Deep Learning Asset Pricing Main Code for Training

Tested Under:

• Tensorflow 1.12.0
• Python 3.6

Example Code

Step 1: Training the SDF network

$ python3 run.py --config=config/config.json --logdir=output --saveBestFreq=128 --printOnConsole=True --saveLog=True --ignoreEpoch=32

Step 2: Run the first 8 cells of model_GAN.ipynb to generate SDF

Step 3: Run create_RF_data.py to generate the data with R * F

Step 4: Train the beta prediction network

$ python3 run_RtnFcst_ensembles.py --config config_RF --logdir output_RF --task_id 1 --trial_id 1

Step 5: Run the remaining cells of model_GAN.ipynb to get EV and XS-R2 pricing results

Datasets can be found at Google Drive