Strategies

Our process leverages Big Data. We start with a variety of traditional and non-traditional datasets. The data is cleaned, stored in our data lake and leveraged in the forecasting process. We process both structured and unstructured data. Structured data refers to financial data, prices, value chain, consumption and population data. Unstructured data is both text and image data. Regardless of the type of data, we have an appropriately chosen, trained AI algorithm to integrate and forecast. Our proprietary Natural Language Processing (NLP) algorithms allow us to analyze unstructured and semi-structured data at scale. We source both from traditional sources (news, transcripts, annual reports) as well as various open source data we collect. We have designed our own scraping platform which enables us to collect data that is customized for our own specific needs. Sentiment analysis relies on reading media outlets, public reports, quarterly / annual filings and whatever is accessible in order to understand how positive or negative the company narrative is. Normally this is a slow and laborious task, involving large groups of analysts with various biases of estimating sentiment. This process becomes particularly intense and sometimes impossible during earnings, four times a year, lasting six weeks. Our systems not only offer more accurate and reliable analysis by reading accessible data at scale but also allow us to read data in various languages across the world – just imagine how hard this would be for a team of analysts and portfolio managers. . Agent Based Modelling (ABM) is a branch of Artificial Intelligence that is not as popular as neural networks. However, it offers unique advantages and is a perfect complement to other computational methods as it allows for reacting to information that has not been seen before (data not trained on) and operating in new environments (as markets transition). Governance is about making decisions and we believe good governance is central to the agency relationship between management and us, as shareholders. Fairness, transparency, good ethics, diversity and a well functioning board are principles which we just use in managing our business but also expect to see high and improving standards in our investments. Our approach to investing in well-governed companies is differentiated by our technology. Our edge in identifying good governance comes from being able to infer governance scores for companies that are too small to be properly understood by the wider investment community. It is also important to “connect the dots” when reading news, reports, quarterly filings and other sources. Which companies, what products, which key people were mentioned? How are they related? Have they been mentioned in the previous discussions? All of these are consistently reviewed and evaluated, across all the sources we see (as well as new ones we add over time). Society is a central pillar to economic activity. B2B enterprises and governments alike exist to ultimately serve consumers. Businesses can not prosper, in the long term, unless they have a positive contribution to society. Social order means a stable society and a stable society means a business can continue operating for a long time. Our investment process centers around looking for companies that are making positive contributions to society, as well as striving to do better. Our approach to investing in a sustainable way is differentiated by our technology. Our edge in social investing comes from being able to infer social scores for companies that are too small to be properly understood by the wider investment community. We borrow computational techniques from Computational Fluid Dynamics in Physics. Our High Performance Computing cluster allows us to run over 650 computing cores in parallel, accomplishing in two hours something which would take weeks to compute with a traditional architecture. The environment is the medium in which we live our lives. It concerns every aspect of what we do, including making investments. Thus, it makes total sense that our investments would also act in accordance with our interests. Our investment process centers around looking for companies that are making positive contributions to the environment, as well as striving to do better. Our approach to investing in a sustainable way is differentiated by our technology. Our edge in environmental investing comes from being able to infer environmental scores for companies that are too small to be properly understood by the wider investment community. Machine Learning is a wonderful tool but it’s not the universal solution the media would make you believe. We believe the best approach is to mix and match best computational techniques with the appropriate tasks. This computational technique works best at automating decisions that have to be taken in a stationary environment. In an investment context, these tend to be more mundane tasks, like detecting whether a company is impacted by a certain theme (or not). These inferences are made continuously, allowing us to detect changes in a thematic for a company ahead of the street. Our data lake is a unique blend of storage solutions - object-based storage, document storage, relational and graph databases. Certain datasets are easier to store and retrieve in a specific storage solution, both for speed and scalability. Security is paramount, which is why our data is stored both in transit and at rest. While our process is end-to-end digital, it does not mean there are no people involved. We prefer to think of our process as “human in the middle” - the algorithms collect data, learn, forecast, build portfolios and manage risks. Portfolio managers oversee the outcomes and make adjustments as needed. Being digital does not have to mean poor or no customer service. On the contrary, we enhance our customer service by being more transparent and accessible to our customers. Technology is not a substitute for human interaction! Best in class customer service is accomplished by blending technology with a personal touch! The evolution of computing, beginning in the 1950s, has traversed the mainframe era, the personal computing era, and currently, we find ourselves approaching the twilight of the smartphone era. While IoT has been defined to some extent, it has yet to witness widespread implementation, widespread adoption, and the consequential impact on productivity. In summary, computers have enabled the execution of index investing and quantitative strategies, resulting in reduced costs for investors and a decline in profitability for the asset management industry, particularly for traditional fundamental asset managers who have seen limited benefits beyond email and Excel. Moreover, this technological advancement has fueled the exponential growth of global information, often referred to as “Big Data,” which traditional fundamental asset managers struggle to harness. We remain committed to fundamental investing, but unlike many others in the market, we have embraced a fully digital approach. Our operations encompass large-scale data ingestion, a proprietary data repository, automated forecasting, portfolio construction, risk management, and client reporting. To succeed in the next generation of fundamental asset management, it’s essential to adopt a technologist’s mindset and leverage cutting-edge tools and techniques. This is the way of the future. With time, more fund managers will have to take an approach like this to remain competitive. This is a seamless combination of a classical approach with cutting-edge AI technology. We love the confluence of deep experience of the team and a technological edge, resulting in a very unique product. Any single one of these modules is a differentiation in and of itself. This innovation is amazing, keep it up!