Athens University Economics and Business (AUEB)
An econometrics analysis of high-frequency financial data
Abstract
dc:descriptionWe present and compare observation driven and parameter driven models for predicting integer price changes of high-frequency financial data. We explore Bayesian inference via Markov chain Monte Carlo (MCMC) and sequential Monte Carlo (SMC) for the observation driven model activity-direction-size (ADS), introduced by Rydberg and Shephard [1998a,2003]. We extend the ADS model by proposing a parameter driven model and use a Bernoulli generalized linear model (GLM) with a latent process in the mean. We propose a new decomposition model that uses trade intervals and is applied on data that allow three possible tick movements: one tick up price change, one tick down price change, or no price change. We model each component sequentially using a Binomial generalized linear autoregressive moving average (GLARMA) model, as well as a GLM with a latent process in the mean. We perform a simulation study to investigate the effectiveness of the proposed parameter driven models using different algorithms within a Bayesian framework. We illustrate the analysis by modelling the transaction-by-transaction data of E-mini Standard and Poor’s (S&P) 500 index futures contract traded on the Chicago Mercantile Exchange's Globex platform between May 16th 2011 and May 24th 2011. In order to assess the predictive performance, we compare the mean square error (MSE) and mean absolute error (MAE) criterion, as well as four scalar performance measures, namely, accuracy, sensitivity, precision and specificity derived from the confusion matrix.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Lamprinakou, Fiori
- Λαμπρινάκου, Φιόρη
Subjects
dc:subject × 14Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/50906