No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
The index premium and its hidden cost for index funds
Abstract
This paper empirically investigates the index premium and its implications from 1990 to 2005. For additions to the S&P 500 and Russell 2000, we find that the price impact from announcement to effective day has averaged + 8.8% and + 4.7%, respectively, and -15.1% and -4.6% for deletions. The premia have been growing over time, peaking in 2000, and declining since then. The implied price elasticity of demand increases with firm size and decreases with idiosyncratic risk, supporting theoretical predictions. We also introduce a new concept that we label the index turnover cost, which represents a hidden cost borne by index funds (and the indexes themselves) due to the index premium. We illustrate this cost and estimate its lower bound as 21-28 bp annually for the S&P 500 and 38-77 bp annually for the Russell 2000.
... Dividing the sample into three subperiods shows that dealer behavior has 4 Blume and Edelen (2004) show that stock index trackers display a similar behavior. 5 See also Chen, Noronha, and Singal (2006), Petajisto (2011), and Pedersen (2018) about this cost. changed after the crisis. ...
... The bid-to-bid returns pick up a hidden cost of index tracking (see, e.g., Chen, Noronha, and Singal, 2006, Pedersen, 2018, Petajisto, 2011. To see this, consider that the bond index return is calculated using the average price from day 0 and that this price will be heavily depressed by the concentrated selling from index trackers. ...
... The cost estimate can be compared to the hidden cost from stock index investment calculated in Petajisto (2011) of 21-28 bps annually for the S&P 500. Note that this annual cost estimate is only indicative, because we do not consider what would happen dynamically when volumes are redistributed away from the exclusion date. ...
Liquidity provision for corporate bonds has become significantly more expensive after the 2008 crisis. Using index exclusions as a natural experiment during which uninformed index trackers request immediacy, we find that the cost of immediacy has more than doubled. In addition, the supply of immediacy has become more elastic with respect to its price. Consistent with a stringent regulatory environment incentivizing smaller dealer inventories, we also find that dealers revert deviations from their target inventory more quickly after the crisis. Finally, we investigate the pricing impact of information, changes in ownership structure, and differences between bank and nonbank dealers.
Received February 22, 2017; editorial decision May 29, 2018 by Editor Itay Goldstein. Authors have furnished an Internet Appendix, which is available on the Oxford University Press Web site next to the link to the final published paper online.
... This obvious quirk of a large simultaneous constituent adjustment of the Top 40 is not unique (in fact, it is common for other indices too, see for example Green andJame (2011) &Petajisto (2009)). Rather, the clearly communicated and timeous nature of the details of locally rebalanced constituents that are known well in advance is more unique. ...
... If the former (along), we should expect abrupt price changes to be followed by an opposite and equal correction to the previous "fundamental" level of the supply and demand intersection (c.f. Petajisto (2009) for a richer discussion into the relevance of demand slopes in the present context). In finance parlance, if a stock has a constant fundamental value, a sudden price change should be followed by the opposite pricing pressure, ceteris paribus. ...
... Unfortunately, as rebalancing rules differ, a true counter-factual to our limited data sample is not possible. For the S&P500, for example, on which various studies have been applied (see, inter-alia Gowri Shankar and Miller (2006), Green and Jame (2011), Petajisto (2011), Chen, Noronha, and Singal (2004, Platikanova (2008) & Jain (1987), changes are not governed by an equally clear and transparent set of ranking criteria (Kappou, Brooks, and Ward (2010b, 117)), making the comparison less relevant. Nonetheless, the literature does provide some interesting findings and theories to place our own experience into context. ...
... The hidden costs of traditional passive indexing have been studied extensively by Petajisto (2011) andChen, Noronha, andSingal (2006), who presented costs as abnormal returns as a result of arbitrageurs front-running additions and deletions to index compositions. Despite promising growth, the level of indexing of factor-investing strategies is low relative to traditional benchmarks; trades in the indexes in our study are unlikely to garner attention from arbitrageurs because of the relatively low absolute dollar amounts. ...
... Multiple studies have attempted to explain the price response to preannounced additions and deletions to traditional benchmark indexes. Petajisto (2011) measured an index premium from the announcement day through the effective day of additions and deletions to index compositions from 1990 through 2005 and identified average costs of 21-28 bps for the S&P 500 Index and 38-77 bps for the Russell 2000 Index (the two benchmarks with the highest amount of indexed assets). Chen et al. (2006) reported higher average costs for the same benchmarks when they used data for a shorter sample. ...
... The findings presented in Figure 2 are crucial because they imply the conjecture made by existing studies is not applicable to factor-investing strategies. Existing studies-for example, Petajisto (2011) and Chen et al. (2006)-analyzed the market impact costs of changes to popular benchmark indexes, such as the S&P 500 and the Russell 2000. Changes to these indexes are informative because they represent large and easily anticipated trades by the index funds. ...
Although hidden, the implicit market impact costs of factor investing may substantially erode a strategy’s expected excess returns. The rebalancing data of a suite of large and long-standing factor-investing indexes are used in this study to model these market impact costs. A framework to assess the costs of rebalancing activities is introduced. These costs are then attributed to characteristics that intuitively describe the strategies’ demands on liquidity, such as rate of turnover and the concentration of turnover. A number of popular factor-investing implementations are identified, and the authors discuss how their index construction methods, when thoughtfully designed, can reduce market impact costs.
... However, there is growing empirical evidence that suggests otherwise. Downward-sloping demand curves are found, among others, by Shleifer (1986), Kaul et al. (2000), Wurgler, and Zhuravskaya (2002), and Petajisto (2009Petajisto ( , 2011, while evidence of upward-sloping supply schedules is found in Bradley et al. (1988), Bagwell (1992), Brown and Ryngaert (1992), and Hodrick (1999). These articles typically rely on event studies-investigating corporate events that affect either demand or supply of an asset without altering its fundamental value. ...
... However, there is growing empirical evidence that suggests otherwise. Downward-sloping demand curves are found, among others, by Shleifer (1986), Kaul et al. (2000), Wurgler, and Zhuravskaya (2002), and Petajisto (2009Petajisto ( , 2011, while evidence of upward-sloping supply schedules is found in Bradley et al. (1988), Bagwell (1992), Brown and Ryngaert (1992), and Hodrick (1999). These articles typically rely on event studies-investigating corporate events that affect either demand or supply of an asset without altering its fundamental value. ...
... On the other hand, evidence for downward-sloping demand curves is found in Shleifer (1986), who studied the demand for stocks newly included in the S&P 500 index, while controlling for informational content such inclusions may convey. Kaul et al. (2000) provided corroborative evidence of less than perfectly elastic demand from the Toronto Stock Exchange, while Petajisto (2011) and Wurgler and Zhuravskaya (2002) presented more extensive results for the Russell 2000 and S&P 500 stocks. Kalay et al. (2004) found elastic demand and supply schedules using the preopening orders. ...
We give a set of identifying conditions for p-dimensional (p ≥ 2) simultaneous equation systems (SES) with heteroscedasticity in the framework of Gaussian quasi-maximum likelihood (QML). Our conditions rely on the presence of heteroscedasticity in the data rather than identifying restrictions traditionally employed in the literature. The QML estimator is shown to be consistent for the true parameter point and asymptotically normal. Monte Carlo experiments indicate that the QML estimator performs well in comparison to the generalized method of moments (GMM) estimator in finite samples, even when the conditional variance is mildly misspecified. We analyze the relationship between traded stock prices and volumes in the setting of SES. Based on a sample of the Russell 3000 stocks, our findings provide new evidence against perfectly elastic demand and supply schedules for equities.
... However, there is growing empirical evidence that suggests otherwise. Downward-sloping demand curves are found, among others, by Shleifer (1986), Kaul et al. (2000), Wurgler, and Zhuravskaya (2002), and Petajisto (2009Petajisto ( , 2011, while evidence of upward-sloping supply schedules is found in Bradley et al. (1988), Bagwell (1992), Brown and Ryngaert (1992), and Hodrick (1999). These articles typically rely on event studies-investigating corporate events that affect either demand or supply of an asset without altering its fundamental value. ...
... However, there is growing empirical evidence that suggests otherwise. Downward-sloping demand curves are found, among others, by Shleifer (1986), Kaul et al. (2000), Wurgler, and Zhuravskaya (2002), and Petajisto (2009Petajisto ( , 2011, while evidence of upward-sloping supply schedules is found in Bradley et al. (1988), Bagwell (1992), Brown and Ryngaert (1992), and Hodrick (1999). These articles typically rely on event studies-investigating corporate events that affect either demand or supply of an asset without altering its fundamental value. ...
... On the other hand, evidence for downward-sloping demand curves is found in Shleifer (1986), who studied the demand for stocks newly included in the S&P 500 index, while controlling for informational content such inclusions may convey. Kaul et al. (2000) provided corroborative evidence of less than perfectly elastic demand from the Toronto Stock Exchange, while Petajisto (2011) and Wurgler and Zhuravskaya (2002) presented more extensive results for the Russell 2000 and S&P 500 stocks. Kalay et al. (2004) found elastic demand and supply schedules using the preopening orders. ...
We give a set of identifying conditions for simultaneous equation systems (SES) with heteroskedasticity in the framework of the Gaussian quasi maximum likelihood (QML) approach. Our conditions rely on the presence of heteroskedasticity rather than exclusion restrictions. The QML estimators are shown to be consistent and asymptotically normal. Monte Carlo experiments show that the QML estimators perform well in finite samples in comparison to the GMM estimators even when volatility is mildly misspecified. In the framework of SES, we analyse the relationships between traded stock prices and volumes. Based on a sample of the Russell 3000 stocks, our estimation results provide new evidence against the homogeneous valuations hypothesis.
... Index reconstitution effects are another possible explanation for the underperformance of the small capitalization indices. Petajisto (2006) points out that this is especially likely for the Russell 2000, which is reconstituted every year at the end of June, due to the combination of relatively large turnover in the index and the large amount of assets indexed and benchmarked to it. In anticipation of the one-time demand shock by index investors at the end of June, stocks being added to the Russell 2000 outperform stocks being deleted in June, and the reverse occurs in July, lowering the returns on the index itself. ...
... Almost all of it comes from the upper and lower boundaries of the index (size deciles 2 and 5-6, while size deciles 3-4 show very little selection alpha). This suggests that index reconstitution may be creating a slight drag on returns, consistent with Petajisto (2006). However, about 70% of the Russell 2000 negative alpha, 169 bp out of 238 bp per year, still comes simply from its exposure to Fama-French portfolios. ...
... As mentioned above, an additional possible explanation for the negative alpha of the small-cap indices is given by the index reconstitution effects discussed by Petajisto (2006) July coefficient captures any extra alpha in these months, which could be due to reconstitution. ...
Standard Fama-French and Carhart models produce economically and statistically significant nonzero alphas even for passive benchmark indices such as the S&P 500 and Russell 2000. We find that these alphas primarily arise from the disproportionate weight the Fama-French factors place on small value stocks which have performed well, and from the CRSP value-weighted market index which is historically a downward-biased benchmark for U.S. stocks. We explore alternative ways to construct these factors and propose alternative models constructed from common and easily tradable benchmark indices. The index-based models outperform the standard models in common applications such as performance evaluation of mutual fund managers.
... It has been empirically observed that the securities added to equity indexes receive positive returns concurrently with their index additions and shortly thereafter. The main indexes on which such effects are observed are the S&P500 and the Russell U.S. indexes, with many studies, such as [30,13,17,15,33], providing evidences in support of the existence of the aforementioned abnormal returns. ...
... In general, rebalance procedure of equity indexes are not necessarily publicly disclosed and sometimes presents some degree of arbitrariness. For example, as discussed in [13,33], Standard & Poor's maintains a proprietary selection process used to discern which stocks will belong to the new issue of the index and make adjustments whenever it considers it to be [26]. ...
We develop a methodology which replicates in great accuracy the FTSE Russell indexes reconstitutions, including the quarterly rebalancings due to new initial public offerings (IPOs). While using only data available in the CRSP US Stock database for our index reconstruction, we demonstrate the accuracy of this methodology by comparing it to the original Russell US indexes for the time period between 1989 to 2019. A python package that generates the replicated indexes is also provided. As an application, we use our index reconstruction protocol to compute the permanent and temporary price impact on the Russell 3000 annual additions and deletions, and on the quarterly additions of new IPOs . We find that the index portfolios following the Russell 3000 index and rebalanced on an annual basis are overall more crowded than those following the index on a quarterly basis. This phenomenon implies that transaction costs of indexing strategies could be significantly reduced by buying new IPOs additions in proximity to quarterly rebalance dates.
... Pedersen offers a few ballpark figures, citing 0.18% per annum from IPOs, 0.06% from secondary equity offerings, and a return drag related to trading costs of upwards of 0.20% for index funds associated with their required turnover. Petajisto (2011) estimates that the market impact costs associated with index rebalancing to have averaged 0.21%-0.28% for the S&P 500 and 0.38%-0.77% ...
... xi For example, see Ritter and Welch (2002). xii The estimates of Petajisto (2011) are collaborated by inputting the averages for yA, yD, PIA and PID and T as reported by Petajisto into the model that appears in Section 2, specifically equation (7d). ...
Against a background of passive investing in the ascendancy, I delve into the issue of whether active fund managers can add value through the prism of Sharpe’s proposition that active investing is a zero sum game prior costs and negative sum game after costs. I explain some of the gaps in Sharpe’s proposition that leave room for active managers to outperform. I make the point that most research on the performance of active funds neither directly tests Sharpe’s proposition, and if anything seems inconsistent with it acting as a constraint on the ability of active managers to create value. I argue that the circumstances of the investor and the nature of the market are more important considerations for the choice between active and passive management, with notable differentiators including the fee paid, investor objectives and the asset category being considered. The main message is that ‘it depends’.
... That is, rather than separately varying the sizes of active and passive as in Figure A1, we could try to determine these endogenously on the basis of an entry decision. See Gâ rleanu and Pedersen (forthcoming 2017) for such an equilibrium level of active management in the context of a more sophisticated model with private information, search, and equilibrium asset management fees; see Petajisto (2009) Petajisto (2009) argued that because some active investors specialize in certain trades, the risk aversion implicit in the aggregate equity premium (e.g., the dividend yield) Ahead of Print: 8 January 2018. This version has not been through the final proofreading process, which may lead to differences with the Version of Record. ...
... That is, rather than separately varying the sizes of active and passive as in Figure A1, we could try to determine these endogenously on the basis of an entry decision. See Gâ rleanu and Pedersen (forthcoming 2017) for such an equilibrium level of active management in the context of a more sophisticated model with private information, search, and equilibrium asset management fees; see Petajisto (2009) Petajisto (2009) argued that because some active investors specialize in certain trades, the risk aversion implicit in the aggregate equity premium (e.g., the dividend yield) Ahead of Print: 8 January 2018. This version has not been through the final proofreading process, which may lead to differences with the Version of Record. ...
I challenge William F. Sharpe’s famous equality that “before costs, the return on the average actively managed dollar will equal the return on the average passively managed dollar.” This equality is based on the implicit assumption that the market portfolio never changes, which does not hold in the real world because new shares are issued, others are repurchased, and indexes are reconstituted—so even “passive” investors must regularly trade. Therefore, active managers can be worth positive fees in aggregate, allowing them to play an important economic role: helping allocate resources efficiently. Passive investing also plays a useful economic role: creating low-cost access to markets.
... However, there is growing empirical evidence that suggests otherwise. Downward-sloping demand curves are found, amongst others, by Shleifer (1986), Kaul et al. (2000), Wurgler, and Zhuravskaya (2002), and Petajisto (2009Petajisto ( , 2011, while evidence of upward-sloping supply schedules is found by Bradley et al. (1988), Bagwell (1992), Brown and Ryngaert (1992), and Hodrick (1999). These studies are typically based on events that affect either demand or supply of an asset without altering its fundamental value such as stock inclusions in major market indices and share repurchases. ...
... On the other hand, evidence for downward-sloping demand curves is reported in Shleifer (1986), who studies the demand for stocks newly included in the S&P 500 index, while controlling for informational content such inclusions may convey. Kaul et al. (2000) provides corroborative evidence of less than perfectly elastic demand from the Toronto Stock Exchange, while Petajisto (2011), and Wurgler and Zhuravskaya (2002) present more extensive results for the Russell 2000 and S&P 500 stocks. Kalay et al (2004) find elastic demand and supply schedules based on the pre-opening orders. ...
We give a set of identifying conditions for p-dimensional (p ≥ 2) simultaneous equation systems (SES) with heteroscedasticity in the framework of Gaussian quasi-maximum likelihood (QML). Our conditions rely on the presence of heteroscedasticity in the data rather than identifying restrictions traditionally employed in the literature. The QML estimator is shown to be consistent for the true parameter point and asymptotically normal. Monte Carlo experiments indicate that the QML estimator performs well in comparison to the generalized method of moments (GMM) estimator in finite samples, even when the conditional variance is mildly misspecified. We analyze the relationship between traded stock prices and volumes in the setting of SES. Based on a sample of the Russell 3000 stocks, our findings provide new evidence against perfectly elastic demand and supply schedules for equities.
... This can be experienced from Figure 2. The daily MCARs have a negative impact on the day prior to the announcement, on the AD, and also on the following day of announcement for both additions and deletions. The AD window results vary from the past research works of Lynch and Mendenhall (1997), Chakrabarti et al (2005), Kumar (2007) and Petajisto (2011) wherein the abnormal returns around the AD were significantly positive (negative) for both additions and deletions, whereas these results are comparable with the results of Rahman and Prabina (2014) where they documented similar kind of results for both additions and deletions of Nifty index. ...
... This positive and negative effect got reversed after 7 days and 3 days of ED. These results are in contrast to the previous studies of Harris and Gurel (1986), Jain (1987), Lynch and Mendenhall (1997), Kumar (2007), Petajisto (2011) and Rahman and Prabina (2014) where the prices reverted in a different time frame for additions and deletions. The short-term price reversal can be evidenced from Figure 4. ...
Stock index revisions to major stock indices usually bring in changes to the price and volume
patterns of stocks getting added/deleted to and from the index. The current study analyzes
stock index revisions of companies added (deleted) to and from the CNX 100 index by testing
DSDC hypothesis and the PPH from 2004 to 2011. The results show that the price and
volume effect is permanent for inclusions and exclusions of CNX 100 index. Hence, this has
led to the support of the Downward Sloping Demand Curve hypothesis.
... ADÀ1, and continued till the day after the announcement. The AD window results depart from the previous studies of Harris and Gurel (1986), Lynch and Mendenhall (1997), Chakrabarti et al. (2005), Kumar (2007) and Petajisto (2011) wherein the abnormal returns around the AD were significantly positive (negative) for both inclusions and exclusions. ...
... These results are in contrast to the findings reported by Kumar (2007) where the prices got reversed in nine days after the ED for inclusions. Also the current results contradict with the previous studies of Harris and Gurel (1986), Jain (1987), Lynch andMendenhall (1997), andPetajisto (2011) where the prices reverted in a different time frame for inclusions and exclusions. The short-term price reversal can be evidenced from Figure 2(d). ...
Purpose
– The purpose of this paper is to test the long-term effects of price and volume with the help of Downward Sloping Demand Curve (DSDC) hypothesis, and also the short-term price and volume effects with the help of Price Pressure Hypothesis (PPH) for the index revisions on the S&P CNX Nifty 50 index.
Design/methodology/approach
– In order to report the long-term and short-term effects, the current study reviews two testable hypotheses, namely, DSDC hypothesis and PPH. The study has used the event study approach by including GARCH (1, 1) conditional variance in the market model.
Findings
– The results report that, the added stocks experienced a significant increase in price and volume on the effective date; whereas the deleted stocks experienced significant volume levels and insignificant price levels on the effective date. Accordingly, the study finds support in favor of PPH.
Research limitations/implications
– The study could not find evidence to support the most studied DSDC hypothesis.
Practical implications
– Index reorganization presumably affects the fund managers, domestic as well as international investors. As a result, studying the effect of index changes is a subject of attention to academicians and investors alike.
Originality/value
– The study contributes to the body of knowledge on index inclusion and exclusion effects by providing Indian evidence on long-term and short-term price and volume effects, and also documenting contrary results to the previous Indian and global research works.
... As a result, existing papers generally relate performance of active portfolios to academic benchmarks, implicitly assuming that passively investing in such benchmarks has zero cost and/or is the same across investors. However, even simple index strategies can have large costs, as documented by both academics (e.g., Petajisto (2011)) and practitioners (e. ...
... As a result, existing papers generally relate performance of active portfolios to academic benchmarks, implicitly assuming that passively investing in such benchmarks has zero cost and/or is the same across investors. However, even simple index strategies can have large costs, as documented by both academics (e.g., Petajisto (2011)) and practitioners (e.g., Norges Bank Investment Management 2009 annual report); Korajczyk and Sadka (2004) document even greater costs for more complex passive strategies such as momentum. While the proper measurement of costs could matter for US equity market research, it is likely to matter even more outside of the US. ...
For sophisticated institutional investors, active management outperforms passive management by more than 180 bps per year
in emerging markets and by about 50 bps in EAFE markets over the 1993 to 2008 period. In U.S. markets, active management underperforms.
Consistent with these patterns in returns, institutions use active management more frequently in non-U.S. markets, particularly
emerging markets. Finally, we provide some evidence that one contributor to the active outperformance is institutional constraints
on flows to non-U.S. markets. Overall, our results suggest that the value of active management depends on the efficiency of
the underlying market and the sophistication of the investor. (JEL G11, G14, G15, G23)
... Subsequently, stock prices reverse as they have been pushed beyond their true value at the equilibrium price. Petajisto (2011) also suggested that price pressure and trading volume are mainly driven by the sudden needs of funds rather than other investors. ...
This study aims to analyze the impact of foreign investors' abnormal trading on the abnormal returns of banking sector stocks in Vietnam from 2006 to 2023. The data sample includes 631 abnormal buy events, 571 abnormal buy events on net purchase days, 477 abnormal sell events, and 432 abnormal sell events on net sell days of foreign investors. The study employs the event study methodology with parametric and non-parametric tests. The results show that abnormal buy events and abnormal buy events on net purchase days by foreign investors convey positive information and have a positive impact on stock prices, creating positive abnormal returns, with stock prices forming new equilibrium levels. Abnormal sell events and abnormal sell events on net sell days convey negative information and have a negative impact on stock prices in the short term, after which prices tend to return to their initial states as before the events. The study suggests that investors can observe foreign investors' trading behavior as reference information in their investment decisions regarding banking sector stocks. Furthermore, it recommends that the government consider relaxing the "room" limit on foreign ownership in Vietnamese commercial banks to align with the objective of upgrading the stock market.
... For example, Biktimirov and Xu (2019b) report a 3.1% cumulative abnormal gain for additions to the Dow Jones industrial average index from the announcement day to five days after, and a -2.75% abnormal decline for deletions during the 1990-2015 period. Similarly, Petajisto (2011) finds comparable results for the small-cap Russell 2000 index, where stocks added experience an average abnormal gain of 4.7% and deletions see a loss of -4.6% from the announcement to the effective day during the period 1990-2005. Despite the extensive literature, disagreement persists regarding the underlying drivers of these stock market reactions, leading to competing hypotheses. ...
We compare the market reactions of African and non-African firms that are added to or deleted from S&P Africa 40, an index of large firms with the majority of their operations on the continent. Both African and non-African additions (deletions) experience permanent stock price gains (losses) around the S&P Africa 40 index revisions. Importantly, we find significant differences in abnormal returns between African and non-African deletions, but not additions. These differences remain robust even when controlling for institutional ownership, size, leverage, profit margin, and capital expenditure variables. Contrary to the predictions of the investor awareness hypothesis, we find no significant differences in cumulative abnormal returns between new and repeated African additions, and we observe a permanent stock price decline for new African deletions.
... Generally, funds with higher turnover ratios tend to have higher transaction costs, which can wipe out returns and lead to lower performance over time as each trade incurs costs such as brokerage charges, SEBON and NEPSE commissions, which can reduce the fund's overall returns. M., & Petajisto, (2009), find that high portfolio turnover is associated with lower risk-adjusted returns. They suggest that fund managers who trade frequently may be generating more costs than benefits for their investors, and those investors should pay attention to a fund's turnover rate when evaluating its potential for long-term success. ...
This dissertation examines the financial performance of selected mutual fund schemes having similar issue date and find out the better mutual funds among the four selected schemes. This study further assesses the impact of variability of NEPSE Index in the financial performance of these mutual fund schemes. The analysis has been done using different statistical and financial tools. Various performance measures like Sharpe ratio, Treynor ratio, Jensen Alpha, holding period return, Net Assets Value and correlation coefficient have been used to analyze the performance. The findings of this study will contribute to understanding the factors influencing mutual fund performance in Nepal, understand the impact of NEPSE Index’s movement in performance of mutual funds and offer valuable insights for investors, fund managers, and policymakers to enhance decision-making and optimize mutual fund performance in the Nepalese market.
... In a later event study, Beneish and Whaley (1996) found that, from October 1989 through June 1994, stocks included in the S&P 500 had an average increase of 3.1% in their opening price the day following the announcement. A more recent event study, by Petajisto (2011), looked at the e↵ect of inclusion and exclusion in the S&P 500 (and the Russell 2000) and found that there was an average 8.8% (4.7%) increase in the price of a newly included stock from the date of the announcement to the actual inclusion date. The e↵ect of exclusion from the S&P 500 (Russell 2000) led to a decrease of 15.1% (4.6%) following the exclusion announcement. ...
We report a laboratory experiment examining how demand for exchange-traded fund (ETF) index products affects the prices and trading volume of assets. We compare an environment where the ETF index includes all assets against an environment where a redundant asset is excluded from the index. We find that (i) subjects place significant value on the ETF index asset beyond the value of its constituent assets; (ii) there is a substantial index premium for included assets; and (iii) the index premium persists even when short selling is permitted. The price increases of the constituent assets and of the ETF itself suggest that ETF products can distort markets to some degree.
This paper was accepted by David Sraer, finance.
Funding: This work was supported the by International Foundation for Research in Experimental Economics Small Grants Program.
Supplemental Material: Data and the online appendix are available at https://doi.org/10.1287/mnsc.2022.02125 .
... In a later event study, Beneish and Whaley (1996) found that, from October 1989 through June 1994, stocks included in the S&P 500 had an average increase of 3.1% in their opening price the day following the announcement. A more recent event study, by Petajisto (2011), looked at the e↵ect of inclusion and exclusion in the S&P 500 (and the Russell 2000) and found that there was an average 8.8% (4.7%) increase in the price of a newly included stock from the date of the announcement to the actual inclusion date. The e↵ect of exclusion from the S&P 500 (Russell 2000) led to a decrease of 15.1% (4.6%) following the exclusion announcement. ...
... 30 This effect is likely driven by demand; index fund managers who replicate an index must buy the stock of each firm that is 30 Similar results have been reported in the academic finance literature since this article first appeared. For example, Harris and Gurel (1986), Whaley (1996, 2002), Lynch and Mendenhall (1997), Wurgler and Zhuravskaya (2002), and Petajisto (2011) all show effects of inclusion in the S&P 500 on stock prices. Researchers have found evidence of price effects for inclusion in other indexes, too. ...
The past couple of decades have seen a significant shift in assets from active to passive investment strategies. We examine the potential effects of this shift for financial stability through four different channels: (1) effects on investment funds’ liquidity transformation and redemption risks; (2) passive strategies that amplify market volatility; (3) increases in asset-management industry concentration; and (4) the effects on valuations, volatility, and comovement of assets that are included in indexes. Overall, the shift from active to passive investment strategies appears to be increasing some types of risk while diminishing others: The shift has probably reduced liquidity transformation risks, although some passive strategies amplify market volatility, and passive-fund growth is increasing asset-management industry concentration. We find mixed evidence that passive investing is contributing to the comovement of asset returns and liquidity.
... Lee et al. (2008) document significant abnormal returns for S&P 500 additions during the post-trading session after the announcement. Other studies that find abnormal post-announcement returns include Harris and Gurel (1986), Goetzmann andGarry (1986), Shleifer (1986), Jain (1987), and Beneish and Whaley (1996) for the S&P 500, Petajisto (2011) andChang et al. (2014) for the Russel2000, Deininger et al. (2000) for the German DAX and MDAX, Doeswijk (2005) for the AEX index, Chakrabarti et al. (2005) for international MSCI indices, Liu (2006) and Liu (2011) for the Nikkei225, Mazouz and Saadouni (2007) and Mase (2007) for the FTSE100, Qiu and Pinfold (2007) for the ASX300, and Yun and Kim (2010) for the KOSPI 200. Similar to the German DAX, many of these indices are based on a publicly available rule-based methodology. ...
Combining market data with a publicly available monthly snapshot of Deutsche Börse’s index ranking list, I create a model that predicts index changes in the DAX, MDAX, SDAX, and TecDAX from 2010 to 2019 before they are officially announced. Even though I empirically show that index changes are predictable, they still earn sizeable post-announcement 1-day abnormal returns up to 1.42% and − 1.54% for promotions and demotions, respectively. While abnormal returns are larger in smaller stocks, I find no evidence that they are related to funding constraints or additional risk for trading on wrong predictions. A trading strategy that trades according to my model yields an annualized Sharpe ratio of 0.83 while being invested for just 4 days a year.
... Sushko and Turner (2018a) 27 Similar results have been reported in the academic finance literature since this article first appeared. For example, Harris and Gurel (1986), Whaley (1996, 2002), Lynch and Mendenhall (1997), Wurgler and Zhuravskaya (2002), and Petajisto (2011) all show effects of inclusion in the S&P 500 on stock prices. Researchers have found evidence of price effects for inclusion in other indexes, too. ...
The past couple of decades have seen a significant shift in assets from active to passive investment strategies. We examine the potential effects of this shift for financial stability through four different channels: (1) effects on investment funds’ liquidity transformation and redemption risks; (2) passive strategies that amplify market volatility; (3) increases in asset-management industry concentration; and (4) the effects on valuations, volatility, and comovement of assets that are included in indexes. Overall, the shift from active to passive investment strategies appears to be increasing some types of risk while diminishing others: The shift has probably reduced liquidity transformation risks, although some passive strategies amplify market volatility, and passive-fund growth is increasing asset-management industry concentration. We find mixed evidence that passive investing is contributing to the comovement of assets. Finally, we use our framework to assess how financial stability risks are likely to evolve if the shift to passive investing continues, noting that some of the repercussions of passive investing ultimately may slow its growth.
... Petajisto (2009Petajisto ( , p. 1015 concludes that "(index and) non-index evidence points in the same direction: demand curves for individual stocks are steep in general". Petajisto (2011) finds that elasticity increases with firm size and decreases with idiosyncratic risk. With downward-sloping demand, it can be expected that index entry, by reducing the quantity available in the open market, will raise share prices and index exit will lower them. ...
In the UK (unlike the US and many other countries), companies enter and exit the main stock market index (FTSE 100) according to a clear set of rules based on market capitalisation. This creates an opportunity to game the system to secure or retain FTSE membership by manipulating capitalisation. There is considerable evidence in extant studies that index membership is beneficial, both for shareholders and managers. Hence, companies may adopt financial strategies designed to acquire or retain membership. We investigate two types of gaming. We define strategic gaming as a situation in which companies, which may initially be a number of places away from the boundary, make abnormal share issues cumulatively over several quarters. We find strong supportive evidence for this. For tactical gaming, which would involve companies in the very closest proximity to the boundary, we do not. Our analysis shows that gaming is limited to companies outside the index trying to get in. Companies that are close to exit do not game to retain their index place. The high natural volatility of market capitalisation makes success of gaming uncertain. Our central estimate is that about 5% of entries to the index appear to be the result of gaming.
... Additional evidence has piled on since these studies. Petajisto (2011) finds that the S&P 500 inclusion effects have grown since the early studies, and also shows that there are inclusion effects for the Russell 2000. Kaul, Mehrota, and Morck (2000) study a unique experiment from the Toronto Stock Exchange 300 and find more evidence of demand-induced price changes, thus extending the evidence on index inclusion effects to international markets. ...
Approximately $10 trillion is benchmarked to Morgan Stanley Capital International’s Developed, Emerging, Frontier, and standalone market indexes. Reclassifications from one index to another require thousands of investors to decide how to react. We study a comprehensive sample of past reclassifications to guide this decision. Reclassified markets’ prices substantially overshoot between the announcement and effective dates—prices fall when a market moves from an index with more benchmarked ownership to one with less, such from Emerging to Frontier, and vice-versa—but revert within a year. We identify alpha-maximizing responses to reclassifications for both tightly benchmarked and more flexible investors.
... These findings generally refute previous results reported in the literature. Such studies were conducted by Beneish and Waley (2002), Chen et al. (2004), Cai (2007), and Petajisto (2011), regarding the S&P 500; Hacibedel (2007), regarding the MSCI Emerging Market Index; Shiu and Wu (2009), regarding the MSCI Taiwan Index; and Joshipura and Janakiramanan (2015), regarding the NIFTY index. All of them generally reported a positive response for added stocks and a negative response for deleted stocks. ...
The presence of a positive (negative) price response to stock additions (deletions) to (from) an index has been getting attention owing to its deviation from the perfectly elastic demand curve of stocks. This study sheds light on this issue in the Indonesian context, using various index categories. The results intriguingly differ from prior research in most developed countries that have confirmed a positive (negative) price response to stocks added (removed) to (from) an index. Price response patterns are even opposite, which are negative (positive) for some additions (deletions), and they seem to reject the prediction of information-based hypotheses. The positive (negative) price response to (added) deleted stocks is associated with (low) high liquidity, (low) high leverage, and (high) low growth opportunity. These results seem to confirm the advantage of value premium strategy in the Indonesian capital market.
... The theme of this article is directly related to the best way to manage investment portfolios (CREMERS; PETAJISTO, 2009;FABOZZI;MARKOWITZ, 2011;FAMA;FRENCH, 2010;FERRI, 2007;PETAJISTO, 2011): actively or passively? While active portfolio managers respond to market expectations to try to obtain risk-adjusted returns that are greater than those of competitors and/or benchmark portfolios, passive managers do not respond to changes in market expectations (LARSEN;RESNICK, 1998;MAGGIN et al., 2007;VOLPERT, 2000). ...
This study considers the problem of indexing fixed-income portfolios to the ANBIMA Market Index – Series B (IMA-B), composed of Brazilian National Treasury Notes – Series B (NTNBs). We propose a mathematical model that minimizes the deviations of the returns of the chosen portfolio in relation to the returns of the index’s theoretical portfolio. The resulting model is a mathematical programming problem with convex objective function, linear constraints and free and non-negative integer variables. Five numerical examples with real data are presented to illustrate the model’s practical use. The results obtained from the fits are analyzed together with data for funds indexed to the IMA-B existing in the Brazilian financial market. The proposed method resulted in fits with optimal control of the tracking errors of the indexed portfolio.
... credit rating changes. 5 The influence of institutions managing portfolios against benchmarks has been related to various pricing anomalies, including: the influence of index inclusion or exclusion on price movements (Petajisto, 2011) and valuation ratios (Belasco et al., 2012); outperformance of low beta and low volatility stocks (Baker et al.2011); and the amplification of volatility and generation of countercyclical Sharpe ratios (Basak and Pavlova, 2013). An overview of the economic implications of index inclusion is provided by Wurgler (2010). ...
This paper outlines the benefits of long-term investing, as well as the pitfalls. Three key advantages held by long-term investors include: the capacity to adopt positions where payoff timing is uncertain; the ability to exploit opportunities generated by the actions of short-term investors; and latitude to invest in unlisted and/or illiquid assets. These advantages provide access to a broader investment opportunity set than available to short-term investors. Strategies suited to long-term investors include: capture of risk premiums arising from the actions of short-term investors; returns from liquidity provision; value investing; exploiting pricing discrepancies across segmented markets; long-term thematic investing; adding economic value to assets through engagement and control; investing in complex assets; and certain types of dynamic strategies. Pitfalls of long-term investing relate to their reliance on expectations about the long-term, when the distant future can be hard to predict; and vulnerabilities related to organizational, agency and alignment issues. Investing in illiquid assets and dynamic strategies are examined in detail.
... The fact that many funds index their performance to benchmark such as the S&P 500 also encourage funds to rebalance their portfolio when the benchmark changes, or in the case of index funds, essentially forcing them to engage in liquidity trades that creates short term price fluctuations. (Petajisto 2008) I find that a liquidity component can be extracted from quarterly change of portfolio holding of mutual funds. This factor is able to predict market return, share the properties of many commonly used proxies for liquidity. ...
This thesis uses equities holdings snapshots of mutual funds to study their trading patterns. Using quarter and semi-annual holdings of mutual funds, I am able to extract a main trading component with the application of the asymptotic principle component method. This component demonstrates short term predictability of returns over three months, suggesting overall mutual fund trades contain a liquidity trading component that temporarily pushes up stock prices that reverse over the next few months. I also demonstrates that this particular type of liquidity risk is related to other measures of liquidity risk. Therefore, this trading component can be a useful building block in creating a comprehensive measure of liquidity.
... Event studies (e.g. Harris and Gurel (1986), Shleifer (1986), Benish and Whaley (1996), Wurgler and Zhuravskaya (2002), and Petajisto (2011)) test the short-term effect of inclusion or deletion on the underlying stock price. Vijh (1994) and Wurgler (2005) examine the changes in correlations and betas of newly listed/delisted stocks and find that stocks added to S&P500 follow the index more closely and their betas are higher relative to matched similar stocks. ...
We find that passive intensity (PI), measured by the passive-linked share of to-tal stock market trading volume, is strongly related to the overall pattern of stock price movements. A one-standard deviation increase in PI is associated with 8 per-cent higher price synchronicity. We further investigate the channels through which this relation is established by separately analyzing its impact on aggregate systematic and idiosyncratic volatility of stock returns. PI has a positive effect on systematic volatility and a negative impact on firm-specific volatility. Consistent with the effect of passive-trading on price dynamics, we find evidence that PI is negatively associated with mutual funds alpha dissimilarity. After controlling for market and idiosyncratic volatility, a one-standard deviation increase in PI corresponds to a 0.20% decrease in fund dissimilarity. Our findings are robust after controlling for various macro and cor-porate factors known to affect systematic or firm-specific volatilities.. All errors are ours.
... First is the inclusion e¤ect. Petajisto (2010) shows that in the period 1990-2005, prices increased an average 8.8% around the event for stocks added to the S&P 500, and dropped -15.1% if the stocks are deleted from the index 3 . The e¤ect generally grew with the size of index fund assets. ...
This paper studies the unique rolling activity of commodity index in futures markets and shows that the resulting price impact is statistically and economically significant. Two trading strategies, devised to exploit this anomaly, yielded excess returns with positive skewness and Sharpe ratios as high as 4.39 from 2000 to March 2010. The profitability of the strategies is positively correlated with the net result of two opposite forces: the size of index investment and the amount of arbitrage capital employed. Due to the price impact, investors forwent 3.6\% annual return, 48\% higher Sharpe ratio, and billions of dollars over the period.
Transfer of stocks to a more regulated section within the same stock exchange is a quasi-natural experiment that enhances the investor base of companies. The purpose of this paper is to examine for the first time this investor base change and its price-impact. Considering the Japanese Exchange Group merger in 2012 and its structural amendments, the author uses a final sample of 181 firms between 2014-2019. An event study methodology is used to examine the abnormal returns and trading activity in relation to the investor base change proxy. The study also uses robust MM regression analysis to investigate whether the expected price-impact has is temporary or permanent. The results demonstrate that companies that had the largest positive shift in investor base also experienced the largest positive abnormal returns (+ 3.74%) and volume gains. Crucially, the author found no evidence of reversal of this price-impact, inconsistent with the price-pressure hypothesis. Instead, the increase in stock prices caused by section transfer to a more regulated section seems to be permanent. Keywords: Section transfers, More regulated section, TSE1, TSE2, Investor Base Change, Permanent price-impact.
This paper studies how the investment horizon of institutional investors affects firms' earnings management strategy in terms of trade‐off decisions between accrual management and operational adjustment. We find that when overall earnings management motives are held fixed, firms in which long‐term investors hold large stakes are more likely to manage earnings by adjusting operational decisions than by manipulating accruals. The preference for real earnings management is more pronounced when long‐term investors face performance pressures and when they have a strong influence on managers. We further document that real earnings management helps smooth earnings, and the future adverse consequences of operational adjustment are relatively less severe for firms with a greater proportion of long‐term investors than for those with more short‐term investors. Our findings are robust to the endogenous choice of ownership structure. Overall, the evidence suggests that firms choose earnings management methods to meet the earnings expectations of institutional investors, who have different earnings target windows.
The practice of periodically reconstituting equity indices suggests that changes to the composition of an index can impact the performance of firms whose stocks are added to or dropped from the index. This paper reviews and provides a comprehensive assessment of the academic literature on how changes in the composition of a stock index impacts prices, trading volume and other firm attributes. The review focuses on post-2000 contributions. It highlights and critically discusses major areas of controversy, disagreement and debate in both the empirical and methodological literature, and puts forth a promising agenda for future research. This is the first comprehensive survey of studies on the index effect and it should be particularly useful to portfolio managers and investors.
The impact of stock market index membership on REIT stock returns is unclear. Returns may become more like those of other indexed stocks and less like those of their underlying properties. Taking advantage of the inclusion of REITs in major S&P indexes starting in 2001, we find that shared index membership significantly increases the correlation between REIT returns. However, index membership also enhances the link between REIT returns and the underlying real estate, consistent with improved pricing efficiency. This article is protected by copyright. All rights reserved
This paper provides an empirical analysis of the index inclusion effect for additions to the S&P 500 index between 1981 and 2015. The analysis finds that between 1990 and 2015 the average excess return for additions from the announcement to effective day was 5.64%. The analysis goes further by exploring the average excess returns due to inclusion in each year. The paper then seeks to determine whether the magnitude of the index inclusion effect is affected by the growth in the passive management industry by regressing a number of different measures of passive holdings against the intra-year inclusion effect averages. Based on past theories, more passively linked funds should create a larger shift in the demand curves for the stocks around inclusion, but the results do not yield such conclusions. The index inclusion effect appears to have peaked in the late 1990s while the passive management industry has continued to grow.
The Russell 1000 and 2000 stock indices comprise the first 1000 and next 2000 largest firms ranked by market capitalization. Small changes in the capitalizations of firms ranked near 1000 move them between these indices. Because the indices are value-weighted, more money tracks the largest stocks in the Russell 2000 than the smallest in the Russell 1000. Using this discontinuity, we find that additions to the Russell 2000 result in price increases and deletions in price declines. We then identify time trends in indexing effects and the types of funds that provide liquidity to indexers.
A company’s entry into (or exit from) a major share index provides a special opportunity to examine price discovery. In an efficient market, we expect the demand curve to remain horizontal and to be unaffected by external events that do not communicate new information to the public, even if demand is affected. However, there is evidence that changes to index composition do impact the value of affected shares. This may be due to the price pressure generated by passively managed investment funds that simultaneously reconstitute their portfolios in order to remain aligned to the index that they are tracking. This study investigates downward sloping demand curves, price pressure and other hypotheses which are related to changes in index composition on the JSE.We calculate abnormal returns using a control portfolio model for shares entering/exiting four major FTSE/JSE indices between 2002 and 2011.In the pre-event window, a long term increasing trend was observed in the share prices of companies that are added to market cap weighted indices, beginning 70 trading days before the effective date. The opposite behaviour was true for index deletions, with some variation in the timing.In the post event window the results show, to some extent, an asymmetric response to share returns; shares entering the index underperform thereafter, whereas those leaving the index out-perform. Although these findings were not significant for all of the indices examined, they do support the price pressure hypothesis of Harris and Gurel (1986).
We ascertain the performance and viability of alternative indexation after including all relevant costs. Onerous costs, associated with forgoing the automatic rebalancing of traditional price-weighted indexes, may occur as a consequence of the frequent trading required to replicate alternatives accurately. The largest cost associated with replication is the cost of adversely moving the stock price as a result of trading the stock, known as price impact cost. It is unknown to what extent alternative index funds suffer degraded performance as a consequence of such costs. We compare the performance of a number of well-known alternative indexes before and after costs with the traditional market-capitalization benchmark index under various alternative rebalancing frequencies, considering differing assets under management, and using various competing price impact cost models in order to measure the extent of the performance erosion that occurs due to costs associated with rebalancing. We find that if we exclude all costs, the alternative indexes generate higher returns compared to the traditional benchmark, mainly as a consequence of higher risk exposures. However, as fund size and consequently the costs of rebalancing increase, the outperformance is reduced to a statistically insignificant level. In effect, we find that as assets under management and consequently rebalancing costs increase, those greater costs almost completely erode the higher returns due to the innate risk exposures of alternative indexes. This finding is robust to the choice of price impact model. We also consider alternative index viability in terms of execution and holdings and find that many alternative indexes are not viable from this perspective either. The salient lesson is that we should not ignore the implementation of alternative indexes when considering their performance. We conclude that the traditional market-capitalization-weighted index will remain popular due to its reliance on elegant theory, simplicity, ease of implementation, vast investment capacity, and inherent low costs.
Purpose
The purpose of this paper is to explore the impact of S&P 500 index fund money flow on the valuations of companies that are constituents of the index and those that are not.
Design/methodology/approach
To examine the impact of passive investing on corporate valuations, the authors run panel regressions of price‐to‐earnings ratio on aggregate money flow into S&P 500 index funds and control for various accounting variables that impact price‐to‐earnings ratio. These regressions involve two samples of stocks. The first sample consists of S&P 500 constituents. The second consists of large‐cap stocks that are not constituents of the S&P 500. The authors also run a set of separate regressions with price‐to‐book ratio rather than price‐to‐earnings ratio as the dependent variable.
Findings
It is found that the valuations of S&P 500 constituents increased by 139 to 167 basis points relative to nonconstituents, depending on valuation metric, due to S&P 500 index fund money flow when evaluated at mean values of money flow and valuation metrics. The valuations of firms within the S&P 500 index respond positively to changes in S&P 500 index fund money flow while the valuations of firms outside the index do not. Additionally, the impact of money flow on valuations persists the month after the flow occurs, suggesting that the impact does not dissipate over time.
Practical implications
Mispricings among individual stocks arising from index fund investing may reduce the allocative efficiency of the stock market and distort investors' performance evaluations of actively managed funds.
Originality/value
The paper is the first to explore the long‐run relationship between S&P 500 index fund money flow and corporate valuations.
The article presents an overview of the assumptions and unintended consequences of the widespread adoption of modern portfolio theory (MPT) in the context of the growth of large institutional investors. We examine the many so-called risk management practices and financial products that have been built on MPT since its inception in the 1950’s. We argue that the very success due to its initial insights had the unintended consequence, given its widespread adoption, of contributing to the undermining the foundation of the financial system in a variety of ways. This study has relevance for both the on-going analyses of the recent financial crisis, as well as for various existing and proposed financial reforms.
A market index summarizes the performance of a group of securities intonumber.1 The use of stock market indices in particular has been growingexponentially for years. Since Charles Dow introduced his indices in1884, the number of distinct stock market indices reported in The WallStreet Journal has increased roughly 5 percent per year, as shown inFigure 1. Today’s Journal reports not just the Dow JonesIndustrial Average (DJIA) and the S&P 500; it also reports on theTurkey Titans 20 and the Philadelphia Stock Exchange Oil Service Index.Markets are being tracked in more and more detail, and Figure 1 suggeststhat there is no end in sight.2
The author argues that equity index funds get more credit for low costs and high efficiency than they deserve. The original idea of a low–cost, low–turnover, diversified portfolio has been replaced by funds that deliver only on their pledges of low turnover and diversification. The concentration of trading around index changes causes investors to incur transaction costs far higher than the pioneers of indexing anticipated. The author concludes that index funds based on popular benchmark indexes have become a much more costly way to invest than they used to be–and than they need to be.
Because of arbitrage around the time of index changes, investors in funds linked to the S&P 500 Index and the Russell 2000 Index lose between 2.1 billion a year for the two indices combined. The losses can be higher if benchmarked assets are considered, the pre-reconstitution period is lengthened, or involuntary deletions are taken into account. The losses are an unexpected consequence of the evaluation of index fund managers on the basis of tracking error. Minimization of tracking error, coupled with the predictability and/or pre-announcement of index changes, creates the opportunity for a wealth transfer from index fund investors to arbitrageurs.
Mean reversion in stock index basis changes has been presumed to be driven by the trading activity of stock index arbitragers. We propose here instead that the observed negative autocorrelation in basis changes is mainly a statistical illusion, arising because many stocks in the index portfolio trade infrequently. Even without formal arbitrage, reported basis changes would appear negatively autocorrelated as lagging stocks eventually trade and get updated. The implications of this study go beyond index arbitrage, however. Our analysis suggests that spurious elements may creep in whenever the price-change or return series of two securities or portfolios of securities are differenced.
Since September, 1976, stocks newly included into the Standard and Poor's 500 Index have earned a significant positive abnormal return at the announcement of the inclusion. This return does not disappear for at least ten days after the inclusion. The returns are positively related to measures of buying by index funds, consistent with the hypothesis that demand curves for stocks slope down. The returns are not related to S & P's bond ratings, which is inconsistent with a plausible version of the hypothesis that inclusion is a certification of the quality of the stock.
This paper analyzes the market for financial assets in a production and exchange economy with several realized outputs and a single unobservable source of nondiversifiable risk. The paper demonstrates that, for a large class of diffusion outputs and preferences, optimizing consumers first estimate the realizations of the unobservable factor and then use these estimates to determine portfolio and consumption rules. Moreover, the explicit consideration of this unobservable productivity factor affects equilibrium demands and prices. The equilibrium spot rate of interest emerges as the “best estimate” of the unobservable factor, and multiperiod default‐free bonds arise as the optimal hedge for the unobservable changes of the stochastic investment opportunity set.
ABSTRACT Two easily measured variables, size and book-to-market equity, combine to capture the cross-sectional variation in average stock returns associated with market {3, size, leverage, book-to-market equity, and earnings-price ratios. Moreover, when the tests allow for variation in {3 that is unrelated to size, the relation between market {3 and average return is flat, even when {3 is the only explanatory variable. THE ASSET-PRICING MODEL OF Sharpe (1964), Lintner (1965), and Black (1972)
Using changes in the MSCI Standard Country Indices for 29 countries between 1998 and 2001, we document that stock returns and volumes exhibit “index effects” in international markets similar to those detected by the studies of US stocks. The stocks added to the indices experience a sharp rise in prices after the announcement and a further rise during the period preceding the actual change, though part of the gain is lost after the actual change date. The stocks that are deleted from the indices, on the other hand, witness a steady and marked decline in their prices. Trading volumes increase significantly and remain at high levels after the change date for the added stocks. There are also considerable cross-country variations in these effects. Tests using data on various measures reflecting the different hypotheses fail to turn up any evidence in support of information effects. Our evidence appears to be more supportive of the downward sloping demand curve hypothesis. There is some evidence of price-pressure and mild evidence of liquidity effect, particularly in Japan and UK.
Several empirical studies document a substantial price premium for stocks in the S&P 500 index. For index investors this creates a recurring cost: as the index is updated, they need to buy stocks with the premium and sell stocks without the premium. Different index rules can produce different index premia due to the different frequency and criteria of updating. We build a model to investigate the behavior of the index turnover cost and the portfolio performance of a mechanical index fund under a market-cap rule, an exogenous random rule, and a deterministic rule. We find that the rational anticipation of future index composition reflected in prices today eliminates any first-order differences in index fund performance across the three index rules. As the index investors become a large part of the market, the non-index investors become less diversified, and this induces hedging motives which hurt the index investors especially under a market-cap rule.
In October 1989, Standard and Poor's began its practice of pre-announcing changes in the S&P 500 index composition The new announcement practice created the "S&P game" - a trading strategy in which so-called "risk-arbitrageurs" buy the shares of newly added index stocks on the day following the announcement, and sell on the day the index change becomes effective. Because S&P 500 index funds tend to wait until the effective day to rebalance their holdings (in order to minimize tracking error) the authors show that the S&P game has been profitable on a risk-adjusted basis. The pricing inefficiency is slowly disappearing, however.
The presented paper addresses the question whether there is a stock price reaction to index replacements on the German stock market. Although it is well known from the empirical literature, which is predominantly related to the US stock market, that an index replacement has a strong impact on stock prices, it is still an open question, whether this stock price adjustment is only transitory or persistent. In order to shed new light on this issue, this paper analysed inclusions into and deletions from the two most important German stock indexes, namely the blue-chip index DAX and the mid-cap index MDAX. Unfortunately, also our evidence seems to be rather mixed with respect to the above mentioned hypotheses. We found a strong abnormal price impact on the day of the announcement of the index replacement. Stocks included in the index rose by 1.72 percent on the announcement date, while stocks deleted form the index fall by 1.19 percent on that day and by a further significant 8.76 percent up to the replacement day. Both reactions were statistically significant and they seem to be persistent, as we found no indication for a reversion in the stock price movement during the following weeks. This evidence does not fit into the price pressure hypothesis. The same is true for our finding that although absolute price reactions on the announcement date are very similar for both type of index replacements abnormal trading volume reactions are larger and more significant for index inclusions. We also found some evidence against the liquidity hypothesis. Especially, an index replacement seems not to have an impact on stock price volatility and price reactions on the announcement date are not correlated with long run volume reactions. Therefore, our evidence may support the imperfect substitute hypothesis, although we found a weakly significant correlation of price and volume reactions on the announcement date only for inclusions. Of course, the economically interesting question remains why investors and fund managers are willing to pay a premium for stocks included in an index.
The equity indexes of the Frank Russell Company are widely used as performance benchmarks for investment managers. Once a year, at the end of June, the Frank Russell Company reconstitutes its stock market indexes. We document economically and statistically significant abnormal returns associated with the annual reconstitution from 1996-2001. The cross-sectional variation in abnormal returns is explained by both permanent changes in liquidity associated with changes in index membership and temporary effects related to price pressure. Our results suggest that passive index funds pay a steep price for minimizing tracking error by rebalancing on the date of reconstitution. Conversely, there are substantial rewards to supplying immediacy to such funds. However, trading on index revisions involves risks arising from sectoral movements and from timing risks as positions are unwound. Indeed, we document dramatic return volatility on the actual day of reconstitution that reflect unanticipated order imbalances.
In this paper, we use index addition to gain insights into the trading behavior of informed traders. When a stock is added to the S&P 500 index, it attracts permanent interest from index funds, which are, by definition, liquidity traders. Consequently, this event is associated with an exogenous increase in the volume of liquidity trading, which is likely to be concentrated at the end of the day and in large-sized trades. These aspects of fund trading, in turn, change the optimal strategies of informed traders. We can characterize the resulting equilibrium by comparing 'steady state' trading patterns before and after index addition. Our key findings are as follows. First, while both daily and intraday volume increase substantially after index addition, the largest increase occurs in the last half-hour of the day. There is a permanent increase in daily and intraday return volatility after addition; once again, the largest increase occurs at the end of the day. The increase in return volatility throughout the day suggests that increased information production follows index addition. However, the larger increase in end-of-day volatility also suggests that informed traders change the timing of their trades to coincide with liquidity trades. Second, consistent with the increased presence of index funds, the distribution of trade sizes shifts rightwards after addition and this shift is most marked at the end of the day. The price impact of individual large trades declines, though the magnitude of the decline is similar through the trading day. The increased frequency of large liquidity trades at the end of the day creates incentives for informed traders to make large trades, since they are now able to 'hide' these trades more easily. Consequently, the proportion of total end-of-day price change occurring on large trades increases significantly after addition. Additional tests reject temporary price pressure from large trades as an alternative explanation for our results. Our results support the conclusion that informed traders alter their trading strategies to optimally match both the timing, and the size, of liquidity trades in the market.
Representative agent models are inconsistent with existing empirical evidence for steep demand curves for individual stocks. This paper resolves the puzzle by proposing that stock prices are instead set by two separate classes of investors. While the market portfolio is still priced by individual investors based on their collective risk aversion, those individual investors also delegate part of their wealth to active money managers who use that capital to price stocks in the cross-section. In equilibrium the fee charged by active managers has to equal the before-fee alpha they earn; this endogenously determines the amount of active capital and the slopes of demand curves. A calibration of the model reveals that demand curves can indeed be steep enough to match the magnitude of many empirical findings, including the price effects for stocks added to (or deleted from) the S&P 500 index.
Excess returns of S&P index replacement stocks are attributed to price pressures and imperfect substitutes in previous research. However, parameter estimates are biased by the use of pre-announcement returns; replacements are characterized by rising stock prices. Using a future estimation period to avoid this bias, we find excess returns do not reverse. Further, we find no relation between excess returns and the amount of stock closely held or the size of index funds. The evidence supports efficient market assumptions: the stock market is liquid and stocks are close substitutes.
I describe asset price dynamics caused by the slow movement of investment capital to trading opportunities. The pattern of price responses to supply or demand shocks typically involves a sharp reaction to the shock and a subsequent and more extended reversal. The amplitude of the immediate price impact and the pattern of the subsequent recovery can reflect institutional impediments to immediate trade, such as search costs for trading counterparties or time to raise capital by intermediaries. I discuss special impediments to capital formation during the recent financial crisis that caused asset price distortions, which subsided afterward. After presenting examples of price reactions to supply shocks in normal market settings, I offer a simple illustrative model of price dynamics associated with slow-moving capital due to the presence of inattentive investors. Copyright (c) 2010 the American Finance Association.
We consider the problem of maximizing terminal utility in a model where asset prices are driven by Wiener processes, but where
the various rates of returns are allowed to be arbitrary semimartingales. The only information available to the investor is
the one generated by the asset prices and, in particular, the return processes cannot be observed directly. This leads to
an optimal control problem under partial information and for the cases of power, log, and exponential utility we manage to
provide a surprisingly explicit representation of the optimal terminal wealth as well as of the optimal portfolio strategy.
This is done without any assumptions about the dynamical structure of the return processes. We also show how various explicit
results in the existing literature are derived as special cases of the general theory.
KeywordsPortfolio-Optimal control-Filtering-Partial information-Stochastic control-Partial observations-Investment
Mathematics Subject Classification (2000)49N30-60H30-93C41-91G10-91G80
Testing the two-parameter asset pricing theory is difficult (and currently infeasible). Due to a mathematical equivalence between the individual return/‘beta’ linearity relation and the market portfolio's mean-variance efficiency, any valid test presupposes complete knowledge of the true market portfolio's composition. This implies, inter alia, that every individual asset must be included in a correct test. Errors of inference inducible by incomplete tests are discussed and some ambiguities in published tests are explained.
The paper investigates how quickly prices attain new equilibrium levels after large-block transactions, and measures the associated temporary and permanent price effects. We find that prices adjust within at most three trades, with most of the adjustment occurring in the first trade. The temporary price effect for seller-initiated transactions is related to block size, but the temporary price effect observed for buyer-initiated transactions is no larger than that observed in 100 share trades. Most of the price effect associated with block trades is permanent and is related to block size, regardless of the initiating party.
I develop a framework to analyze demand curves for multiple risky securities at extended horizons in a setting with limits-to-arbitrage. Following an unexpected change in uninformed investor demand for several assets, I predict returns of each security to be proportional to the contribution of that security's demand shock to the risk of a diversified arbitrage portfolio. I show that securities that are not affected by demand shocks but are correlated with securities undergoing changes in demand should experience returns related to their hedging role in arbitrageurs' portfolios. Finally, I predict a negative cross-sectional relation between post-event returns and the initial return associated with the change in demand. I confirm these predictions using data from a unique redefinition of the Nikkei 225 index in Japan, in which 255 stocks simultaneously undergo significant changes in index investor demand, causing more than ¥2,000 billion of trading in one week and large price changes followed by subsequent reversals for all of the reweighted stocks.
The incomplete information financial economic equilibrium (IIE) literature has been growing at an increasing rate since its inception in the early 1980s. This paper examines issues and concepts essential to understanding, implementing, and testing IIE and understanding its relation to complete information equilibria (CIE). Concepts include the number of state variables in an IIE vis-a-vis the number of state variables in a corresponding CIE; the irrelevance of separation theorems to IIE and the relevance, instead, of a more general state space (re-)representation theorem; the identification of unobservable productivity processes that lead to complete information; the relative level of variable variances in a CIE and the corresponding IIE; stochastic CIE with corresponding deterministic IIE and deterministic CIE with corresponding stochastic IIE; the relationship between IIE and incomplete markets; the (im)persistence of heterogeneous beliefs; and the relation of IIE to the model uncertainty/ambiguity approaches. Understanding these concepts under IIE facilitates understanding the CIE, a special case of IIE.
Stocks added to the S&P 500 generally experience positive abnormal returns following the announcement. Several competing explanations exist for this reaction, but small sample sizes and other issues make it difficult to distinguish among them. We examine this subject using the small-cap Russell 2000 index, which has several advantages over the S&P 500 in this context. Our primary finding is that stocks added to or deleted from the Russell 2000 experience significant changes in stock price and trading volume, but the effect is transitory. The results support the price pressure hypothesis. 2004 The Southern Finance Association and the Southwestern Finance Association.
We introduce a new measure of active portfolio management, Active Share, which represents the share of portfolio holdings that differ from the benchmark index holdings. We compute Active Share for domestic equity mutual funds from 1980 to 2003. We relate Active Share to fund characteristics such as size, expenses, and turnover in the cross-section, and we also examine its evolution over time. Active Share predicts fund performance: funds with the highest Active Share significantly outperform their benchmarks, both before and after expenses, and they exhibit strong performance persistence. Nonindex funds with the lowest Active Share underperform their benchmarks. The Author 2009. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org, Oxford University Press.
Bibliography: leaves 53-56. Robert C. Merton.
We use 2 years of daily flows for three major Standard and Poor's index funds to analyze the relationship among index funds, asset prices, and volatility. We find strong contemporaneous correlation between inflows and returns, no evidence for positive feedback trading, and evidence that negative market returns may induce subsequent sales. Market volatility affects investors as dynamic risk sharing, but higher volatility does not drive investors from the market. Bullish newsletter sentiment is associated with greater inflows. We report high correlation among investor disagreement and market uncertainty and flows. Dispersion in advice and open interest correlate with lower inflows.
In textbook theory, demand curves for stocks are kept flat by riskless arbitrage between perfect substitutes. In reality, however, individual stocks do not have perfect substitutes. We develop a simple model of demand curves for stocks in which the risk inherent in arbitrage between imperfect substitutes deters risk-averse arbitrageurs from flattening demand curves. Consistent with the model, stocks without close substitutes experience higher price jumps upon inclusion into the S&P 500 Index. The results suggest that arbitrage is weaker and mispricing is likely to be more frequent and more severe among stocks without close substitutes.
Since October 1989, Standard and Poor's has (when possible) announced changes in the composition of the S&P 500 index one week in advance. Because index funds hold S&P 500 stocks to minimize tracking error, index composition changes since this date provide an opportunity to examine the market reaction to an anticipated change in the demand for a stock. Using post-October 1989 data, the authors document significantly positive (negative) postannouncement abnormal returns that are only partially reversed following additions (deletions). These results indicate the existence of temporary price pressing and downward-sloping long-run demand curves for stocks and represent a violation of market efficiency. Copyright 1997 by University of Chicago Press.
In this article, we consider the possibility that some liquidity traders preannounce the size of their orders, a practice
that has come to be known as “sunshine trading”. Two possible effects preannouncement might have on the equilibrium are examined.
First, since it identifies certain trades as informationless, preannouncement changes the nature of any informational asymmetries
in the market. Second, preannouncement can coordinate the supply and demand of liquidity in the market. We show that preannouncement
typically reduces the trading costs of those who preannounce, but its effects on the trading costs and welfare of other traders
are ambiguous. We also examine the implications of preannouncement for the distribution of prices and the amount of information
that prices reveal.
Anecdotal accounts imply that founding families routinely engage in opportunistic activities that exploit minority shareholders. We gauge the severity of these moral hazard conflicts by examining whether founding families--as large, undiversified blockholders--seek to reduce firm-specific risk by influencing the firm's diversification and capital structure decisions. Surprisingly, we find that family firms actually experience less diversification than, and use similar levels of debt as, nonfamily firms. Consistent with these findings, we also find that direct measures of equity risk are not related to founding-family ownership, which suggests that family holdings are not limited to low-risk businesses or industries. Although founding-family ownership and influence are prevalent and significant in U.S. industrial firms, the results do not support the hypothesis that continued founding-family ownership in public firms leads to minority-shareholder wealth expropriation. Instead, our results show that minority shareholders in large U.S. firms benefit from the presence of founding families.
This paper investigates whether abnormal returns permanently exist in transparent U.S. Russell index reconstitution and provides evidence to disentangle the competing hypotheses associated with the index effect in the literature. Additions to Russell 1000 generate cumulative excess returns of 10.9% from 2 days before May 31 to June 30 while stocks deleted from Russell 2000 Growth Index suffer cumulative loss of 6.6%. The effect of index reconstitution on stocks in the style switching groups is moderate while it is much smaller for stocks in the retention groups. Based on daily trading volume, there is evidence that money managers tied to Russell style indexes tend not to rebalance their portfolios actively until the time of index reconstitution to avoid tracking error. However, for stocks generating large excess returns, money managers trade them actively prior to the reconstitution. This study is supportive of the imperfect substitutes hypothesis in explaining the index effect, given the absence of complete reversal of the event period abnormal returns and of consistent improvement in liquidity for the index additions. In the joint test, the price pressure hypothesis and the liquidity hypothesis explain the marginal index effect at most by 0.12% and 3.05%, respectively, while the imperfect substitutes hypothesis explains it at least by 9.21%. Furthermore, the index effect is not purely driven by individual stock price momentum. Copyright Springer Science + Business Media, LLC 2006
We study the price effects of changes to the S&P 500 index and document an asymmetric price response: There is a permanent increase in the price of added firms but no permanent decline for deleted firms. These results are at odds with extant explanations of the effects of index changes that imply a symmetric price response to additions and deletions. A possible explanation for asymmetric price effects arises from the changes in investor awareness. Results from our empirical tests support the thesis that changes in investor awareness contribute to the asymmetric price effects of S&P 500 index additions and deletions. Copyright 2004 by The American Finance Association.
There is an exact linear relation between expected returns and true 'betas' when the market portfolio is on the ex ante mean-variance efficient frontier but empirical research has found little relation between sample mean returns and estimated betas. A possible explanation is that market portfolio proxies are mean-variance inefficient. The authors categorize proxies that produce particular relations between expected returns and true betas. For the special case of a zero relation, a market portfolio proxy must lie inside the efficient frontier but it may be close to the frontier. Copyright 1994 by American Finance Association.
This paper analyzes an economy in which investors operate under partial information about technology‐relevant state variables. It is shown that for Gaussian information structures under incomplete observations, the consumer's problem can be transformed into an equivalent program with a completely observed state: the conditional expectation of the underlying unobservable state variables. A consequence of this transformation is that classic results in finance remain valid under an appropriate reinterpretation of the state variables.
This paper documents the effects of large (block) transactions on the prices of common stocks traded on the New York Stock Exchange. We examine whether mean temporary and permanent price effects associated with large and small transactions differ and whether the price effects vary cross-sectionally according to the size of the block. Alternative definitions of block size are investigated – percentage of the equity traded, block volume in relation to normal trading volume, and dollar value of the block. The results suggest that price effects are predominantly temporary for seller-initiated transactions and permanent for buyer-initiated transactions.
Examines the arbitrage model of capital asset pricing as an alternative to the mean variance capital asset pricing model introduced by Sharpe, Lintner and Treynor. Overview of the arbitrage theory; Role of the arbitrage model in explaining phenomena observed in capital markets for risky assets; Influence of the presence of noise on the pricing relation. (Из Ebsco)
This study analyzes the effects of changes in S&P 500 index composition from January 1986 through June 1994, a period during which Standard and Poor's began its practice of preannouncing changes five days beforehand. The new announcement practice has given rise to the 'S&P game' and has altered the way stock prices react. The authors find that prices increase abnormally from the close on the announcement day to the close on the effective day. The overall increase is greater than under the old announcement policy, although part of the increase reverses after the stock is included in the index. Copyright 1996 by American Finance Association.
Mean reversion in stock index basis changes has been presumed to be driven by the trading activity of stock index arbitragers. The authors propose here instead that the observed negative autocorrelation in basis changes is mainly a statistical illusion, arising because many stocks in the index portfolio trade infrequently. Even without formal arbitrage, reported basis changes would appear negatively autocorrelated as lagging stocks eventually trade and get updated. The implications of this study go beyond index arbitrage, however. The authors' analysis suggests that spurious elements may creep in whenever the price-change or return series of two securities or portfolios of securities are differenced. Copyright 1994 by American Finance Association.
Two easily measured variables, size and book-to-market equity, combine to capture the cross-sectional variation in average stock returns associated with market "beta", size, leverage, book-to-market equity, and earnings-price ratios. Moreover, when the tests allow for variation in "beta" that is unrelated to size, t he relation between market "beta" and average return is flat, even when "beta" is the only explanatory variable. Copyright 1992 by American Finance Association.
The authors study the price elasticity of demand for the common stock of an individual corporation. Despite the prevalance of assumptions that demand is perfectly elastic, there is little, if any, direct evidence in the literature to either support or reject that contention. Consistent with the notion of finite price elasticities, the authors find that the announcement of primary stock offerings by regulated firms depresses their stock prices and little, if any, evidence that this decline is the result of adverse information about future cash flows. Attempts to relate offer announcement effects directly to possible determinants of price elasticities, however, are inconclusive. Copyright 1991 by American Finance Association.
This paper examines the effects of the mechanism by which securities are traded on their price behavior. We compare the behavior of open‐to‐open and close‐to‐close returns on NYSE stocks, given the differences in execution methods applied in the opening and closing transactions. Opening returns are found to exhibit greater dispersion, greater deviations from normality and a more negative and significant autocorrelation pattern than closing returns. We study the effects of the bid‐ask spread and the price‐adjustment process on the estimated return variances and covariances and discuss the associated biases. We conclude that the trading mechanism has a significant effect on stock price behavior.
Weights in the Toronto Stock Exchange 300 index are determined by the market values of the included stocks' public floats. In November 1996, the exchange implemented a previously announced revision of its definition of the public float. This revision, which increased the floats and the index weights of 31 stocks, conveyed no information and had no effect on the legal duties of shareholders. Affected stocks experienced statistically significant excess returns of 2.3 percent during the event week, and no price reversal occurred as trading volume returned to normal levels. These findings support downward sloping demand curves for stocks. An obvious event with which to examine the slope of demand curves for stocks is one that changes supply. In the absence of new information, a shift in supply should not affect stock prices if demand curves for stocks are flat. Scholes (1972), using a sample of secondary equity distributions, asks whether stocks are "...unique works of art..." or merely ...
In the present paper we analyze the relationship between index funds and asset prices. In particular, our analysis of daily index fund flows indicates a strong contemporaneous correlation between fund inflows and S&P market returns. We also document a strong negative correlation between fund out-flows and S&P market returns with the exception of outflows from a fund with very high initial investment requirement. These effects may be interpreted in two ways. Either investor supply and demand affects S&P market prices, or investors condition their demand and supply on intra-day market fluctuations. To sort out these effects, we examine trailing investor reaction to market moves. Our results suggest the market reacts to daily demand. However, only negative reactions appear due to past returns. We investigate whether index investor demand shocks ("liquidity shocks") are permanent or temporary and provide evidence that supports the hypothesis that they are permanent. This result may help ex...
On S&P 500 Index Replication Strategies
- M E Blume
- R M Edelen
Blume, M.E. and R.M. Edelen, 2001, "On S&P 500 Index Replication Strategies,"
working paper, University of Pennsylvania.
- J Wurgler
- K Zhuravskaya
Wurgler, J. and K. Zhuravskaya, 2002, " Does Arbitrage Flatten Demand Curves for
Stocks? " Journal of Business, vol. 75, no. 4, 583-608.
7.3%
61
8.2%
174
2.3%
74
15.5%
1991
403
18.3%
78
8.8%
344
2.3%
67
16.8%
1992
439
17.4%
81
10.2%
388
3.8%
56
12.3%
1993
344
13.0%
93
10.6%
314
3.9%
60
13.0%
1994
462
18.1%
97
10.0%
415
6.4%
65
13.3%
1995
363
11.6%
67
6.9%
251
3.9%
79
13.5%
1996
413
14.1%
106
10.9%
312
4.8%
91
16.4%
1997
436
14.3%
94
7.3%
320
4.7%
101
15.8%
1998
427
14.0%
97
8.9%
243
4.2%
108
17.7%
1999
412
13.7%
101
10.2%
250
3.9%
124
23.8%
2000
546
20.0%
137
12.6%
334
4.8%
117
26.7%
2001
501
12.3%
108
8.3%
279
1.8%
152
26.1%
2002
382
9.0%
104
7.8%
233
1.9%
138
23.6%
2003
283
6.3%
82
9.7%
187
1.6%
94
15.7%
2004
310
6.8%
67
8.1%
214
2.5%
67
11.7%
2005
206
4.1%
80
8.7%
215
2.1%
89
14.3%
1990-2000
412
14.7%
92
9.5%
304
4.1%
86
16.8%
2001-2005
336
7.7%
88
8.5%
226
2.0%
108
18.3%
1990-2005
388
12.5%
91
9.2%
280
3.4%
93
17.3%
- P A Gompers
- A Metrick
Gompers, P.A. and A. Metrick, 2001, "Institutional Investors and Equity Prices,"
Quarterly Journal of Economics, 229-259.
An Anatomy of the 'S&P Game': The E¤ect of Changing the Rules
- M D Beneish
- R E Whaley
Beneish, M.D. and R.E. Whaley, 1996, "An Anatomy of the 'S&P Game': The E¤ect
of Changing the Rules," Journal of Finance, vol. 51, no. 5, 1909-1930.
- J Wurgler
- K Zhuravskaya
Wurgler, J. and K. Zhuravskaya, 2002, "Does Arbitrage Flatten Demand Curves for
Stocks?" Journal of Business, vol. 75, no. 4, 583-608.
1992
439
17.4%
81
10.2%
388
3.8%
56
12.3%
1993
344
13.0%
93
10.6%
314
3.9%
60
13.0%
1994
462
18.1%
97
10.0%
415
6.4%
65
13.3%
1995
363
11.6%
67
6.9%
251
3.9%
79
13.5%
1996
413
14.1%
106
10.9%
312
4.8%
91
16.4%
1997
436
14.3%
94
7.3%
320
4.7%
101
15.8%
1998
427
14.0%
97
8.9%
243
4.2%
108
17.7%
1999
412
13.7%
101
10.2%
250
3.9%
124
23.8%
2000
546
20.0%
137
12.6%
334
4.8%
117
26.7%
2001
501
12.3%
108
8.3%
279
1.8%
152
26.1%
2002
382
9.0%
104
7.8%
233
1.9%
138
23.6%
2003
283
6.3%
82
9.7%
187
1.6%
94
15.7%
2004
310
6.8%
67
8.1%
214
2.5%
67
11.7%
2005
206
4.1%
80
8.7%
215
2.1%
89
14.3%
1990-2000
412
14.7%
92
9.5%
304
4.1%
86
16.8%
2001-2005
336
7.7%
88
8.5%
226
2.0%
108
18.3%
1990-2005
388
12.5%
91
9.2%
280
3.4%
93
17.3%