Decentralization: A Brief History in Tech
The world wide web has history with decentralized architectures. At its inception, the internet was a collection of services built on open protocols, which allowed participants to build presence with the assurance of permanence. Tech giants, capable of mobilizing more resources and scaling with more direction, changed this landscape drastically while eliminating the richness of available content. This history reflects the antagonistic relationship between efficiency and variety: in the early days of the web, the factor driving the rapid growth of tech firms was efficient resource deployment, at the cost of functional variety.
This sort of efficiency was necessary due to the large expense inherent to emerging technologies: compute power and data competency (collection, storage and analytics) were an enormous expense, and with AI still in the cradle, the manpower needed to manage operations was valued at premium. Today, these advantages to centralized structures are not as decisive as they once were: compute power and data storage are cheaper than ever, and AI, while toddling, is further along in functional implementation than it once was. At the same time, technological evolution is driving towards ever faster innovation, creating a hunger for variety centralized architectures can’t satiate.
The tech sector has taken note of the changing landscape, and responded in kind: companies like GitHub and Automattic (the entity behind WordPress) have long eschewed hierarchical systems in favour of flat, decentralized architectures, reaping the rewards of collective intelligence- faster innovation and more motivated workers- in the process. Similarly, organizations like InnoCentive harness the intelligence of groups to solve scientific problems for clients, faster and on a smaller budget than possible within closed networks.
Decentralization in Finance: A New Frontier
Per Thomas Malone, MIT organizational theorist, in “high tech, R&D-oriented industries, the critical factors of business success are often precisely those benefits of decentralized decision making: freedom, flexibility, motivation, creativity” because these economies are “knowledge-based and innovation-driven” . Many of the factors driving the tech, engineering and biomedical sectors hold true for the finance vertical- and a semblance of disruptive decentralization has begun to emerge here as well. Innovative platforms like QuantConnect, Quantopian, Numerai, and Kaggle have taken on the task to bring asset management to the masses in the form of competitions and crowd-sourced algorithms. However the potential of a decentralized ecosystem that keeps IP in the hands of participants while providing analysts with powerful tools to solve problems remains untapped.
Aided by advances in cloud computing, data analytics and AI, open availability of a platform for quantitative analysis, collaboration and exchange will give rise to collective intelligence- a faster, more efficient, and widely more innovative creative mechanism. Collective intelligence can most effectively take hold when the crowd is able to behave in an intelligent manner, meaning a quantitative research platform aiming to empower its user base must provide the necessary tools to help users make sense of the problem- in this case, the problem being too much (yet not enough) data, and the rapidly changing nature of the markets.
Availability of better tools for purchase from independent agents than those that can be generated in-house will lead to desiloing a more equitable distribution of profits from fund manager pockets to best-in-class innovators as funds try to compete with emerging players. Equeum intends on making marketplace essentially ownerless; will be compensated proportionally for building and maintaining infrastructure, but will not act as an intermediary so as to allow for true competitive market.
Collective Intelligence: Precipitating Factors
The term “collective intelligence” is often associated with socialist notions of collaborative problem solving or consensus decision making, and it is easy to overlook that these behaviours arise in different forms within a free-market capitalist framework. The mechanisms that drive productivity and innovation in decentralized systems are based on efficiency of work allocation, iteration, competition, and efficiency of knowledge transfer.
Unlike hierarchies, decentralized networks are self-organizing, meaning tasks are self selected based on an individual perspective of personal competencies, project interest, and anticipated reward. This individual-driven approach to work allocation results in markedly more enthusiastic participants who typically find themselves suited to and excited about the projects they take on. In addition, the quality of self-determination decentralized structures bestow on participants has a side benefit of increasing efficiency; compensation in such an environment is truly commensurate to contribution.By tying compensation to delivery as opposed to process, decentralized systems can produce solutions in a more cost effective manner- buyers pay for a solution, not its development. For reference, it is estimated that successful on-boarding of a quantitative analyst runs funds up to 500k.
The self-organizing nature of decentralized systems is facilitating to innovation. By distributing work based on interest and rewarding only successful product, these networks naturally facilitate diversity, competition, synthesis, and evolution, creating consistently better products on an accelerated cycle. While centralized organizations are typically built to a certain level of completion (with a capacity to transform only very slowly), decentralized networks are built with extension and enrichment by participants in mind- meaning the richness of the content and services available grows exponentially.
In environments requiring complex decision making, it is more efficient to move intellectual product than it is to move the information itself. Decentralized networks excel in these circumstances, as they allow an endless number of participants to specialize at each discrete step in a process, making it vastly more efficient. Given the proper tools, a set of standard protocols, and an environment that empowers participants to set terms for and maintain ownership of their work, the financial vertical will be transformed with ever better work product. Data scientists, analysts, and any investors will reap the rewards.