In this paper, we propose a formal test for density forecast evaluation in presence of dependent data. Apart from accepting or rejecting the tested model, our smooth test identifies the possible sources (such as the location, scale and shape of the distribution) of rejection, thereby helping in revising the initial model. We also propose how to augment the smooth test to investigate explicit forms of dependence in the data within the same test framework. An extensive application to S&P 500 returns indicate capturing time-varying volatility and non-gaussianity significantly improve the performance of the model. Although we are dealing with index returns, the proposed smooth test can be applied to other financial data for exchange rates, futures or forward markets, options prices, inflation rate, analyst forecasts among many others.
Score test, Probability integral transform, Model selection, GARCH model, Simulation based method, Sample size selection
Business | Corporate Finance | Finance and Financial Management
China International Conference in Finance 2015, July 9-12
GHOSH, Aurobindo and BERA, Anil K..
Density forecast evaluation for dependent financial data: Theory and applications. (2015). China International Conference in Finance 2015, July 9-12. 1-57. Research Collection Lee Kong Chian School Of Business.
Available at: http://ink.library.smu.edu.sg/lkcsb_research/5087
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.