Autoregressive Conditional Duration Models in Finance: A Survey of the Theoretical and Empirical Literature

Pacurar, Maria

doi:10.1111/j.1467-6419.2007.00547.x

Cited by 136 publications

(80 citation statements)

References 148 publications

(304 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…empirical regularities exhibited by a wide range of financial time series, are commonly used to validate ABM designs and parameterizations. A wide range of stylized facts, both for high-frequency and aggregated data, are described in the literature, e.g., Chen et al (2012), Cont (2001), Daniel (2006, Pacurar (2008). The class of intraday stylized facts can be associated to transaction Order placement Parameters book depth levels N = 3 OBI base µ = 2 size penalty exponent η = 0.8 risk function weights α 0 = 0.1, α 1 = 0.5, α 2 = 0.25 volatility-threshold penalty weight β = 2.5…”

Section: Experimental Setup Results and Discussionmentioning

confidence: 99%

Microstructure-based order placement in a continuous double auction agent based model

Mandes

2015

View full text Add to dashboard Cite

Abstract. This contribution proposes a novel order placement strategy which can be used for simulating continuous double auction financial markets, within an agent-based model framework. The order placement decision is given by an optimization problem which minimizes the risk adjusted execution cost, taking into consideration relevant market microstructure factors and intrinsic agent characteristics. This order submission process is more realistic than has been done previously and contributes to a higher fidelity of the intraday market dynamics. The results show that, as opposed to random submission strategies, highfrequency stylized facts such as the concave shape of the market price impact function and the power-law decaying relative price distribution of off-spread limit orders are replicated. Therefore, the resulting model can be used as a realistic test environment for high-frequency trading strategies, in the context of the current, heated debate over the impact of high-frequency trading. Not only the impact of individual trading strategies can be analyzed, but also the interdependencies and the global emergent behavior of multiple coexistent strategies. Moreover, innovative regulatory policies, which have not been tested yet under real market conditions, could be inspected.

show abstract

Section: Experimental Setup Results and Discussionmentioning

confidence: 99%

Microstructure-based order placement in a continuous double auction agent based model

Mandes

2015

View full text Add to dashboard Cite

show abstract

“…The ACD model is the counterpart of GARCH models for dealing with trade duration (TD) data and it is used to capture the clustering structure, which conveys meaningful information, observed in high frequency financial data; see Duchesne and Pacurar (2008), Liu and Heyde (2008) and Pacurar (2008). TD data possess a number of unique characteristics such as: an irregular nature in which they are collected; a large number of observations or cases; a diurnal pattern, where activity is higher at the beginning and closing than in the middle of the trading day; asymmetry and an inverse bathtub shaped hazard rate (HR); see Bhatti (2010) and Leiva et al (2014b).…”

Section: Introductionmentioning

confidence: 99%

Birnbaum–Saunders autoregressive conditional duration models applied to high-frequency financial data

et al. 2017

View full text Add to dashboard Cite

Modern financial markets now record the precise time of each stock trade, along with price and volume, with the aim of analysing the structure of the times between trading events -leading to a big data problem. In this paper, we propose and compare two Birnbaum-Saunders autoregressive conditional duration models specified in terms of time-varying conditional median and mean durations. These models provide a novel alternative to the existing autoregressive conditional duration models due to their flexibility and ease of estimation. Influence diagnostic tools are developed to allow goodness-of-fit assessment and to detect departures from assumptions, including the presence of outliers and influential cases. Both global and local influence tools are considered based on the parameter estimates under different perturbation schemes. A thorough Monte Carlo study is presented to evaluate the performance of the maximum likelihood estimators, and the forecasting ability of the models is assessed using the traditional and density forecast evaluation techniques. The Monte Carlo study suggests that the parameter estimators are asymptotically unbiased, consistent and normally distributed. Finally, a full analysis of a real-world financial transaction data set, from the German DAX in 2016, is presented to illustrate the proposed approach and to compare the fitting and forecasting performances with existing models in the literature. One case related to the duration time is identified as potentially influential, but its removal does not change resulting inferences demonstrating the robustness of the proposed approach. Fitting and forecasting performances favor the proposed models and, in particular, the median-based approach gives additional protection against outliers, as expected.

show abstract

“…The availability of trade-by-trade (commonly known as ultra high-frequency (UHF)) data on transactions, in the recent decade, has revolutionised data processing and statistical modelling techniques in finance (McCulloch & Tsay, 2001;Tsay, 2005;Liesenfeld, Nolte, & Pohlmeier, 2006;Pacurar, 2008). The unique characteristics of UHF data has introduced new theoretical and computational challenges to both statistical and financial studies (Liesenfeld, Nolte, & Pohlmeier, 2006;Pacurar, 2008;Tsay, 2005).…”

Section: Applicationmentioning

confidence: 99%

“…The unique characteristics of UHF data has introduced new theoretical and computational challenges to both statistical and financial studies (Liesenfeld, Nolte, & Pohlmeier, 2006;Pacurar, 2008;Tsay, 2005). Such data consist of (i) discrete-valued observations, as the price change in consecutive transactions is in a multiple of tick size, where one tick is defined as the minimum amount by which the price of the market can change and (ii) unequally spaced time intervals; see Tsay (2005, Chapter 5) and Liesenfeld, Nolte, and Pohlmeier (2006) and references therein for further details on analysing UHF data sets.…”

Section: Applicationmentioning

confidence: 99%

Bayesian Modelling of Integer Data Using the Generalised Poisson Difference Distribution

Shahtahmassebi¹,

Moyeed²

2014

IJSP

View full text Add to dashboard Cite

Integer-valued random variables arising from the difference of two discrete variables can be seen frequently in various applications. In this paper, we obtain the distribution and derive the properties of the difference of two generalised Poisson variables with unequal parameters. This distribution is adopted to model a set of ultra high frequency (UHF) data relating to FTSE100 index futures using covariates. The unique characteristics of UHF data have introduced new theoretical and computational challenges to both statistical and financial studies. Such data consist of discrete-valued observations and unequally spaced time intervals. We also extend the model to its zero inflated version in order to capture the excess of zeros in the given data set. The analysis is carried out in a Bayesian framework using Markov Chain Monte Carlo methods. Various model diagnostics and model comparisons were undertaken which showed that index changes were explained well by the fitted model.

show abstract

Autoregressive Conditional Duration Models in Finance: A Survey of the Theoretical and Empirical Literature

Cited by 136 publications

References 148 publications

Microstructure-based order placement in a continuous double auction agent based model

Microstructure-based order placement in a continuous double auction agent based model

Birnbaum–Saunders autoregressive conditional duration models applied to high-frequency financial data

Bayesian Modelling of Integer Data Using the Generalised Poisson Difference Distribution

Contact Info

Product

Resources

About