Volatility forecast comparison using imperfect volatility proxies

Patton, Andrew J.

doi:10.1016/j.jeconom.2010.03.034

Cited by 955 publications

(452 citation statements)

References 52 publications

Supporting

Mentioning

440

Contrasting

Unclassified

Order By: Relevance

“…Recent papers addressing this include [13,14]. Generally, the suggested volatility proxy is the realized volatility in some particular form, coupled with an MSFE evaluation criterion.…”

Section: Resultsmentioning

confidence: 99%

Forecasting Realized Volatility Using Subsample Averaging

Huang¹,

Lee²

2013

OJS

View full text Add to dashboard Cite

When the observed price process is the true underlying price process plus microstructure noise, it is known that realized volatility (RV) estimates will be overwhelmed by the noise when the sampling frequency approaches infinity. Therefore, it may be optimal to sample less frequently, and averaging the less frequently sampled subsamples can improve estimation for quadratic variation. In this paper, we extend this idea to forecasting daily realized volatility. While subsample averaging has been proposed and used in estimating RV, this paper is the first that uses subsample averaging for forecasting RV. The subsample averaging method we examine incorporates the high frequency data in different levels of systematic sampling. It first pools the high frequency data into several subsamples, then generates forecasts from each subsample, and then combines these forecasts. We find that in daily S&P 500 return realized volatility forecasts, subsample averaging generates better forecasts than those using only one subsample.

show abstract

“…Recent papers addressing this include [13,14]. Generally, the suggested volatility proxy is the realized volatility in some particular form, coupled with an MSFE evaluation criterion.…”

Section: Resultsmentioning

confidence: 99%

Forecasting Realized Volatility Using Subsample Averaging

Huang¹,

Lee²

2013

OJS

View full text Add to dashboard Cite

show abstract

“…QLIKE function shares robustness on ranking the models with respect to an unbiased estimator of the unkown conditional variance, see for example Patton (2011) . Even if metric criteria are important, it is useful to have statistical tests that assess if the difference between loss functions of two competing models is significant or not.…”

Section: Data and Methodsologymentioning

confidence: 99%

A Family of Stochastic Unit GARCH Models

Konte

2012

IJEF

View full text Add to dashboard Cite

A class of Asymmetric GARCH models is presented. It shares the same unconditional variance and volatility forecast formula as the standard GARCH(P,Q) model under the assumption of a symmetric conditional distribution for innovations. use three models of this class to assess their ability to forecast S&P 500 market volatility and to make better decisions for the purpose of risk management and investment. Subsequently, a comparison is made with respect to competing models (GARCH, EGARCH, GJR). It was found that for the in-sample evaluation, the best model is obtained from the Stochastic Unit GARCH (SUGARCH) model where leverage effects are introduced through the GARCH (i.e 1  ) parameter. For the out-of-sample evaluation (QLIKE loss function), it is better to use the SUGARCH class where the asymmetry appears on the ARCH (i.e 1  )parameter.

show abstract

“…First, the use of imperfect volatility proxy could lead to inconsistent or undesirable outcomes of standard methods for comparing volatility estimators (see, e.g., McMillan & Speight (2004); Hansen & Lunde (2006); Patton (2011)). For example, as suggested by Andersen and Bollerslev (1998), the most common proxy for 'true volatility', i.e., squared daily returns, may be a 'noisy' or imprecise estimator and thus, as an alternative, the volatility can be calculated by summing the squared intraday returns.…”

Section: Literature Reviewmentioning

confidence: 99%

Relative performance of VIXC vs. GARCH in predicting realised volatility changes

Jung

2016

Investment Analysts Journal

View full text Add to dashboard Cite

This paper examines the forecasting power of the volatility index of Canada (VIXC) and GARCH-family volatility. Specifically, this paper is motivated by an enquiry into how well volatility estimators predict volatility changes. To this end, we use a daily series of VIXC from 1 October 2009 through 30 April 2015. To estimate out-of-sample parameters for GARCH volatilities, a series of daily returns of the TSX60 since 29 November 2002 is used roll-forwardly. Then we run the forecasting regressions for a full-sample and five subsamples grouped by the daily percentage change in realised volatility, and compare R 2 and the loss functions to assess the forecasting power of volatility estimators. Additionally, this paper proposes a new measurement called mean-directional-error (MDE), which can comprehensively evaluate the forecast error in both direction and size of movement. The key findings of this paper can be summarised as follows: even though the information content of VIXC is highest on average and intensifies when volatility falls high, the directional accuracy measured by MDE is worst for VIXC in all subsamples. GJR-GARCH (1,1) achieves the highest accuracy judged by the traditional loss functions. Focusing on the directional accuracy, GARCH (1,1) demonstrates the best predictability. ARTICLE HISTORY

show abstract

Volatility forecast comparison using imperfect volatility proxies

Cited by 955 publications

References 52 publications

Forecasting Realized Volatility Using Subsample Averaging

Forecasting Realized Volatility Using Subsample Averaging

A Family of Stochastic Unit GARCH Models

Relative performance of VIXC vs. GARCH in predicting realised volatility changes

Contact Info

Product

Resources

About