Skip to main content
SpringerLink
Log in

Time-varying price discovery in the eighteenth century: empirical evidence from the London and Amsterdam stock markets

  • Original Paper
  • Published:
Cliometrica Aims and scope Submit manuscript

Abstract

This paper examines the time-varying nature of price discovery in eighteenth century cross-listed stocks. Specifically, we investigate how quickly news is reflected in prices for two of the great moneyed companies, the Bank of England and the East India Company, over the period 1723–1794. These British companies were cross-listed on the London and Amsterdam stock exchange and news between the capitals flowed mainly via the use of boats that transported mail. We examine in detail the historical context surrounding the defining events of the period and use these as a guide to how the data should be analysed. We show that both trading venues contributed to price discovery, and although the London venue was more important for these stocks, its importance varies over time.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Notes

  1. Since its development, the information shares approach has been extended and successfully applied to cross-listed stocks (see, e.g. Huang 2002; Pascual et al. 2006; Chan et al. 2007; Hoag and Norman 2013), government bonds (see, e.g. Campbell and Hendry 2007; Caporale and Girardi 2013), derivatives (Booth et al. 1999; Chakravarty et al. 2004), and the market for carbon emissions (Mizrach and Otsubo 2014). The vast majority of previous studies assume that price discovery is time-invariant (see Aitken et al. 2012; Hoag and Norman 2013, for recent exceptions).

  2. The use of the VEC model is common in the extant literature and is consistent with a number of market microstructure models; see, e.g. section II in Hasbrouck (1995).

  3. The only other contribution in this area is the study by Dempster et al. (2000). They examine the extent to which the two markets are integrated using a common features approach. To this end, they implement a VEC model too and proceed to test for common cycles rather than using the price discovery approach as we do. They find that the two markets were indeed highly interlinked, and that price movements across the two markets were matched in both the short and long runs. Dempster et al. argue that, “relevant movements in the market seem to have originated in London...”

  4. This act forbade options and forwards being traded in London (Neal 1990, p. 150).

  5. The correlations between the corresponding first differences of prices vary slightly more over the sub-periods but since no formal tests of significance are conducted, it may be that they merely represent standard variations in correlation estimates that would occur over time even if the data generating process remained constant throughout.

  6. This finding corroborated Neal’s (1985) earlier finding that the London, Amsterdam, Paris and New York markets were well integrated in the nineteenth century, and the findings of Harrison (1998), who shows that the distribution of price changes existing at that time, was very similar to that of today.

  7. The intercept in (1) is suppressed for presentational purposes only.

  8. The use of the Cholesky decomposition means that the estimated measure is not invariant to the ordering of the data (Lütkepohl 1991). To avoid this dependence, we augment the original information share measure by replacing \({\mathbf {D}}^\prime {\mathbf {L}}\) with \({\mathbf {D}}^\prime \varvec{\varSigma }/\sqrt{\sigma _{jj}}\); see Pesaran and Shin (1998) for motivation within the context of impulse response functions.

  9. Time variation in the residual covariance matrix associated with (1) will also affect the information share measure via \(\mathbf {L}\).

  10. The residual covariance matrix is recalculated after each estimation of the model.

  11. While the original data source contains price data for the South Sea company as well as the BOE and EIC, we do not use this series since it is very thinly traded after the bubble of 1720 and so is likely to lead to inaccurate results. There were around 150 companies that were publicly quoted at some point between 1690 and 1834 (see Shea 2000), but most of these were extremely small and with very incomplete records making them not amenable to formal statistical analysis.

  12. The data frequency is known to be irregular (see Table 3 in Neal 1987). However, it should be noted that the information share measure we employ is robust to situations in which the sampling frequency used does not coincide with the true frequency over which the data are generated (see Table 5 in Yan and Zivot 2010).

  13. See https://www.icpsr.umich.edu/icpsrweb/ICPSR/studies. The title of the archive is: Course of the Exchange, London, 1698–1823 and Amsterdamsche Beurs, Amsterdam, 1723–1794 (ICPSR 1008).

  14. Indeed, as we show in the subsequent analysis, the proportion of price discovery attributable to Amsterdam rose slightly rather than falling.

  15. Dempster et al. (2000), who also employ a VEC model, show that each of the London stocks has a \([1 -1]\) cointegrating relationship with its Amsterdam counterpart, and so, this specification is well founded.

  16. This system is estimated via the Kalman filter technique using the state-space object in Eviews 8.0 (×64).

  17. We use the Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test statistic to examine the null hypothesis of stationary residuals.

  18. The effect of this choice of mispricing model specification on the subsequent price discovery measures is examined later in the paper.

  19. The VEC model is estimated via ordinary least squares in Eviews 8.0 (\(\times 64\)).

  20. We have also tested the null hypothesis that the London component/information share equals one, representing a flow of information only from London to Amsterdam. In all cases, this null hypothesis is strongly rejected.

  21. Neal’s peace and war periods are 09/08/23–19/10/39, 11/11/48–14/07/56, 18/02/63–04/03/78, 06/12/82–22/09/90, and 21/10/39–23/10/48, 04/08/56–05/02/63, 02/03/78–20/11/82, 08/10/90–19/12/94, respectively. It should be noted that for the American War of Independence, it is arguable that a longer period could be considered, i.e. 1775–1783.

  22. We follow Neal (1987) and use the 1723–1737 and 1738–1794 pre- and post-Barnard act samples, respectively.

  23. However, we should also consider a panic in London beginning in April 1761 and mentioned by Neal (1990, p. 170). Charles Mackay placed the blame for this panic on two earthquakes and the fear of a third, which had been prophesised by “a crack-brained fellow called Bell” (Mackay 1995, p. 224). Hence it may well be that this reduction in London price discovery may be unconnected to the succession of George III.

  24. Thus, we begin with the January 1723 to December 1732 sample, followed by the January 1724 to December 1733 sample, and so on. For each window, we determine the optimal lag structure in the VEC model using the AIC.

  25. We define regularly spaced observations as those with a time interval between reported prices of between 11 and 17 days. Consequently, the only variation in data frequency that is permitted in due to the occurrence of religious holidays and thus the data “reflect precisely irregularities in trading activity on the Effectenbeurs” (Neal 1987, p. 105).

  26. See Koudijs (2013) for a detailed discussion of flows between London and Amsterdam and how volatility is affected by the arrival of news.

References

  • Aitken M, Di Marco E, Harris F (2012) Price discovery efficiency and information impounding on NYSE Euronext Paris. SSRN working paper http://ssrn.com/abstract=2029338

  • Beach B, Norman S, Wills D (2013) Time or spot? A revaluation of Amsterdam market data prior to 1747. Cliometrica 7:61–85

    Article  Google Scholar 

  • Bowen H (1989) Investment and empire in the later eighteenth century: east India stockholding, 1756–1791. Econ Hist Rev 42:186–206

    Article  Google Scholar 

  • Booth G, So R, Tse Y (1999) Price discovery in the German equity index derivatives market. J Futures Mark 19:619–643

    Article  Google Scholar 

  • Campbell B, Hendry S (2007) Price discovery in Canadian and U.S. 10-Year government bond markets. Bank of Canada working paper 2007–43

  • Caporale G, Girardi A (2013) Price discovery and trade fragmentation in a multi-market environment: evidence from the MTS system. J Bank Finance 37:227–240

    Article  Google Scholar 

  • Chan K, Menkveld A, Yang Z (2007) The informativeness of domestic and foreign investors’ stock trades: evidence from the perfectly segmented Chinese market. J Financ Mark 10:391–415

    Article  Google Scholar 

  • Chakravarty S, Gulen H, Mayhew S (2004) Informed trading in stock and option markets. J Finance 59:1235–1257

    Article  Google Scholar 

  • Chu Q, Hsieh W, Tse Y (1999) Price discovery on the S&P 500 index markets: an analysis of spot index, index futures and SPDRs. Int Rev Financ Anal 8:21–34

    Article  Google Scholar 

  • Dale R (2004) The first crash: lessons from the South Sea Bubble. Princeton University Press, Princeton

    Google Scholar 

  • Davison A, Hinkley D (1997) Bootstrap methods and their application. Cambridge series in statistical and probabilistic mathematics. Cambridge University Press, Cambridge

    Book  Google Scholar 

  • Dempster G, Wells J, Wills D (2000) A common-features analysis of Amsterdam and London financial markets during the eighteenth century. Econ Inquiry 38:19–33

    Article  Google Scholar 

  • Harris F, McInish T, Wood R (2002) Security price adjustment across exchanges: an investigation of common factor components for Dow stocks. J Financ Mark 5:277–308

    Article  Google Scholar 

  • Harrison P (1998) Similarities in the distribution of stock market price changes between the eighteenth and twentieth Centuries. J Bus 71:55–79

    Article  Google Scholar 

  • Hasbrouck J (1995) One security, many markets: determining the contributions to price discovery. J Finance 50:1175–1199

    Article  Google Scholar 

  • Hoag C, Norman S (2013) Transatlantic capital market price discovery during a financial crisis. Bull Econ Res 65:1–9

    Article  Google Scholar 

  • Homer S, Sylla R (2005) A history of interest rates, 4th edn. Wiley, Chichester

    Google Scholar 

  • Huang R (2002) The quality of ECN and Nasdaq market maker quotes. J Finance 57:1285–1319

    Article  Google Scholar 

  • Koudijs P (2013) The boats that did not sail: Evidence on the sources of asset price volatility from an 18th century natural experiment. NBER Working Paper No. 18831

  • Lehmann B (2002) Some desiderata for the measurement of price discovery across markets. J Financ Mark 5:259–276

    Article  Google Scholar 

  • Lütkepohl H (1991) Introduction to multiple time series analysis. Springer, Berlin

    Book  Google Scholar 

  • Mackay C (1995) Extraordinary popular delusions and the madness of crowds. Crown Publications, London

    Google Scholar 

  • Mammen E (1993) Bootstrap and wild bootstrap for high dimensional linear models. Ann Stat 21:255–285

    Article  Google Scholar 

  • Mizrach B, Otsubo Y (2014) The market microstructure of the European climate exchange. J Bank Financ 39:107–116

    Article  Google Scholar 

  • Neal L (1985) Integration of international capital markets: quantitative evidence from the eighteenth to twentieth centuries. J Econ Hist 45:219–226

    Article  Google Scholar 

  • Neal L (1987) The integration and efficiency of the London and Amsterdam stock markets in the eighteenth century. J Econ Hist 47:97–115

    Article  Google Scholar 

  • Neal L (1990) The rise of financial capitalism. Cambridge University Press, New York

    Google Scholar 

  • Pascual R, Pascual-Fuster B, Climent F (2006) Cross-listing, price discovery and the informativeness of the trading process. J Financ Mark 9:144–161

    Article  Google Scholar 

  • Pesaran H, Shin Y (1996) Cointegration and speed of convergence to equilibrium. J Econom 71:117–143

    Article  Google Scholar 

  • Pesaran H, Shin Y (1998) Generalized impulse response analysis in linear multivariate models. Econ Lett 58:17–29

    Article  Google Scholar 

  • Shea G (2000) The course of the exchange: Measuring and interpreting returns processes in 18th and early 19th century Britain. Mimeo, Department of Economics, University of St. Andrews

  • Stock J, Watson M (1988) Testing for common trends. J Am Stat Assoc 83:1097–1107

    Article  Google Scholar 

  • Van Dillen JG (1931) Effectenkoersen aan de Amsterdamsche Beurs, 1723–1794. Economische Historische Jaarboek 17:1–46

    Google Scholar 

  • Wilson C (1941) Anglo-Dutch commerce and finance in the eighteenth century. Cambridge University Press, Cambridge

    Google Scholar 

  • Wright J (1999) British government borrowing in wartime, 1750–1815. Econ Hist Rev 52:355–361

    Article  Google Scholar 

  • Yan B, Zivot E (2010) A structural analysis of price discovery measures. J Financ Mark 13:1–19

    Article  Google Scholar 

Download references

Acknowledgments

The authors are grateful to Gary Shea for kindly providing us with his data on the dividend payments for the companies. We are also grateful for useful comments on previous versions of this paper by the Founding Managing Editor, Claude Diebolt, and two anonymous referees.

Author information

Authors and Affiliations

  1. ICMA Centre, Henley Business School, University of Reading, Reading, UK

    Adrian R. Bell & Chris Brooks

  2. School of Economics, Finance and Management, University of Bristol, Bristol, BS8 1TN, UK

    Nick Taylor

Authors
  1. Adrian R. Bell
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chris Brooks
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Nick Taylor
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Nick Taylor.

Appendix

Appendix

1.1 Bootstrap procedure

The wild bootstrap used in this paper consists of the following steps:

  1. Step 1:

    Estimate the VEC model based on the original data to obtain the following vector of fitted return values

    $$\begin{aligned} \Delta \widehat{\mathbf {y}}_t = \widehat{\varvec{\alpha }} \varvec{\beta }^\prime {\mathbf {y}}_{t-1} + \sum _{i=1}^{p} \widehat{\varvec{\varGamma }}_i \Delta {\mathbf {y}}_{t-i}, \end{aligned}$$

    with residuals given by \(\widehat{\varvec{\epsilon }}_t\). The estimation of this model can be based on the full sample, a sub-sample, or a rolling window of data. In each case, the lag length is determined by the AIC.

  2. Step 2:

    Construct bootstrap residuals as follows

    $$\begin{aligned} \widetilde{\varvec{\epsilon }}_t = \widetilde{u}_t \widehat{\varvec{\epsilon }}_t, \end{aligned}$$

    where \(\widetilde{u}_t\) is obtained using the two-point distribution proposed by Mammen (1993), that is

    $$\begin{aligned} \widetilde{u}_t = {\left\{ \begin{array}{ll} -(\sqrt{5}-1)/2, &{}\quad \text{ with } \text{ probability } p=(\sqrt{5}+1)/(2\sqrt{5}),\\ +(\sqrt{5}+1)/2, &{}\quad \text{ with } \text{ probability } 1-p. \end{array}\right. } \end{aligned}$$
  3. Step 3:

    Construct the bootstrap data generating process (DGP) using the bootstrap residuals

    $$\begin{aligned} \varDelta \widetilde{\mathbf {y}}_t = \varDelta \widehat{\mathbf {y}}_t + \widetilde{\varvec{\epsilon }}. \end{aligned}$$
  4. Step 4:

    Re-estimate the model using the bootstrap DGP and construct the component and information shares using the respective formulae given by (3a) and (3b), denoting these by \(\widetilde{M}_j\) if a time-invariant version is required (achieved via estimation over the full sample), or by \(\widetilde{M}_{j,t}\) if a time-varying version is required (achieved via sub-period or rolling window estimation).

  5. Step 5:

    Repeat Steps 2–4 a total of 1,000 times.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bell, A.R., Brooks, C. & Taylor, N. Time-varying price discovery in the eighteenth century: empirical evidence from the London and Amsterdam stock markets. Cliometrica 10, 5–30 (2016). https://doi.org/10.1007/s11698-014-0120-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11698-014-0120-z

Keywords

JEL Classification

Search

Navigation