RISK MANAGEMENT AND MONITORING

RISK MANAGEMENT

Our sophisticated Risk Management system is stratified to capture the different dimensions of risk and leveraged by the predictive power of deep neural networks. 

We consider the current risk per trade, per trading system (daily and weekly), and at the portfolio level.

Our risk management system considers structural breaks and market regimes to identify the interdependencies between risks and their pricing.

Finally, we combine these results with the ones coming from options (which are naturally forward-looking) in a Bayesian framework.

colorful-2468874_1920 (1)-min

MONITORING

watercolor-2681039_1920-min

Our monitoring system is based on five different levels:

  1. We backtest each trading system weekly to compare the recorded performance with the expected one.
  2. Similarly, we compare the ex-post performance of the execution algorithms with the ex-ante one to verify that the execution costs are aligned with the expectations.  
  3. We compare the current level of drawdown and risk with the related boundaries from the backtesting to identify performance deterioration on time.
  4. We monitor the business cycle dynamics and compare the recorded performances with one expected, given the prevailing market regime.
  5. We propose some market specifications and, following the simulation approach proposed by De Prado (2018), assess the related trading systems’ performances. Then we monitor the statistical properties of the current data generating process for the relevant time series of returns (equity indexes, commodities, …). This allows us to see how close a market is to a given market specification and which return we should expect for each trading system.