Machine learning model using neural networks y for predicting stock price in the future.

When we do day trading, we look for a way to know how the price of the stock of a company we acquire will close. This model simplifies predicting a neural network to forecast the price as well as an ARCH model to predict the variation the next day.

It is very important since one can take a buy position when the model indicates an increase in the stock the next day.

The model is recursive and is updated at the close of the market (3pm) and does not include post-market or publication of results.

Efficiency of 80% vs previous executions.