We examine the impact of temporal and portfolio aggregation on the quality of Value-at-Risk (VaR) forecasts over a horizon of 10 trading days for a well-diversified portfolio of stocks, bonds and alternative investments. The VaR forecasts are constructed based on daily, weekly, or biweekly returns of all constituent assets separately, gathered into portfolios based on asset class, or into a single portfolio. We compare the impact of aggregation with that of choosing a model for the conditional volatilities and correlations, the distribution for the innovations, and the method of forecast construction. We find that the level of temporal aggregation is most important. Daily returns form the best basis for VaR forecasts. Modeling the portfolio at the asset or asset class level works better than complete portfolio aggregation, but differences are smaller. The differences from the model, distribution, and forecast choices are also smaller compared with temporal aggregation.

Additional Metadata
Keywords Aggregation, Forecast evaluation, Model comparison, Value-at-risk
JEL Time-Series Models; Dynamic Quantile Regressions (jel C22), Time-Series Models; Dynamic Quantile Regressions (jel C32), Model Evaluation and Testing (jel C52), Forecasting and Other Model Applications (jel C53), Financial Forecasting (jel G17)
Persistent URL dx.doi.org/10.1093/jjfinec/nbx019, hdl.handle.net/1765/102524
Journal Journal of Financial Econometrics
Kole, H.J.W.G, Markwat, T.D, Opschoor, A, & van Dijk, D.J.C. (2017). Forecasting Value-at-Risk under temporal and portfolio aggregation. Journal of Financial Econometrics, 15(4), 649–677. doi:10.1093/jjfinec/nbx019