Irithmics was launched in 2012 to develop deep learning technology for public health research as part of an EU backed academic research call.
Our strong background in investments, finance and risk led us to look at ways to apply our technology and deep learning to institutional investment dynamics and fund strategy analysis.
Irithmics grew out of this desire to explore the implications for AI technology on modern portfolio analysis with the ultimate goal of providing a powerful new way for asset managers, corporates and professional financial advisors to understand and anticipate collective institutional investor behaviours.
Our mission is to fundamentally change the way investment decisions are understood by providing powerful insights into how and why markets move as well as the risks of this movement and subsequent consequences.
Using our state-of-the-art AI powered platform, our clients are able to measures, monitor and anticipate the dynamics behind market expectations and appetite, identify core trends in investor behaviour and compare investor’s intentions and views.
The Irithmics platform examines data from over 250,000 global institutional investors, funds and hedge funds daily. Our data provides clarity on changes in portfolio weighting as well as subtle shifts in views and expectations. We empower our clients with the tools to analyse differences in sensitivity to market forces: from the dynamics between longer-term strategic and shorter-term tactical allocations of capital and risk; sensitivity towards news, announcements and events; and vulnerability to selling by systematic strategies or short selling.
The result is a powerful combination of high value data that allows our users to measure, monitor and anticipate market views and the resulting behaviours, identify trends as they develop and make intelligent decisions with increased confidence.
Our technology is trusted by clients around the world and is used by major global Stock Exchanges to provide their issuers with deeper market insights and analysis. Irithmics data is also used by leading investment banks to help with their broker research, with our insights used to provide enhance understanding of ratings, recommendations and estimated target prices.
The technology that we have built to power the Irithmics platform is designed to measure, monitor and evaluate shifting market views and the behavioural changes behind institutional investor decision making. Our platform uses state-of-the-art deep learning originally designed to help describe and understand the spread and impact of long-term chronic human diseases, like diabetes. We have applied these technologies to learn from over 250,000 global institutional investors and funds, across hundreds of thousands of portfolios with a combined value of many trillions of dollars.
Because institutional investors base their decisions on a broad number of factors, such as expectations of return, risk exposure and changes in outlook, understanding the reasons behind how and why investors make these decisions is incredibly valuable. Institutional investors make rational, measured decisions based on their views of current market conditions and our technology helps to identify the reasons behind these decisions – it helps understand the views behind investors’ allocation of capital and risk.
Modern portfolio theory can help us understand how portfolios may be constructed and optimised taking into account the expected return of the market, the volatility of that return and the correlation of returns between different assets. The views (or expectations) of individual fund managers also influence the portfolio weightings and can help explain why one portfolio might be overweight a particular company, while another might be underweight.
How expected returns, volatility, correlation and manager views evolve over time affect changes that are made to portfolios, and ultimately help explain why selling or buying is seen in the market. The true value in these types of insights however lies in the ability to analyse large numbers of portfolios, discovering the patterns and trends affecting how they are built, managed and rebalanced.
Our powerful technology helps our clients anticipate these changes in market expectations, behaviour and dynamics for a deeper market understanding.
Why what we do is different
At Irithmics we use deep neural network technology to allow us to evaluate the views of global institutional investors and how these views affect the portfolios they manager and the risks there are exposed to.
Reinforcement learning and generative adversarial networks enable us to map investor behaviour, identifying and establishing patterns and running powerful analysis between present and potential future events.
Our technology also provides deeper insights into past behaviour and can help identify key historical factors which have influenced market shifts.
The deep neural network technology that powers our platform learns from and evaluates market data against a specific set of key attributes and characteristics which indicate how collective fund managers’ market views develop and how these changing views influence decisions on portfolio weighting. This helps our clients identify and anticipate how these changing views affect investor appetite for particular markets and how they reflect investors’ expectations of a company.
We have identified around 80 of these unique characteristics which are used to classify and describe the behaviour and dynamics of institutional investors. These characteristics all us to measure and anticipate changes in investor views which reflect shifting views and outlook in the wider market.
Valuable, real-time insights
Our platform works by identifying how changes in investors’ holdings for particular companies within a portfolio reflect their collective views and expectations. Our deep learning algorithms help to evaluate why investors’ views may have changed or whether these changes are likely to be part of a wider strategic or tactical changes.
The platform also also evaluate changes in global macro data to estimate whether these are likely to strengthen or weaken a fund manager’s view of the market, or position on a particular holding.
Our state-of-the-art deep neural network learns from changes in markets, anticipating the likely views of fund managers which have resulted in them rebalancing a portfolio or changing their risk exposure to a specific company or sector.
If you are a fund manager, a corporate client or a financial PR professional, our state-of-the-art technology could help you gain a deeper understanding of market dynamics. Get in touch today to find out how our powerful platform can help you evolve your understanding of investor behaviour.