Skip to main content
SpringerLink
Log in

Did institutions herd during the internet bubble?

  • Original Research
  • Published:
Review of Quantitative Finance and Accounting Aims and scope Submit manuscript

Abstract

We examine the trading behavior of institutional investors during the internet bubble and crash of 1998–2001, and its impact on stock prices. Similar to some recent findings concerning the trading behavior of hedge funds and NASDAQ 100 stocks, we find that during the bubble all types of institutions herded with great intensity into internet stocks for a comprehensive sample of institutional investors and internet stocks. In addition to this, we present three entirely new results. First, institutional herding was much greater than what can be explained by momentum trading. Second, institutions as a group continued to increase their holdings of internet stocks for two quarters past the market peak during the first quarter of 2000, and three quarters past the peak for individual stock prices, suggesting that institutions were unable to time the price peaks. Finally and most importantly, we find positive abnormal returns contemporaneous with institutional herding and negative abnormal returns (reversals) at the point that herding ceased. This finding suggests that institutions’ trading created temporary price pressures, and may have contributed to the bubble.

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

Similar content being viewed by others

Notes

  1. The return we report for our internet index is similar to that reported by Ofek and Richardson (2003) for theirs. This return is only indicative of price changes during this time [or’ the period in question’], and does not represent the return an institution could have earned.

  2. Pastor and Veronesi (2006) suggest that the significant increase in the prices of NASDAQ stocks during the 1990s does not necessarily represent a bubble. They present a model in which higher uncertainty about future growth results in higher prices, and argue that the higher prices of NASDAQ stocks during the 1990s could potentially be explained by the high uncertainty in their expected growth rate.

  3. In a survey conducted by Goetzmann and Dhar (2006), investors willingly admitted buying overvalued stocks on the expectation that their prices would continue to rise.

  4. For example, short sale restrictions could have limited arbitrage for overvalued internet stocks. However, Battalio and Schultz (2006) argue that investors could, in fact, have used synthetic options to circumvent short-sales restriction; that they did not is because the overpricing was not obvious to them. Lakonishok et al. (2007) show similar findings.

  5. Our choice of sample period is consistent with DeLong and Konstantin’s argument (2006) that the dot-com bubble did not begin until 1998.

  6. Several prior studies examine the impact of herding-induced trading on prices. An incomplete list includes Sias (2004), Lu et al. (2012), Hoitash and Krishnan (2008), Khanthavit (1998).

  7. There is no strict definition as to what constitutes an internet firm and no unique SIC code is associated with internet firms, since any company can perform both internet-related and non-internet related activities.

  8. These requirements set out the conditions for the mandatory reporting of holdings. Some institutions voluntarily report holdings of all stocks. A check of our data reveals that about 17 % of the reported positions did not surpass either of these levels, and thus did not have to be reported.

  9. According to Thomson Financial, a mapping error occurred when Thomson data and data from the former Technimetrics were merged; many institutions were erroneously reclassified as type 5. Many of the reclassifications also saw minor changes in firms’ names, for instance, the adding or deleting of the word “The.” In order to reduce reclassification errors, we reassign all institutions to their original classification if the reclassified institution was initially type 1, 2, 3 or 4, and its classification was subsequently changed to type 5.

  10. This is done since we do not know the exact number of shares held by these institutions in the quarter before they first reported the holdings data on 13-F forms. Similarly, we do not know the exact number of shares held by them in the quarter following their final report. Any assumption with respect of number of shares would impact the pit and pt for these quarters, and thus our herding measure.

  11. To determine the robustness of our results with respect to the screen of the number of traders, we calculate herding statistics with the requirement that at least twenty reporting institutions traded the stock during the quarter. While requiring more traders for each stock reduces the number of stock-quarters in the analysis, especially for smaller firms, the results are not meaningfully affected.

  12. Plausibly, new additions to the S&P 500 index and seasoned equity offerings could induce buy herding by institutional investors. We find only 4 stocks from our sample of firms that were added to the S&P 500 index as of the end of 2000. Similarly, using CRSP data, we identify only 28 possible seasoned equity offerings in our sample of stocks for the entire period 1998–2001.

  13. The total number of stock-quarters available for this panel is 3,210, because 364 of the internet firms went through an initial public offering during the prior quarter, and the return cannot be calculated for the quarter during which the IPO was released. Throughout the paper, we exclude these 364 stock-quarters any time we include the prior quarter return in our analysis. Finally, there was one other firm that had a missing prior quarter return for a reason other than the release of its IPO.

  14. The inferences from this comparison are the same when compared with the 2,553 internet stock-quarters with matches as if we were to compare them with the 3,210 internet stock-quarters with prior quarter returns (including both those with matches and those without matches).

  15. Dass et al. (2008) find that mutual fund herding during bubble and crash periods in high price NASDAQ stocks was related to the incentives in the compensation contracts of fund managers. They find that funds with high incentive contracts diverged from the herd and had lower exposure to bubble stocks. This lower exposure had a significant impact on their performance. Sias (2004) argues that compared to banks, insurance firms, endowments and foundations, mutual funds may experience greater changes in inflows/outflows as a result of reputational changes.

  16. These results do not rule out the distinct possibility that some small number of institutions were successful in timing the onset of the crash. See Brunnermeirer and Nagel (2004).

  17. Of the 430 internet firms in the sample, 139 firms had a price peak in the fourth quarter of 1999, 119 in the first quarter of 2000, 41 in the second quarter of 1999, and 37 in the first quarter of 1999. 36 peaks occurred during 1998, and only three in 2001.

  18. Nofsinger and Sias (1999) find that the decile of NYSE stocks, which experiences the largest annual increase in institutional ownership, outperform the decile of stocks that have seen the largest decrease in institutional ownership. Similarly, GTW (1995) and Wermers (1999) find a similar relationship for quarterly changes in mutual funds holdings and stock returns.

  19. Wermers finds evidence of price pressures in the current quarter. He finds no evidence of reversals in any of the four subsequent quarters. Instead, he finds evidence of continuations in quarter +1 for buy herded stocks and in quarters +1 and +2 for sell herded stocks. He interprets this as evidence that herding leads to efficient prices, even if adjustments take some time.

  20. We also perform simultaneous industry and size adjustment; the results are similar. Size adjustment alone will not take into account the systematic industry level return differences that were experienced by internet firms during this period.

  21. There were 112 stock-quarters used to compute the buy herding intensities reported in Table 1 that are not included in Table 6. Of these, 83 were stock-quarters for which the current quarter is the 4th quarter of 2001; these are omitted in Table 6 because our holdings data ends in that quarter. Thus, we cannot classify the subsequent quarter (1st quarter of 2002) as buy herding, no herding or sell herding. In addition, 29 stock-quarters had a missing return in the subsequent quarter because of delisting.

  22. There were 148 stock-quarters used to compute the sell herding intensities reported in Table 1 that are not included in Table 6. Of these, 47 were stock-quarters for which the current quarter is the 4th quarter of 2001. In addition, 101 stock-quarters had a missing return in the subsequent quarter because of delisting.

  23. Furthermore, the abnormal returns in the subsequent quarter for firms that are in the no herding group are between the group that remained in the buy herding group and the one that switched to sell herding.

  24. The higher AF(i,t)s for smaller firms resulting from fewer traders makes it possible that smaller firms are overrepresented in the “no herding” group. We investigate [or ‘look into’ or ‘consider’] this possibility by comparing the size quintile distribution of firms that were in the “no herding” group for the subsequent quarter to the size distribution of the current quarter (i.e., quarter of herding). While there is a slight tendency in the subsequent quarter distribution to over represent the lowest quintile and under represent the upper size quintile, the % differences are small. Thus, we conclude that the size distributions are not different.

References

  • Abreu D, Brunnermeier M (2003) Bubbles and crashes. Econometrica 71(1):73–204

    Article  Google Scholar 

  • Banerjee A (1992) A simple model of herd behavior. Q J Econ 107(3):797–817

    Article  Google Scholar 

  • Battalio R, Schultz P (2006) Options and the bubble. J Financ 61(5):2071–2102

    Article  Google Scholar 

  • Bhattacharya U, Galpin N, Ray R, Yu X (2009) The role of the media in the internet bubble. J Financ Quant Anal 44(3):657–682

    Article  Google Scholar 

  • Bikhchandani S, Hirshleifer D, Welch I (1992) A theory of fads, fashion, custom, and cultural change as informational cascades. J Polit Econ 100(5):992–1026

    Article  Google Scholar 

  • Brunnermeirer M, Nagel S (2004) Hedge funds and the technology bubble. J Financ 59(5):2013–2040

    Article  Google Scholar 

  • Dasgupta A, Prat A, Verardo M (2011a) Institutional trade persistence and long-term equity returns. J Financ 66(2):635–653

    Article  Google Scholar 

  • Dasgupta A, Prat A, Verardo M (2011b) The price impact of institutional herding. Rev Financ Stud 24(3):892–925

    Article  Google Scholar 

  • Dass N, Massa M, Patgiri R (2008) Mutual funds and bubbles: the surprising role of contractual incentives. Rev Financ Stud 21(1):51–99

    Article  Google Scholar 

  • DeLong B, Konstantin M (2006) A short note on the size of the dot-com bubble. NBER Working Paper No. W12011

  • Dreman D (1979) Contrarian investment strategy: the psychology of stock market success. Random House, New York

    Google Scholar 

  • Friedman B (1984) A comment: stock prices and social dynamics. Brook Papers Econ Activity 25(2):504–508

    Google Scholar 

  • Froot K, Scharfestein D, Stein J (1992) Herd on the street: informational inefficiencies in a market with short-term speculation. J Financ 47(4):1461–1484

    Article  Google Scholar 

  • Goetzmann W, Dhar R (2006) Bubble investors: what were they thinking?. Yale University, Working Paper

    Google Scholar 

  • Griffin J, Harris J, Shu T, Topaloglu S (2011) Who drove and burst the tech bubble? J Financ 66(4):1251–1290

    Article  Google Scholar 

  • Grinblatt M, Titman S, Wermers R (1995) Momentum investment strategies, portfolio performance, and herding: a study of mutual fund behavior. Am Econ Rev 85(5):1088–1105

    Google Scholar 

  • Hand J (2003) Profits losses and the non-linear pricing of internet stocks. In: Hand JRM, Lev B (eds) Intangible assets: values, measures and risks. Oxford University Press, Oxford

    Google Scholar 

  • Hirshleifer D, Subrahmanyam A, Titman S (1994) Security analysis and trading patterns when some investors receive information before others. J Financ 49(5):1665–1698

    Article  Google Scholar 

  • Hoitash R, Krishnan M (2008) Herding, momentum and investor over-reaction. Rev Quant Financ Acc 30(1):25–47

    Article  Google Scholar 

  • Hong H, Scheinkman J, Xiong W (2006) Asset float and speculative bubbles. J Financ 61(3):1073–1117

    Article  Google Scholar 

  • Khanthavit A (1998) Information structure, trading behavior and stock prices in the thai market. Rev Pac Bassin Financ Mark Pol 1(3):321–353

    Article  Google Scholar 

  • Lakonishok J, Shleifer A, Vishny R (1992) The impact of institutional trading on stock prices. J Financ Econ 32(1):23–43

    Article  Google Scholar 

  • Lakonishok J, Lee I, Pearson N, Poteshman A (2007) Option market activity. Rev Financ Stud 20(3):813–857

    Article  Google Scholar 

  • Lu Y, Fang H, Nieh C (2012) The price impact of foreign institutional herding on large-size stocks in the taiwan stock market. Rev Quant Financ Acc 39(2):189–208

    Article  Google Scholar 

  • Maug E, Naik N (1995) Herding and delegated portfolio management: the impact of relative performance evaluation on asset allocation. Duke University and London Business School, Working Paper

    Google Scholar 

  • Nofsinger J, Sias R (1999) Herding and feedback trading by institutional investors. J Financ 54(6):2263–2316

    Article  Google Scholar 

  • Ofek E, Richardson M (2003) Dotcom mania: the rise and fall of internet stock prices. J Financ 58(3):1113–1137

    Article  Google Scholar 

  • Pastor L, Veronesi P (2006) Was there a NASDAQ bubble in the late 1990 s? J Financ Econ 81(1):61–100

    Article  Google Scholar 

  • Scharfstein D, Stein J (1990) Herd behavior and investment. Am Econ Rev 80(3):465–479

    Google Scholar 

  • Sias R (2004) Institutional herding. Rev Financ Stud 17(1):165–206

    Article  Google Scholar 

  • Wermers R (1999) Mutual fund trading and the impact on stock prices. J Financ 54(2):581–622

    Article  Google Scholar 

  • Wermers R (2003) Is money really ‘smart’? New evidence on the relation between mutual fund flows, managerial behavior, and performance persistence. University of Maryland, Working Paper

    Google Scholar 

Download references

Acknowledgments

I am grateful for the valuable suggestions from the editor, C. F. Lee, and an anonymous referee that greatly improved the quality of this paper. This paper is based on my first essay for my dissertation. I deeply appreciate the guidance of my dissertation committee especially that of John Easterwood and Raman Kumar, both of whom were of immense help in guiding me at every stage of this essay. I sincerely thank them for their invaluable contributions. I would also like to thank the discussants and participants of FMA and European FMA for their helpful comments.

Author information

Authors and Affiliations

  1. Dearborn, MI, USA

    Vivek Singh

Authors
  1. Vivek Singh
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Vivek Singh.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Singh, V. Did institutions herd during the internet bubble?. Rev Quant Finan Acc 41, 513–534 (2013). https://doi.org/10.1007/s11156-012-0320-1

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11156-012-0320-1

Keywords

JEL Classification

Search

Navigation