Markets crash, and so does a paper explaining why

Markets undergo flash crashes — when stocks or bonds rapidly nosedive in value and then just as rapidly recover — every day. On May 6, 2010, for example, the entire equity market flamed out and then nearly recovered its value all in the matter of hours.

Economic papers can do the same, apparently. Take the recent withdrawal of an paper from the Journal of Financial Markets:

This article has been withdrawn at the request of the authors and editor. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at http://www.elsevier.com/locate/withdrawalpolicy.

Here’s the background: Puzzled economists and policy wonks sorted through equity market data and a trio of researchers came up with a theory to help predict why the May 6, 2010 flash crash occurred. The theory, published in the Journal of Trading, says it’s not commonly-suspected insiders who drive flash crashes, but a concept the trio call “Volume-Synchronized Probability of INformed trading” (VPIN). The VPIN measures the dynamics of buys versus sells in a way that the authors conclude can help predict these mini-crashes.

The two economists who published the Journal of Financial Markets paper — still available here — disagreed with the assessment. But editors at the journal  initiated the withdrawal after receiving complaints about the paper, according to study co-author Torben G. Andersen, a finance professor and director of the International Business & Markets Program and Research Center at the Kellogg School of Management, Northwestern University:

There is (perhaps not surprisingly) outside opposition to aspects of the paper and the editors would like “enhanced evidence” included before the article is published.

Andersen said he and his co-author weren’t aware of the nuances of a withdrawal and a retraction — a distinction Elsevier makes but which we think lacks a difference— nor were they aware of the withdrawal notice itself:

After your inquiry, we contacted the editors, and they promised to contact Elsevier in order to change the wording surrounding the ‘withdrawal’ notice. Specifically, the editor promised to forward the following statement to Elsevier: “the paper has been withdrawn pending a revision and an updated version will be available soon.”

The bottom line is that we were not informed about the language or type of notification that would appear on the Elsevier web-page. But we did agree to augment the evidence in the paper, so we were aware that the “accepted” paper would not appear in print as written. There is an implicit agreement that an updated version will be published, however. We will be extremely upset if this does not happen in the near future, but we currently have no reason to suspect “foul play” on that dimension. We shall see …