About Universal Ratings

Founded in Singapore in 2016 by J. Marczyk, G. Graci and J. Glinski, RTS Ratings Pte Ltd is a privately held and independent fintech company. RTS Ratings measures the complexity of stocks in stock universes. The goal is to identify high-complexity stocks which increase volatility and risk and which may reduce the performance of a portfolio.


 Rationale

Our stock rating system is based on the following rationale:

  • Excessive complexity is a formidable source of fragility and vulnerability. Complexity-induced risk is a huge problem. It makes things fragile.

  • The economy is punctuated by crises, bubbles and destabilizing events. In addition, it is characterized by turbulence and volatility. In such a context it is often more important to know what not to do. Wealth preservation is often more important than sheer performance.

  • Inexperienced investors should stay away from complex products. UR provides a systems which measures this complexity.

  • Our stock rating provides unique information: a quantitative breakdown - a sort of a ‘CAT scan’ - of the complexity of a stock universe.

The bottom line: we don’t tell you how to pick stocks or how to build your portfolios - we tell you which stocks to avoid.

 

The Math

We do not resort to conventional approaches such as statistics, Monte Carlo methods, fuzzy logic, cluster analysis, fractals, chaos theory, Big Data, traditional AI, neural networks or PCA. Instead, we use a new and proprietary solution which has been engineered specifically for a highly unstable, complex and non-stationary regime in which mainstream analytics is not applicable because of the nonlinearities present in real-life data. Our approach is model-free and has its roots in quantum mechanics and biology. This means we don’t need to build and validate a model each time we have to solve a new problem. We don’t use machine learning techniques because our clients cannot afford to suffer enough failures to teach an algorithm how to react properly.

Benoît Mandelbrot said: “Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.” Using linear tools to an evidently non-linear context is wrong and risky to say the least. Assuming that a distributions is Gaussian just because certain closed-form solutions are elegant or simple is bending the reality to suit a tool. Using PCA with evidently non-linear data is like using laminar flow methods to solve turbulent flow fluid dynamics problems. The art of doing mathematics lies in using the right method for a given problem.

Our stock rating is based on a quantitative theory of complexity developed by J. Marczyk. This approach constitutes the foundation of the World's first complexity-specific standard, published in Italy in December 2015, the UNI 11613 Business Complexity Assement guidelines and the ISO 22375 which was published in October, 2018.