Chartered Financial Analyst (CFA) Practice Exam Level 2

What does the Bayesian Information Criterion (BIC) measure?

1. Model complexity

2. Model fit

3. Statistical significance

4. Forecast accuracy

The correct answer is: Model fit

The Bayesian Information Criterion (BIC) is a criterion used for model selection among a finite set of models. It is particularly focused on balancing the trade-off between model fit and model complexity. Specifically, BIC provides a means to evaluate how well a model describes the data while penalizing the model for its complexity—typically measured by the number of parameters it includes. BIC is calculated using the likelihood of the model, which reflects how well it fits the data, and incorporates a penalty term that increases with the number of parameters in the model. As a result, a model that fits the data well but is overly complex will receive a lower BIC score compared to a simpler model with adequate fit. This penalty helps to prevent overfitting, making BIC a valuable criterion for selecting a model that achieves a good balance between complexity and fit. In contrast, the other options do have relevance in different contexts but do not encapsulate what BIC measures. For instance, while model fit indeed refers to how well a model represents the observed data, BIC specifically contextualizes that fit by incorporating model complexity. Statistical significance relates to hypothesis testing rather than model selection, and forecast accuracy refers to the predictive power of a model rather than a criterion for selecting the