Pick a winner based on log-likelihood
In this exercise, you will practice using log-likelihood to choose a model with the best fit.
GARCH models use the maximum likelihood method to estimate parameters. In general, the bigger the log-likelihood, the better the model since it implies a bigger probability of having observed the data you got.
Two GARCH models with different distribution assumptions have been defined and fitted with the S&P 500 return data. The normal distribution GARCH is saved in normal_result, and the skewed Student's t-distribution GARCH is saved in skewt_result.
This exercise is part of the course
GARCH Models in Python
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Print normal GARCH model summary
print(____.____())
# Print skewed GARCH model summary
print(____.____())
# Print the log-likelihood of normal GARCH
print('Log-likelihood of normal GARCH :', ____.____)
# Print the log-likelihood of skewt GARCH
print('Log-likelihood of skewt GARCH :', ____.____)