Does systematic sampling preserve granger causality with an application to high frequency financial data?

Gulasekaran Rajaguru, Tilak Abeysinghe, Michael O'Neill

Research output: Contribution to conferencePresentationResearchpeer-review

30 Downloads (Pure)

Abstract

In applied econometric literature, the causal inferences are often made based on temporally aggregated or systematically sampled data. A number of theoretical studies have pointed out that temporal aggregation has distorting effects on causal inference and systematic sampling of stationary variables preserves the direction of causality. This paper examines the issue in detail by plugging in theoretical cross covariances into the limiting values of least squares estimates in a VAR framework. The asymptotic distributions of the estimates of systematically sampled process are expressed in terms of the cross covariances of the disaggregated process. An extensive Monte Carlo study is conducted to examine small sample results. Quite contrary to the stationary case, this paper shows that systematic sampling of integrated series may induce spurious causality. In particular, systematic sampling induces spurious bi-directional Granger causality among the variables if the uni-directional causality runs from a non-stationary series to either a stationary or a non-stationary series. On the other hand, systematic sampling preserves the uni-directional causality among the variables if the uni-directional causality runs from a stationary series to either a stationary or a non-stationary series. It is observed that in general the most distorting causal inferences are likely at low levels of sampling intervals where the order of sampling-span just exceeds the actual causal lag. At high levels of systematic sampling, causal information concentrates in contemporaneous correlations. An empirical exercise illustrates the relative usefulness of the results further.
Original languageEnglish
Publication statusPublished - 2017
EventAustralian Conference of Economists: Economics for Better Lives - Sydney, Sydney, Australia
Duration: 19 Jul 201721 Jul 2017
Conference number: 46th
http://ace2017.org.au/ (ACE 2017 Home Page)
https://vimeo.com/250391912 (ACE Conference 2017 [Promotional Video])

Conference

ConferenceAustralian Conference of Economists: Economics for Better Lives
Abbreviated titleACE 2017
CountryAustralia
CitySydney
Period19/07/1721/07/17
OtherThe theme this year is “Economics for Better Lives” and sessions will consider issues such as inequality and poverty, financial regulation, taxation reform, education reform, the economics of mental health, utility regulation, transport economics, infrastructure, regional trade agreements, international migration flows, international aid and development, the role of behavioral economics the economics of populism, and macroeconomics in a low interest rate world. We are planning a special session on professional ethics for economists, and the future of the profession.

This is the premier Australian conference for economists in universities, public policy, business and finance. Each year we welcome many overseas economists including this year’s keynote speakers Carol Graham and Kip Viscusi. Generations of young economists and students have presented papers and connected with others at the conference, launching their careers. Our policy panels with distinguished academic and government economists have introduced new ideas and shaped public policy.
Internet address

Fingerprint

Financial data
Granger causality
Sampling
Causality
Causal inference
Integrated
Usefulness
Lag
Asymptotic distribution
Small sample
Exercise
Least squares
Monte Carlo study
Contemporaneous correlation
Temporal aggregation
Applied econometrics

Cite this

Rajaguru, G., Abeysinghe, T., & O'Neill, M. (2017). Does systematic sampling preserve granger causality with an application to high frequency financial data?. Australian Conference of Economists: Economics for Better Lives, Sydney, Australia.
Rajaguru, Gulasekaran ; Abeysinghe, Tilak ; O'Neill, Michael. / Does systematic sampling preserve granger causality with an application to high frequency financial data?. Australian Conference of Economists: Economics for Better Lives, Sydney, Australia.
@conference{16f99087c3c84057b62d156a935280b0,
title = "Does systematic sampling preserve granger causality with an application to high frequency financial data?",
abstract = "In applied econometric literature, the causal inferences are often made based on temporally aggregated or systematically sampled data. A number of theoretical studies have pointed out that temporal aggregation has distorting effects on causal inference and systematic sampling of stationary variables preserves the direction of causality. This paper examines the issue in detail by plugging in theoretical cross covariances into the limiting values of least squares estimates in a VAR framework. The asymptotic distributions of the estimates of systematically sampled process are expressed in terms of the cross covariances of the disaggregated process. An extensive Monte Carlo study is conducted to examine small sample results. Quite contrary to the stationary case, this paper shows that systematic sampling of integrated series may induce spurious causality. In particular, systematic sampling induces spurious bi-directional Granger causality among the variables if the uni-directional causality runs from a non-stationary series to either a stationary or a non-stationary series. On the other hand, systematic sampling preserves the uni-directional causality among the variables if the uni-directional causality runs from a stationary series to either a stationary or a non-stationary series. It is observed that in general the most distorting causal inferences are likely at low levels of sampling intervals where the order of sampling-span just exceeds the actual causal lag. At high levels of systematic sampling, causal information concentrates in contemporaneous correlations. An empirical exercise illustrates the relative usefulness of the results further.",
author = "Gulasekaran Rajaguru and Tilak Abeysinghe and Michael O'Neill",
year = "2017",
language = "English",
note = "Australian Conference of Economists: Economics for Better Lives, ACE 2017 ; Conference date: 19-07-2017 Through 21-07-2017",
url = "http://ace2017.org.au/, https://vimeo.com/250391912",

}

Rajaguru, G, Abeysinghe, T & O'Neill, M 2017, 'Does systematic sampling preserve granger causality with an application to high frequency financial data?' Australian Conference of Economists: Economics for Better Lives, Sydney, Australia, 19/07/17 - 21/07/17, .

Does systematic sampling preserve granger causality with an application to high frequency financial data? / Rajaguru, Gulasekaran; Abeysinghe, Tilak; O'Neill, Michael.

2017. Australian Conference of Economists: Economics for Better Lives, Sydney, Australia.

Research output: Contribution to conferencePresentationResearchpeer-review

TY - CONF

T1 - Does systematic sampling preserve granger causality with an application to high frequency financial data?

AU - Rajaguru, Gulasekaran

AU - Abeysinghe, Tilak

AU - O'Neill, Michael

PY - 2017

Y1 - 2017

N2 - In applied econometric literature, the causal inferences are often made based on temporally aggregated or systematically sampled data. A number of theoretical studies have pointed out that temporal aggregation has distorting effects on causal inference and systematic sampling of stationary variables preserves the direction of causality. This paper examines the issue in detail by plugging in theoretical cross covariances into the limiting values of least squares estimates in a VAR framework. The asymptotic distributions of the estimates of systematically sampled process are expressed in terms of the cross covariances of the disaggregated process. An extensive Monte Carlo study is conducted to examine small sample results. Quite contrary to the stationary case, this paper shows that systematic sampling of integrated series may induce spurious causality. In particular, systematic sampling induces spurious bi-directional Granger causality among the variables if the uni-directional causality runs from a non-stationary series to either a stationary or a non-stationary series. On the other hand, systematic sampling preserves the uni-directional causality among the variables if the uni-directional causality runs from a stationary series to either a stationary or a non-stationary series. It is observed that in general the most distorting causal inferences are likely at low levels of sampling intervals where the order of sampling-span just exceeds the actual causal lag. At high levels of systematic sampling, causal information concentrates in contemporaneous correlations. An empirical exercise illustrates the relative usefulness of the results further.

AB - In applied econometric literature, the causal inferences are often made based on temporally aggregated or systematically sampled data. A number of theoretical studies have pointed out that temporal aggregation has distorting effects on causal inference and systematic sampling of stationary variables preserves the direction of causality. This paper examines the issue in detail by plugging in theoretical cross covariances into the limiting values of least squares estimates in a VAR framework. The asymptotic distributions of the estimates of systematically sampled process are expressed in terms of the cross covariances of the disaggregated process. An extensive Monte Carlo study is conducted to examine small sample results. Quite contrary to the stationary case, this paper shows that systematic sampling of integrated series may induce spurious causality. In particular, systematic sampling induces spurious bi-directional Granger causality among the variables if the uni-directional causality runs from a non-stationary series to either a stationary or a non-stationary series. On the other hand, systematic sampling preserves the uni-directional causality among the variables if the uni-directional causality runs from a stationary series to either a stationary or a non-stationary series. It is observed that in general the most distorting causal inferences are likely at low levels of sampling intervals where the order of sampling-span just exceeds the actual causal lag. At high levels of systematic sampling, causal information concentrates in contemporaneous correlations. An empirical exercise illustrates the relative usefulness of the results further.

UR - http://ace2017.org.au/wp-content/uploads/2015/12/ACE-2017-Program-1.pdf

UR - http://ace2017.org.au/speaker-abstractsfull-papers/gulasekaran-rajaguru/

UR - https://vimeo.com/250391912 (

M3 - Presentation

ER -

Rajaguru G, Abeysinghe T, O'Neill M. Does systematic sampling preserve granger causality with an application to high frequency financial data?. 2017. Australian Conference of Economists: Economics for Better Lives, Sydney, Australia.