VRP AI

How it works for VRP AI Strategy

Moneza uses deep learning - Convolutional Neural Networks (CNNs), Deep Reinforcement Learning (DRL) and Long Short-Term Memory Networks (LSTM) to increase its edge by predicting standard deviation and mean of closing SPX prices in 1- and 5-day sessions. The same principles can be applied to raw market data to determine whether to buy or sell binary spreads.

Convolutional Neural Networks (CNN)

Convolutional Neural Networks (CNN) are series of algorithms that work to identify underlying relationships in complex datasets. In other words, it’s a computerized method to mimic the functionality of the human brain. In our case, this is used to analyze market movements.

Deep Reinforcement Learning (DRL)

Deep Reinforcement Learning (DRL) essentially helps these algorithms learn how to achieve complex objectives.

Long Short-Term Memory Networks (LSTM)

Long Short-Term Memory Networks then helps by allowing previously viewed information to persist. This is like watching a movie while remembering what transpired in the previous scene or reading a book while knowing what happened in the previous chapter.

Together, our technology leverages these systems to analyze the market and extract key insights needed to execute our strategy. In doing so, we developed Ultra, a family of proprietary convolutional nets which we use to extract variance risk premiums more efficiently with increased win rates while reducing maximum drawdowns.