Autoregressive conditional heteroskedasticity

In econometrics, the autoregressive conditional heteroskedasticity (ARCH) model is a statistical model for time series data that describes the variance of the current error term or innovation as a function of the actual sizes of the previous time periods' error terms;[1] often the variance is related to the squares of the previous innovations. The ARCH model is appropriate when the error variance in a time series follows an autoregressive (AR) model; if an autoregressive moving average (ARMA) model is assumed for the error variance, the model is a generalized autoregressive conditional heteroskedasticity (GARCH) model.[2]

ARCH models are commonly employed in modeling financial time series that exhibit time-varying volatility and volatility clustering, i.e. periods of swings interspersed with periods of relative calm. ARCH-type models are sometimes considered to be in the family of stochastic volatility models, although this is strictly incorrect since at time t the volatility is completely pre-determined (deterministic) given previous values.[3]

Model specification edit

To model a time series using an ARCH process, let  denote the error terms (return residuals, with respect to a mean process), i.e. the series terms. These   are split into a stochastic piece   and a time-dependent standard deviation   characterizing the typical size of the terms so that

 

The random variable   is a strong white noise process. The series   is modeled by

 ,
where   and  .

An ARCH(q) model can be estimated using ordinary least squares. A method for testing whether the residuals   exhibit time-varying heteroskedasticity using the Lagrange multiplier test was proposed by Engle (1982). This procedure is as follows:

  1. Estimate the best fitting autoregressive model AR(q)  .
  2. Obtain the squares of the error   and regress them on a constant and q lagged values:
     
    where q is the length of ARCH lags.
  3. The null hypothesis is that, in the absence of ARCH components, we have   for all  . The alternative hypothesis is that, in the presence of ARCH components, at least one of the estimated   coefficients must be significant. In a sample of T residuals under the null hypothesis of no ARCH errors, the test statistic T'R² follows   distribution with q degrees of freedom, where   is the number of equations in the model which fits the residuals vs the lags (i.e.  ). If T'R² is greater than the Chi-square table value, we reject the null hypothesis and conclude there is an ARCH effect in the ARMA model. If T'R² is smaller than the Chi-square table value, we do not reject the null hypothesis.

GARCH edit

If an autoregressive moving average (ARMA) model is assumed for the error variance, the model is a generalized autoregressive conditional heteroskedasticity (GARCH) model.[2]

In that case, the GARCH (p, q) model (where p is the order of the GARCH terms   and q is the order of the ARCH terms   ), following the notation of the original paper, is given by

 

 

 

Generally, when testing for heteroskedasticity in econometric models, the best test is the White test. However, when dealing with time series data, this means to test for ARCH and GARCH errors.

Exponentially weighted moving average (EWMA) is an alternative model in a separate class of exponential smoothing models. As an alternative to GARCH modelling it has some attractive properties such as a greater weight upon more recent observations, but also drawbacks such as an arbitrary decay factor that introduces subjectivity into the estimation.

GARCH(p, q) model specification edit

The lag length p of a GARCH(p, q) process is established in three steps:

  1. Estimate the best fitting AR(q) model
     .
  2. Compute and plot the autocorrelations of   by
     
  3. The asymptotic, that is for large samples, standard deviation of   is  . Individual values that are larger than this indicate GARCH errors. To estimate the total number of lags, use the Ljung–Box test until the value of these are less than, say, 10% significant. The Ljung–Box Q-statistic follows   distribution with n degrees of freedom if the squared residuals   are uncorrelated. It is recommended to consider up to T/4 values of n. The null hypothesis states that there are no ARCH or GARCH errors. Rejecting the null thus means that such errors exist in the conditional variance.

NGARCH edit

NAGARCH edit

Nonlinear Asymmetric GARCH(1,1) (NAGARCH) is a model with the specification:[6][7]

 ,
where   and  , which ensures the non-negativity and stationarity of the variance process.

For stock returns, parameter   is usually estimated to be positive; in this case, it reflects a phenomenon commonly referred to as the "leverage effect", signifying that negative returns increase future volatility by a larger amount than positive returns of the same magnitude.[6][7]

This model should not be confused with the NARCH model, together with the NGARCH extension, introduced by Higgins and Bera in 1992.[8]

IGARCH edit

Integrated Generalized Autoregressive Conditional heteroskedasticity (IGARCH) is a restricted version of the GARCH model, where the persistent parameters sum up to one, and imports a unit root in the GARCH process.[9] The condition for this is

 .

EGARCH edit

The exponential generalized autoregressive conditional heteroskedastic (EGARCH) model by Nelson & Cao (1991) is another form of the GARCH model. Formally, an EGARCH(p,q):

 

where  ,   is the conditional variance,  ,  ,  ,   and   are coefficients.   may be a standard normal variable or come from a generalized error distribution. The formulation for   allows the sign and the magnitude of   to have separate effects on the volatility. This is particularly useful in an asset pricing context.[10][11]

Since   may be negative, there are no sign restrictions for the parameters.

GARCH-M edit

The GARCH-in-mean (GARCH-M) model adds a heteroskedasticity term into the mean equation. It has the specification:

 

The residual   is defined as:

 

QGARCH edit

The Quadratic GARCH (QGARCH) model by Sentana (1995) is used to model asymmetric effects of positive and negative shocks.

In the example of a GARCH(1,1) model, the residual process   is

 

where   is i.i.d. and

 

GJR-GARCH edit

Similar to QGARCH, the Glosten-Jagannathan-Runkle GARCH (GJR-GARCH) model by Glosten, Jagannathan and Runkle (1993) also models asymmetry in the ARCH process. The suggestion is to model   where   is i.i.d., and

 

where   if  , and   if  .

TGARCH model edit

The Threshold GARCH (TGARCH) model by Zakoian (1994) is similar to GJR GARCH. The specification is one on conditional standard deviation instead of conditional variance:

 

where   if  , and   if  . Likewise,   if  , and   if  .

fGARCH edit

Hentschel's fGARCH model,[12] also known as Family GARCH, is an omnibus model that nests a variety of other popular symmetric and asymmetric GARCH models including APARCH, GJR, AVGARCH, NGARCH, etc.

COGARCH edit

In 2004, Claudia Klüppelberg, Alexander Lindner and Ross Maller proposed a continuous-time generalization of the discrete-time GARCH(1,1) process. The idea is to start with the GARCH(1,1) model equations

 
 

and then to replace the strong white noise process   by the infinitesimal increments   of a Lévy process  , and the squared noise process   by the increments  , where

 

is the purely discontinuous part of the quadratic variation process of  . The result is the following system of stochastic differential equations:

 
 

where the positive parameters  ,   and   are determined by  ,   and  . Now given some initial condition  , the system above has a pathwise unique solution   which is then called the continuous-time GARCH (COGARCH) model.[13]

ZD-GARCH edit

Unlike GARCH model, the Zero-Drift GARCH (ZD-GARCH) model by Li, Zhang, Zhu and Ling (2018) [14] lets the drift term   in the first order GARCH model. The ZD-GARCH model is to model  , where   is i.i.d., and

 

The ZD-GARCH model does not require  , and hence it nests the Exponentially weighted moving average (EWMA) model in "RiskMetrics". Since the drift term  , the ZD-GARCH model is always non-stationary, and its statistical inference methods are quite different from those for the classical GARCH model. Based on the historical data, the parameters   and   can be estimated by the generalized QMLE method.

Spatial GARCH edit

Spatial GARCH processes by Otto, Schmid and Garthoff (2018) [15] are considered as the spatial equivalent to the temporal generalized autoregressive conditional heteroscedasticity (GARCH) models. In contrast to the temporal ARCH model, in which the distribution is known given the full information set for the prior periods, the distribution is not straightforward in the spatial and spatiotemporal setting due to the interdependence between neighboring spatial locations. The spatial model is given by   and

 

where   denotes the  -th spatial location and   refers to the  -th entry of a spatial weight matrix and   for  . The spatial weight matrix defines which locations are considered to be adjacent.

Gaussian process-driven GARCH edit

In a different vein, the machine learning community has proposed the use of Gaussian process regression models to obtain a GARCH scheme.[16] This results in a nonparametric modelling scheme, which allows for: (i) advanced robustness to overfitting, since the model marginalises over its parameters to perform inference, under a Bayesian inference rationale; and (ii) capturing highly-nonlinear dependencies without increasing model complexity.[citation needed]

References edit

  1. ^ Engle, Robert F. (1982). "Autoregressive Conditional Heteroskedasticity with Estimates of the Variance of United Kingdom Inflation". Econometrica. 50 (4): 987–1007. doi:10.2307/1912773. JSTOR 1912773.
  2. ^ a b Bollerslev, Tim (1986). "Generalized Autoregressive Conditional Heteroskedasticity". Journal of Econometrics. 31 (3): 307–327. CiteSeerX 10.1.1.468.2892. doi:10.1016/0304-4076(86)90063-1. S2CID 8797625.
  3. ^ Brooks, Chris (2014). Introductory Econometrics for Finance (3rd ed.). Cambridge: Cambridge University Press. p. 461. ISBN 9781107661455.
  4. ^ Lanne, Markku; Saikkonen, Pentti (July 2005). "Non-linear GARCH models for highly persistent volatility" (PDF). The Econometrics Journal. 8 (2): 251–276. doi:10.1111/j.1368-423X.2005.00163.x. JSTOR 23113641. S2CID 15252964.
  5. ^ Bollerslev, Tim; Russell, Jeffrey; Watson, Mark (May 2010). "Chapter 8: Glossary to ARCH (GARCH)" (PDF). Volatility and Time Series Econometrics: Essays in Honor of Robert Engle (1st ed.). Oxford: Oxford University Press. pp. 137–163. ISBN 9780199549498. Retrieved 27 October 2017.
  6. ^ a b Engle, Robert F.; Ng, Victor K. (1993). "Measuring and testing the impact of news on volatility" (PDF). Journal of Finance. 48 (5): 1749–1778. doi:10.1111/j.1540-6261.1993.tb05127.x. SSRN 262096. It is not yet clear in the finance literature that the asymmetric properties of variances are due to changing leverage. The name "leverage effect" is used simply because it is popular among researchers when referring to such a phenomenon.
  7. ^ a b Posedel, Petra (2006). "Analysis Of The Exchange Rate And Pricing Foreign Currency Options On The Croatian Market: The Ngarch Model As An Alternative To The Black Scholes Model" (PDF). Financial Theory and Practice. 30 (4): 347–368. Special attention to the model is given by the parameter of asymmetry [theta (θ)] which describes the correlation between returns and variance.6 ...
    6 In the case of analyzing stock returns, the positive value of [theta] reflects the empirically well known leverage effect indicating that a downward movement in the price of a stock causes more of an increase in variance more than a same value downward movement in the price of a stock, meaning that returns and variance are negatively correlated
  8. ^ Higgins, M.L; Bera, A.K (1992). "A Class of Nonlinear Arch Models". International Economic Review. 33 (1): 137–158. doi:10.2307/2526988. JSTOR 2526988.
  9. ^ Caporale, Guglielmo Maria; Pittis, Nikitas; Spagnolo, Nicola (October 2003). "IGARCH models and structural breaks". Applied Economics Letters. 10 (12): 765–768. doi:10.1080/1350485032000138403. ISSN 1350-4851.
  10. ^ St. Pierre, Eilleen F. (1998). "Estimating EGARCH-M Models: Science or Art". The Quarterly Review of Economics and Finance. 38 (2): 167–180. doi:10.1016/S1062-9769(99)80110-0.
  11. ^ Chatterjee, Swarn; Hubble, Amy (2016). "Day-Of-The-Whieek Effect In Us Biotechnology Stocks—Do Policy Changes And Economic Cycles Matter?". Annals of Financial Economics. 11 (2): 1–17. doi:10.1142/S2010495216500081.
  12. ^ Hentschel, Ludger (1995). "All in the family Nesting symmetric and asymmetric GARCH models". Journal of Financial Economics. 39 (1): 71–104. CiteSeerX 10.1.1.557.8941. doi:10.1016/0304-405X(94)00821-H.
  13. ^ Klüppelberg, C.; Lindner, A.; Maller, R. (2004). "A continuous-time GARCH process driven by a Lévy process: stationarity and second-order behaviour". Journal of Applied Probability. 41 (3): 601–622. doi:10.1239/jap/1091543413. hdl:10419/31047. S2CID 17943198.
  14. ^ Li, D.; Zhang, X.; Zhu, K.; Ling, S. (2018). "The ZD-GARCH model: A new way to study heteroscedasticity" (PDF). Journal of Econometrics. 202 (1): 1–17. doi:10.1016/j.jeconom.2017.09.003.
  15. ^ Otto, P.; Schmid, W.; Garthoff, R. (2018). "Generalised spatial and spatiotemporal autoregressive conditional heteroscedasticity". Spatial Statistics. 26 (1): 125–145. arXiv:1609.00711. doi:10.1016/j.spasta.2018.07.005. S2CID 88521485.
  16. ^ Platanios, E.; Chatzis, S. (2014). "Gaussian process-mixture conditional heteroscedasticity". IEEE Transactions on Pattern Analysis and Machine Intelligence. 36 (5): 889–900. arXiv:1211.4410. doi:10.1109/TPAMI.2013.183. PMID 26353224. S2CID 10424638.

Further reading edit