Talk:Granger causality

Statistics Low‑importance

	This article is within the scope of WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.StatisticsWikipedia:WikiProject StatisticsTemplate:WikiProject StatisticsStatistics articles
Low	This article has been rated as Low-importance on the importance scale.

Economics Low‑importance

	Business and economics portal This article is within the scope of WikiProject Economics, a collaborative effort to improve the coverage of Economics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.EconomicsWikipedia:WikiProject EconomicsTemplate:WikiProject EconomicsEconomics articles
Low	This article has been rated as Low-importance on the project's importance scale.

I just wanted to add the original paper here:

Granger, C.W.J., 1969. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37 (3), 424– 438. —Preceding unsigned comment added by 62.143.180.152 (talk) 20:01, 7 March 2008 (UTC)[reply]

What can I read about this please? Any easy to understand explanations?

You can google "Granger Causality" and you will get different articles and Powerpoint presentations from either researchers or university professors. Some of those are easier to understand than others. It is easier to explain Granger Causality by walking someone through an example as I have by answering your second question just below.

Does this mean that eg. forecasts could be made of sectors in the stock market, whereas previously you could not tell if these lagged correlations were just coincidence or not?

You could use Granger Causality to test whether certain independent variables have an impact on stock market sectors. For instances, testing that Oil prices "Granger cause" energy companies stock prices you would do the following:

You would first build a base model with just lagged stock prices (last month for instance) as the independent variable to predict current stock price.
You build a test model by adding lagged oil prices as a second independent variable.
You compare the squared residuals of the base model and the test model and you test whether they are different using the F test. You could just as well use the unpaired t test or its nonparametric counterpart the Mann-Whitney test. Using either of these tests, you will get a p value giving you the probability that the tested independent variable truly provides you additional explanatory information.

Sympa 19:15, 18 September 2006 (UTC)Sympa[reply]

mathematical section is flawed

Because it makes reference to an index "p" that does not appear in the formulae.

Moscholar (talk) 08:56, 29 August 2013 (UTC)[reply]

Deletion of reference and discussion of Grassmann (2020)

On 4/21/2023 Wikivdt deleted a reference and a brief (two-sentence) discussion of a paper by Grassmann et al challenging the validity of Granger's methodology:

http://en.wiki.x.io/w/index.php?title=Granger_causality&diff=1150523783&oldid=1122114761

I don't know whether Grassmann's criticism is valid, but her paper was published in a respected journal, and it seems pertinent to this article. Is there a good reason that the material was deleted? Should it be restored? NCdave (talk) 13:12, 23 June 2023 (UTC)[reply]

Seeing no objection, I'm going to revert the deletion, but tone down the criticism. NCdave (talk) 22:50, 3 April 2024 (UTC)[reply]