/Active-Stock-Market-Management

NASDAQ Technical and fundamental analysis, and prediction of beating the index over the coming year

Primary LanguagePython

Active-Stock-Market-Management

NASDAQ Technical and fundamental analysis, and prediction of beating the index over the coming year

Data

Data from WRDS and analysis process in another file.

Variables includes:

Fondamentals:

# Yield quarter
# marge de profit
# P/E
# Total assets
# Total liabilities
# B
# P/B
# Nb shares
# % shares issued
# % shares repurchased

Technical:

# Price variation
# Momentum 3 months
# Momentum 6 months
# Median quarter
# Beat median

Models

# Random forest
# Gradient Boosting
# Linear Support Vector Classification

The prodiction can be quickly impove by using another imputer (mean imputations were used) and with a model tunnning.

Results

Models are better to prediction the 'non-beating index'stocks than the one that will beat the index.

On test set (2017) the non beating median precision were 67%. This result was also found on the S&P500 and the DOW.

The use of such model can be in a context of FNB, where the model can suggest sotcks that won't perform compare to the index, en then have a better portfolio than the Index.