Methodology

We make decisions based on the empirical analysis of financial data. Our focus is rooted in science and technology, transforming the way investment decisions are made.

  The forecasting and anticipation capacity that humans have regarding the price of a given instrument is very limited. That is why at Ursa, we use the scientific method to identify recurring behavioral patterns in the various markets and the specific variables that determine the price. In this way, we detect potential opportunities that can be taken advantage of by our investment systems.

  Alpha, which is how we quantify our mathematical investment advantage over the markets, comes from empirical and rigorous metrics that, together with the creation of indicators and use of alternative data, allow us to develop successful investment systems.

  Data management is a fundamental part of the alpha determination process. The more extensive and diverse the data sets available, the greater the scope and probability of success in corroborating or refuting hypotheses, ideas and inferences about the behavior of a market under specific conditions.

  The formulation of specific hypotheses in the search for alpha typically aligns with observations that occur recurrently in the data, or that have a consistent logic depending on the nature of the market and the human psychology on which it is based. That is why Ursa's experience in operating financial markets is essential.

  Algorithmic investment systems that are obtained from the hypotheses that successfully generate alpha are tested extensively under different market conditions and different execution mechanisms. Before operating them, they go through an incubation period, where we confirm that our results are robust, and are kept under constant monitoring to confirm that the alpha calculated during development is consistent with the alpha observed during operation. The result of this process allows us to create algorithms with proven mathematical advantage.

Design Process

Hypothesis and Quantitative Analysis

Preparation of premises and quantitative analysis of market behavior.

Design and BackTesting

Ruleset and filters formulation that capture presented opportunities, tested on historical data.

Evaluation and Optimization

Analysis and sensitivity tests of statistical performance to optimize parameters.

Risk Management

Definition of risk management policies, protections, position sizes and rebalancing frequencies.

Operational Implementation

Selection of intermediaries for execution considering costs, reliability, degree of service and liquidity. Development of algorithms for automated execution.

Periodic Review

Evaluation of the performance of the algorithm in its theoretical basis and real operation. Continuous improvement on parameters and risk measurements.