No label defined (Q4955684)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1449649
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 1449649 |
Statements
21 May 2000
0 references
GARCH processes
0 references
statistical process control
0 references
control charts
0 references
CUSUM
0 references
EWMA
0 references
0.8478606939315796
0 references
0.8049783706665039
0 references
0.7903989553451538
0 references
0.7880174517631531
0 references
0.7840938568115234
0 references
The authors consider a problem of Control Charts algorithms adaptation to the change-point detection in GARCH processes. I.e., the data are considered of the form \( X_t=Y_t+a1\{t>s\}, \) where \(s\) is the change-point to be detected, \(Y_t=\varepsilon_t\sqrt{h_t}\) is a GARCH process, and \(h_t=\alpha_0+\sum_{i=1}^q\alpha_i Y^2_{t-i}+\sum_{j=1}^p\beta_j h_{t-j},\) \(\varepsilon_j\) are i.i.d. with \(E\varepsilon_j=0\) and \(E\varepsilon_j^2=1\). Modified Shewhart, EWMA and CUSUM charts are considered for the process \(Y_t\) and for the residuals process \(\hat\varepsilon_t\). Note, that by classical definition of the residuals \(Y_t-E (Y_t |Y_{t-1},Y_{t-2},\dots)=Y_t\) since \(E (Y_t |Y_{t-1},Y_{t-2},\dots)=0\) for GARCH processes. So the authors propose to use two new versions of the residuals: \[ \hat\varepsilon_t= X_t(\alpha_0+\sum_{i=1}^q\alpha_i X_{t-i})^{-1/2} \] (for \(p=0\), the definition for \(p>0\) is more complicated), resp., \[ \hat\varepsilon_t=X_t^2-E(Y_t^2 |Y_{t-1},Y_{t-2},\dots). \] The average run length for such control charts is investigated via simulations. The charts are applied to the analysis of the Lufthansa and AEG stocks. For the entire collection see [Zbl 0929.00046].
0 references