Stochastic volatility with missing data: Assessing the effects of holidays

Abstract

In empirical finance, it is usual to consider holidays as if they do not exist. The main goal of this paper is to assess the effects of holidays on volatility estimation and prediction. Holidays are taken into account by assuming they are missing values in a time series of returns generated by a Stochastic volatility (SV) model. Estimation is evaluated through Monte Carlo experiments. In addition, we assess the effects of holidays on one-step ahead Value-at-Risk forecasting using several time series returns. The results are slightly better when we take into account the missing values, especially for VaR forecasting.

Keywords:

• Estimation
• forecasting
• value-at-Risk

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 As in Cascone and Hotta (2019).

2 On the contrary, it is assumed that weekends do not produce missing values.

3 Here we are interested in the number of missing prices, given by v_p , while in Section 2.1 we are interested in number of missing returns expressed as v. It can be easily seen that $v=v_p+1$.

4 We also have results for $n\in \{550,1050\}$ which are available upon request.

5 Thus, they are the same returns used in the Standard Method.

Additional information

Funding

The second author acknowledges financial support from the São Paulo State Research Foundation (FAPESP), grant number 2018/04654-9. The authors also acknowledge the support of the Center for Applied Research on Econometrics, Finance and Statistics (CAREFS).