency to succumb to emotions. Investing should be entrusted instead to dispassionate computer algorithms that supposedly improve with each incremental data point.
“Rules-based” investing is the outcome of years of study in behavioral economics. We humans are not rational beings, at least not consistently. Rules-based investing, therefore, was an attempt to save humans from themselves – from making horrible financial decisions at the worst possible times because of the innate tendency to succumb to emotion. Watching this evolve, companies in the business of creating financial products developed “rules-based” approaches (e.g., index funds) and “automated rules-based” approaches (e.g., robo-advisors; factor based investing; smart Beta; low volatility funds).
“Rules-based” investing works well during periods of extended low volatility and single direction price action because the approach excises judgment from the investment process. Buying and selling of securities are determined by index construction or some pre-established set of rules or formulae. Transactions determined by these factors often contribute to market momentum, both up and down. “Rules-based” investing, which involves “forced” or “indiscriminate” transactions, is the province of algorithmically driven investing strategies.
Automated transactions do not allow for judgments under uncertainty. The flaw with this approach is that “rules-based” investing ultimately generates transactions determined singularly by pre-established rules and purposely disregards or discounts judgments, as judgments reflect fundamental considerations. “Rules-based” approaches, therefore, often conflict with fundamental approaches and are, in effect, a form of investing based on technical factors. Because of the proliferation of “rules-based” investment strategies and the influence of social media platforms, is it possible that “fake news” could cause a stock market crash? Certainly. As of yet, an algorithm cannot distinguish between fake and real news. Until the information economy improves the integrity of content, the impact of “rules-based” investing will likely be to exacerbate volatility, not suppress it.