Advanced

Long Memory of Financial Time Series and Hidden Markov Models with Time-Varying Parameters

Nystrup, Peter; Madsen, Henrik and Lindström, Erik LU (2016) In Journal of Forecasting
Abstract

Hidden Markov models are often used to model daily returns and to infer the hidden state of financial markets. Previous studies have found that the estimated models change over time, but the implications of the time-varying behavior have not been thoroughly examined. This paper presents an adaptive estimation approach that allows for the parameters of the estimated models to be time varying. It is shown that a two-state Gaussian hidden Markov model with time-varying parameters is able to reproduce the long memory of squared daily returns that was previously believed to be the most difficult fact to reproduce with a hidden Markov model. Capturing the time-varying behavior of the parameters also leads to improved one-step density... (More)

Hidden Markov models are often used to model daily returns and to infer the hidden state of financial markets. Previous studies have found that the estimated models change over time, but the implications of the time-varying behavior have not been thoroughly examined. This paper presents an adaptive estimation approach that allows for the parameters of the estimated models to be time varying. It is shown that a two-state Gaussian hidden Markov model with time-varying parameters is able to reproduce the long memory of squared daily returns that was previously believed to be the most difficult fact to reproduce with a hidden Markov model. Capturing the time-varying behavior of the parameters also leads to improved one-step density forecasts. Finally, it is shown that the forecasting performance of the estimated models can be further improved using local smoothing to forecast the parameter variations.

(Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
epub
subject
keywords
Adaptive estimation, Daily returns, Hidden Markov models, Long memory, Time-varying parameters
in
Journal of Forecasting
publisher
Wiley Online Library
external identifiers
  • scopus:84987602233
ISSN
0277-6693
DOI
10.1002/for.2447
language
English
LU publication?
yes
id
6168f868-bddf-48e3-ae40-2ee4cc4e4a01
date added to LUP
2016-09-28 09:54:18
date last changed
2017-02-22 15:27:45
@article{6168f868-bddf-48e3-ae40-2ee4cc4e4a01,
  abstract     = {<p>Hidden Markov models are often used to model daily returns and to infer the hidden state of financial markets. Previous studies have found that the estimated models change over time, but the implications of the time-varying behavior have not been thoroughly examined. This paper presents an adaptive estimation approach that allows for the parameters of the estimated models to be time varying. It is shown that a two-state Gaussian hidden Markov model with time-varying parameters is able to reproduce the long memory of squared daily returns that was previously believed to be the most difficult fact to reproduce with a hidden Markov model. Capturing the time-varying behavior of the parameters also leads to improved one-step density forecasts. Finally, it is shown that the forecasting performance of the estimated models can be further improved using local smoothing to forecast the parameter variations.</p>},
  author       = {Nystrup, Peter and Madsen, Henrik and Lindström, Erik},
  issn         = {0277-6693},
  keyword      = {Adaptive estimation,Daily returns,Hidden Markov models,Long memory,Time-varying parameters},
  language     = {eng},
  month        = {09},
  publisher    = {Wiley Online Library},
  series       = {Journal of Forecasting},
  title        = {Long Memory of Financial Time Series and Hidden Markov Models with Time-Varying Parameters},
  url          = {http://dx.doi.org/10.1002/for.2447},
  year         = {2016},
}