Chartered Financial Analyst (CFA) Practice Exam Level 2

What does an auto-regressive (AR) model primarily focus on?

1. Time-invariant variables

2. Correlation of the current value with its past values

3. Heteroskedasticity testing

4. Variance of multiple independent variables

The correct answer is: Correlation of the current value with its past values

An auto-regressive (AR) model primarily focuses on the correlation of the current value of a time series with its past values. This type of model is fundamentally built on the premise that past values of the series can provide valuable information for predicting its future values. In essence, it assumes that the past behavior of the variable being studied influences its future behavior. In an AR model, the current observation is expressed as a linear combination of its past observations and a stochastic error term. The coefficients of the past values represent the degree of their influence on the current value, highlighting the importance of historical data in forecasting. Other options, while related to various aspects of time series analysis and econometrics, do not capture the core function of an AR model. Time-invariant variables relate to constant factors across observations and are not the primary concern of AR models, which focus on dynamics over time. Heteroskedasticity testing pertains to the variability of the error terms in regression analysis, and variance of multiple independent variables is associated with multivariate regression rather than the autoregressive approach that emphasizes lagged values of a single variable. Thus, the central theme of predicting current values based on their past states makes the chosen answer the most appropriate in the context of AR models.