A comparison between qualitative and quantitative histories: the example of the efficient market hypothesis

ABSTRACT

This paper uses the example of the history of the efficient market hypothesis (EMH) and citation analysis in order to investigate some differences between qualitative history and a quantitative history. The history of the EMH provides a telling example of the way quantitative analyses can supply different perspectives on the qualitative history of this hypothesis or complement it. For instance, since the EMH was proposed, several criticisms emerged. In addition, the definition and the scope of this hypothesis have been modified several times. Although the qualitative history of the EMH refers to these criticisms and these alternative definitions and scopes, qualitative tools cannot provide a clear measure of the impact of these criticisms and these modifications among economists. By studying the dissemination of the EMH, its major criticisms, and the answers economists provided, citation analysis sheds a different light on the history of the EMH.

KEYWORDS:

• Efficient market hypothesis
• history of finance
• qualitative history
• quantitative history

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Franck Jovanovic is a Full Professor in Economics and Finance at the School of Administrative Sciences of TELUQ University (the University of Québec’s Distance Learning). He has published or edited 9 books or special issues and more than 40 articles and book chapters. He participated, among other things, in two reference economics textbooks: Principes d'économie moderne, with Joseph E. Stiglitz (Nobel Prize in Economics), Carl E. Walsh and Jean-Dominique Lafay, and Principes de l'économie with Gregory N. Mankiw and Mark P. Taylor. His latest book, Econophysics and financial economics: An emerging dialogue, is published by Oxford University Press.

Notes

1 For a historical surveys of the history of bibliometrics, see Edwards (2017).

2 As Stone (1979) reminded, we can oppose the narrative history, which is descriptive, to the structural history, which is analytical.

3 The main lines of qualitative history of the EMH are nowadays well known (Delcey, 2017; Frydman & Goldberg, 2010; Jovanovic, 2008, 2010; Sewell, 2011; Thicke, 2017).

4 Bernstein (1992), Merton (1998), Scholes (1998), Dimson and Mussavian (1999, 2000), Whelan, Bowie, and Hibbert (2002), or Davis and Etheridge (2006).

5 See Courtault et al. (2002), Ben-El-Mechaiekh and Dimand (2018) and Jovanovic (2000).

6 In 1914, Bachelier published a book for non-specialists, Le Jeu, la Chance et le Hasard.

7 It is worth mentioning that Bachelier presented his Theory of speculation in all of his books. In his last book, published in 1938, his presentation ‘was more concise and readable and more mathematically elegant than his earlier statements of the theory’ (Ben-El-Mechaiekh & Dimand, 2018, p. 41).

8 During the 1960s, Cootner supervised more than twenty M.A. and Ph.D. theses in financial economics and became an essential figure in the development of the discipline at MIT, before moving to Stanford University where Working was. In 1964, he edited the first anthology of articles dedicated to The Random Character of Stock Market Prices. This book also provided the first translation of Bachelier’s thesis and contributed enormously to the spread of the random walk model.

9 I took as the benchmark the seminal papers of financial economics published during the 1960s and the 1970s and cited in social science journals. Cootner (1962) was cited 47 times compared to 473 times in average (with a maximum of 1997 times and the minimum 22 times); it is included in the third-lowest decile.

10 See also Polillo (2015) on the role of empirical methodology in the diffusion of the EMH defended by Fama.

11 A joint-test refers to the fact that, on a given market, any test of the efficiency (i.e. the fact prices fully reflect available information) tests at the same time the notion of efficiency and the asset-pricing model used to price securities on this market. In other words, any empirical refutation (or validation) can be due either to the fact that the market is not efficient (or efficient) or that the model used is not appropriate (or appropriate) for the test. In other words, such a joint-test implies that market efficiency per se is not testable per se (Campbell, Lo, & MacKinlay, 1997; Cuthbertson, 2004; Fama, 1976a, 1991; Findlay & Williams, 2000; Jovanovic, 2010; LeRoy, 1976, 1989; Lo, 2000).

12 It is worth noting that LeRoy (1976) is not necessary a paper that will be cited because it is a criticism or a respond. In other terms, this paper could have an impact while it was not cited, the last section will come back on this point.

13 The University of Rochester was a satellite of the University of Chicago (Fourcade & Khurana, 2017, p. 363).

14 However, as remind, that Fama (1976a) explanation was reproduced in Fama largely diffused book, Foundations of finance (1976b).Therefore, it is extremely difficult to measure the impact of Fama (1976a) because Foundations of finance was a textbook largely diffused and the definition of the EMH was one topic among numerous ones.

15 It can be noticed that I did a ‘basic’ co-citation analysis about Fama (1976a) and LeRoy (1976) manually (i.e. the fact that 55% of the articles which cite these two articles are the same), because the number of citations is very low. However, for publications largely cited I would have used a deeper document co-citation analysis (Leydesdorff, 1998).

16 More precisely, the starting point of the qualitative history of the EMH was what the economists knew and that can be called a canonical history (Jovanovic, 2008).

17 It is worth mentioning that citation analysis does not differentiate between references that are in agreement with the source paper and references that criticize the source paper. However, Polillo (2015, p. 23), discussing the history of the EMH, suggested to ‘shift attention to networks of prestige, as recently suggested by Collins and Guillen (2012) as a part of a larger model of “mutual halos”’ in order to reduce the similar bias we have in citation networks.

18 Similar problems existed for Econophysics, see Jovanovic and Schinckus (2017) or Gingras and Schinckus (2012).

19 Black died in 1995 and thus was not eligible for the prize. Scholes shared the prize with Merton.

20 It is worth mentioning that the ‘inside-knowledge history’ is a very limited vision of qualitative history.