You are on page 1of 18

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0307-4358.htm

The impact
The impact of passive investing of passive
on corporate valuations investing
Eric Belasco
Department of Agricultural Economics and Economics, 1067
Montana State University, Bozeman, Montana, USA
Michael Finke Received August 2012
Revised January 2012,
Department of Personal Financial Planning, Texas Tech University, March 2012
Lubbock, Texas, USA, and Accepted May 2012

David Nanigian
The American College, Bryn Mawr, Pennsylvania, USA

Abstract
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.
Keywords Investments, Indexing, Fund management, Active management, Passive management,
Index fund, Index premium, Demand curves for stocks, S&P 500
Paper type Research paper

JEL classification G11 (investment decisions), G12 (asset pricing), G23 (private financial
institutions)
The authors appreciated the feedback given by Ozzy Akay, David Blitzer, Dale Domian,
Harold Evensky, Scott Hein, Jason Hsu, Vitali Kalesnik, Masha Rahnama, John Salter, William
Waller, and Walter Woerheide, and the excellent research assistance provided by Melissa Cenneno.
Managerial Finance
The authors also thank seminar participants at Texas Tech University (hosted by the Division of Vol. 38 No. 11, 2012
Personal Financial Planning) and The American College, as well as participants at the Academy of pp. 1067-1084
q Emerald Group Publishing Limited
Financial Services Annual Meeting, the Financial Management Association International Annual 0307-4358
Meeting, and the Journal of Indexes Editorial Board Meeting for their comments and suggestions. DOI 10.1108/03074351211266793
MF I. Introduction
38,11 Neoclassical asset pricing theory assumes that the price of a stock changes as a result
of new information about a corporations value to its shareholders. Event studies
on changes to the composition of an index are frequently used to examine whether a
non-valuation motivated change in the quantity demanded of a given security results in
a change in its price. While these studies suggest an increase in valuation resulting from
1068 inclusion in the index, it is possible that the increase accurately reflects greater intrinsic
value of the stock from increased liquidity. This study explores 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. We find that money flow increases the price-to-earnings and
price-to-book ratios of companies that are constituents relative to those that are not. This
provides evidence of downward-sloping demand curves for stocks.
In a broad review of studies investigating indexation price effects, Brealey (2000)
identifies a trend over the period 1966 to 1995 toward larger positive abnormal returns
following the addition of a stock to the S&P 500 and larger negative abnormal returns
following a deletion of a stock from the index. Petajisto (2011) finds that the average
abnormal returns to additions are 8.8 percent in the period 1990 through 2005, and the
average abnormal returns to deletions are 2 15.1 percent. The percentage of US equity
mutual fund assets invested in index funds increased from 1.0 percent in 1984 to 12.4
percent in 2002 (French, 2008), and more than half of all US equity index fund assets
are invested in S&P 500 index funds (Investment Company Institute, 2011). This
increase in passive equity investment among S&P 500 firms suggests that the growth
of the event effects may be largely due to the shift toward passive investing.
The slope of a demand curve for a companys stock is gauged by the (absolute)

magnitude of its price elasticity of demand, E DP . In the stock market setting:
DQ=Q
D ;
E P 1
DP=P
where Q denotes the quantity of shares of a companys stock demanded by investors and
P denotes the price per share of the companys stock.
This concept is shown in Figure 1.

Shleifer (1986) and Petajisto (2011) estimate that E DP is near unity. This runs contrary to
the predictions of Famas (1970) efficient market hypothesis, which predicts that
arbitrage trading will correct mispricings among stocks[1]. Arbitrageurs have not
flattened the demand curve. Petajisto (2011) also discovers that in the period 1990
through 2005 the average abnormal returns to additions to the Russell 2000 was
8.0 percent and the average abnormal returns to deletions was 2 13.4 percent. Price
effects in the purely market cap-based Russell 2,000 that are similar to those observed in
the S&P 500 provide further evidence of downward-sloping demand curves for stocks.
In a related strand of research, Baberdis et al. (2005) theorize that, in the presence of
limited arbitrage, money flowing into a segment of the market impacts the correlation
of returns between stocks in that segment. Consistent with the authors habitat theory
of return comovement, they find that a stocks correlation with other stocks in the S&P
500 increases when it is added to the index and, commensurate with the trend towards
S&P 500 index fund investing, the correlations have increased in recent years.
Relatively higher stock ownership by active funds reduces the magnitude of the
correlation between stocks and the index; however, index additions result in increased
S The impact
D0 D1 of passive
investing
Price (per share of stock)

P1
1069

P0

Quantity (Volume of stock)


Note: This figure shows how the price of a stock responds to a
change in the quantity of the stock demanded by market participants Figure 1.
Supply and demand
when the demand curve is downward-sloping

passive ownership share and a decrease in firm-specific price changes (Ye, 2011).
Wermers and Yao (2011) find stronger pricing anomalies, such as those related to
momentum and accruals, among stocks with high levels of passive fund ownership,
which provides additional evidence that index fund investing impedes the efficiency of
capital markets. Increasing correlation among funds within the S&P 500 and pricing
anomalies associated with increased passive investment may counteract alternative
explanations for the gains from S&P 500 inclusion, including reduced agency costs
from increased monitoring by analysts (Yu, 2008).
Event studies only partially address price effects associated with index fund investing.
This is because investor cash is continuously flowing in and out of index funds, yet
changes to the composition of an index only occur a few times a year. Changes in the
quantity of a stock demanded by index fund managers as a result of fund flows may also
impact the price of stocks if stocks within the index exhibit downward-sloping demand
curves. Goetzmann and Massa (2003) examine the relationship between index fund flows
and returns on the index and find that a strong same-day relationship exists. However, it is
difficult to determine whether the flows are driving the returns or vice versa. To
disentangle demand effects from potential feedback effects, the authors perform a
Geweke-Messe-Dent (1982) (GMD) test. The GMD test confirms that the direction of
causality is from flows to returns. However, the GMD test is weak if the causality is at a
higher frequency than the data. This motivated the authors to conduct a series of tests
using higher frequency data, which arrive at the same results as the GMD test. Hence,
strong evidence suggests that flows drive returns. However, the question of whether or not
S&P 500 index fund money flow has inflated the fundamental value of companies that are
constituents of the index relative to nonconstituents has not yet been determined.
MF We hypothesize that flows into S&P 500 index funds positively impact the PE ratio
38,11 of companies that are in the index relative to those that are not. In an unpublished
manuscript, Morck and Yang (2002) find evidence that money flow into index funds
alters the Tobins Q ratios. Our empirical work is an extension of Morck and Yang and
is different in several ways. First, we perform regressions that involve net cash flows
(NCF) rather than time-series plots. Second, we take a comprehensive view of index
1070 fund assets by looking at all S&P 500 index funds, including exchange traded funds
(ETF). Third, we focus on ratios that are more commonly used by practitioners to
gauge valuation. Fourth, we use more recent data. This paper examines the extent to
which price multiple measures of valuation are impacted by the trend toward passive
investing. To conduct this analysis, we run panel regressions of price-to-earnings ratio
(PE) on aggregate NCF into S&P 500 index funds controlling for various accounting
variables that impact PE. Additionally, regressions are run with price-to-book ratio
(PB), an alternative valuation measure, as the dependent variable[2].
The results from regression models that control for aggregate US equity fund NCF
demonstrate that the valuations of constituents increased by 139 to 167 basis points
relative to nonconstituents, depending on valuation metric, due to S&P 500 index fund
NCF when evaluated at mean value of NCF and valuation metrics. Valuations of
firms within the S&P 500 index respond positively to changes in S&P 500 NCF while
the values of firms outside the index do not increase during the same time period.
Additionally, flow induced price-pressure effects do not appear to dissipate over time.

II. Data
To maximize coverage of index fund assets, we incorporate both mutual fund and ETF
data into our analysis. Mutual fund and ETF data come from Morningstar Direct.
Index and individual stock data come from Compustat. Because Morningstars
coverage of fund flows begins in February 1993, we control for lagged one-month fund
flows, and because Compustats price-dividends-earnings file ends in February 2007,
the sample period spans March 1993 to February 2007.

A. Description of data on fund flows


Monthly NCFs are gathered from Morningstar Directs fund flows database on all
US-domiciled open-end mutual funds and ETFs excluding money market funds and
funds of funds[3]. We gather two time-series of NCFs. The first consists of all S&P 500
index funds which, based on conversations with Morningstar, is defined as all funds
with an institutional category of S&P 500 Tracking. The second consists of all US
equity funds and is used to control for market-wide fund NCF in our regression
analysis; it consists of all funds with a US broad asset class of US Stock.

B. Description of data on S&P 500 index valuation


For each month in the sample, information on month-end closing prices, earnings
per share, book value per share, and a variable indicating historical S&P 500 index
constituency status is gathered from Compustat on all publicly-traded large-cap
companies. Following Fama and French (1996), we define large-cap as greater than
median NYSE market capitalization and gather monthly market capitalization
breakpoints from Kenneth Frenchs web site. We restrict our analysis to only large-cap
companies in order to prevent the size premia from biasing our results. We exclude
non-operating establishments, financial services companies (due to non-reporting of The impact
working capital data), and companies headquartered outside of the USA (ADRs, ADSs, of passive
AMs, and GDRs)[4] from our analysis. We define earnings per share as 12 month
moving average quarterly as reported basic earnings per share. We define book value investing
per share as the total value of common equity excluding intangible assets based on
fiscal year end data divided by common shares outstanding.
Additionally, for each month in the sample, data on various accounting line items 1071
are gathered from Compustat to control for return on invested capital (ROIC), free
cash flow (FCF) growth, and the internal growth rate of dividends, which are
commonly regarded to impact price multiple ratios[5]. To control for seasonality, all
income statement and statement of cash flow items are defined as the cumulative
value of the associated variables over the prior four-quarters. Balance sheet items are
defined as the associated value as of the most recent quarter. Companies lacking
four-quarters of continuous data in a given month are excluded from analysis for that
period. Line item values that are coded by Compustat as not meaningful or
insignificant are imputed as zero.
Following Damodaran (2007), ROIC is calculated as after-tax operating income
divided by the book value of lagged four-quarter invested capital, where after-tax
operating income is calculated as net income plus after-tax interest expense minus
after-tax non-operating income. The applicable tax rate is assumed to be the maximum
marginal tax rate in the associated time period. Based on information provided by the
Urban Institute and the Brookings Institution (2010), the effective tax rate was
34 percent prior to 1993 and 35 percent from 1993 through 2007.
Following Brealey et al. (2006), FCF is calculated as net income plus depreciation
and amortization expense minus the change in working capital minus capital
expenditures. In the interest of preserving degrees of freedom, FCF Growth is defined
as the percentage change in FCFs between the current and lagged fourth quarter.
Following Brealey et al. (2006), the internal growth rate of dividends is calculated as
the product of the retention ratio (also known as the plowback ratio) and return
on equity. The retention ratio is equal to one minus the indicated annual dollar value of
dividends paid to common shareholders divided by income before extraordinary items
(adjusted for common stock equivalents). The return on equity is calculated as income
before extraordinary items divided by lagged four-quarter book value of equity.
To control for the effect of outliers and data entry errors from the data feeds, all
accounting variables are winsorized (i.e. truncated) at the 1 percent tails.

III. Methods
A series of least-squares regressions is used to analyze the impact of money flow into
S&P 500 index funds on the valuation of individual companies. Our choice of a panel
study involving individual companies in contrast to time-series analysis involving
portfolios of stocks was driven by two rationales. First, the panel study approach
allows us to better control for effects across the cross-section and time domains.
Second, more information is utilized through the panel approach as our results are not
limited to a single portfolio or a moment in time.
We employ a fixed effects model with a one-way error term in our regression analysis
to account for company-specific individual effects. The use of this model is supported by
the results from the Hausmans specification test for fixed versus random effects.
MF A. Description of regression model
38,11 To test our hypothesis, we estimate regressions that are variants of the following form:

PE i;t ai;t b1 ROIC i;t b2 FCF growthi;t b3


2
Expected growth rate of dividendsi;t b4 Aggregate
1072 where subscript i denotes the individual stock and subscript t denotes the month. We
also run a set of regressions with an alternative dependent variable, PB.

IV. Results
Table I presents summary statistics on the main variables of interest. NCFs are
reported in billions of dollars. It is interesting to note that S&P 500 index fund NCFs
are relatively more volatile than those of the aggregate market of US equity funds. For
example, the standard deviation of S&P 500 index fund NCF is 2.25 times greater than
its mean value, yet the standard deviation of aggregate US equity fund NCF is only
0.90 times the size of its mean value. The factors that affect aggregate US equity fund
NCF, such as aggregate demand for liquidity among market participants, also affect
S&P 500 index fund NCF. However, the Pearson correlation coefficient between the two
money flow variables is only 0.55. The correlation coefficient is not indicative of
multicollinearity, suggesting that the increased popularity of indexing over our sample
period, in addition to market-wide factors, contribute to the high volatility of S&P 500

Mean SD 25th percentile Median 75th percentile

Panel A: net cash flow variables (T 168)


S&P 500 index fund net cash flow 0.97 2.18 0.14 0.88 2.06
US equity fund net cash flow 8.22 7.43 4.07 8.44 12.20
Panel B: financial variables constituents sample PE regressions (n 622 )
Price-to-earnings ratio 21.03 39.19 13.95 19.61 28.23
Return on invested capital 0.16 0.38 0.06 0.12 0.19
Free cash flow growth (1.23) 4.94 (2.07) (1.23) (0.41)
Expected growth rate of dividend stream 0.17 0.26 0.08 0.16 0.24
Panel C: financial variables nonconstituents sample PE regressions (n 1,864 )
Price-to-earnings ratio 22.20 47.13 13.47 20.69 32.04
Return on invested capital 0.15 0.38 0.06 0.12 0.19
Free cash flow growth (1.23) 5.79 (2.34) (1.40) (0.38)
Expected growth rate of dividend stream 0.15 0.31 0.08 0.16 0.25
Panel D: financial variables constituents sample PB regressions (n 619 )
Price-to-book ratio 4.25 4.30 2.03 3.07 4.91
Return on invested capital 0.16 0.39 0.06 0.12 0.19
Free cash flow growth (1.23) 4.97 (2.08) (1.24) (0.41)
Expected growth rate of dividend stream 0.16 0.24 0.07 0.16 0.24
Panel E: financial variables nonconstituents sample PB regressions (n 1,851 )
Price-to-book ratio 4.73 5.15 2.14 3.30 5.47
Return on invested capital 0.15 0.38 0.06 0.12 0.19
Free cash flow growth (1.24) 5.80 (2.35) (1.41) (0.39)
Expected growth rate of dividend stream 0.15 0.29 0.07 0.16 0.24
Table I. Notes: This table reports summary statistics of regression variables; net cash flow variables are
Summary statistics reported in billions of dollars
index fund NCF relative to its mean. The heterogeneity is shown in the time plots of the The impact
two series in Figure 2. of passive
It is also interesting to note that the ROIC, FCF growth, and expected growth rate of
dividends of S&P 500 constituents are similar to those of nonconstituents. investing
A. Description of results from the analysis of price-to-earnings ratio
Table II presents the results from equations that relate PE ratio to the explanatory 1073
variables. The odd columns present results from regressions involving the sample of S&P
500 index constituents. The even columns present results from regressions involving the
sample of large-cap stocks that are not constituents of the S&P 500 index. Column (5)
shows that the PE ratios of constituents are positively and significantly (at the 1 percent
level) impacted by S&P 500 index fund NCF. The regression coefficient is economically
meaningful as well. The overall mean of PE was 21.03 and the time-series mean of S&P 500
index fund NCF was $0.97 billion. When evaluated at these average values, S&P 500 index
fund NCF corresponds to a 1.7 percent increase in the value of constituents:
0:37 0:97
1:7%:
21:03
The results from a regression of the same model on the sample of nonconstituents is
conveyed in column (6). The results from the sample of nonconstituents are not significant,
perhaps due to S&P 500 index fund NCF subsuming the effect of aggregate US equity fund
NCF. Therefore, regressions which control for aggregate US equity fund NCF are run. The
results from these regressions for constituents and nonconstituents are conveyed in

40,000,000,000
Aggregate US equity fund NCF
30,000,000,000
S&P 500 index fund NCF

20,000,000,000

10,000,000,000

0
Mar-93
Mar-94
Mar-95
Mar-96
Mar-97
Mar-98
Mar-99
Mar-00
Mar-01
Mar-02
Mar-03
Mar-04
Mar-05
Mar-06

(10,000,000,000)

(20,000,000,000)

(30,000,000,000)

(40,000,000,000)
Notes: This figure plots the time-series of aggregate NCF into US equity funds
and into the subset of such funds that Morningstar categorizes as having an
objective of tracking the S&P 500 Index; the sample period spans March Figure 2.
Time plot of money flows
1993-February 2007
MF
38,11

1074

Table II.

1993-February 2007)
Price-to-earnings ratio
regression results (March
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 47.97 * * * 31.94 * 47.19 * * * 31.78 * 48.17 * * * 31.95 * 47.46 * * * 31.81 *


(11.04) (16.35) (11.03) (16.35) (11.03) (16.35) (11.03) (16.35)
Return on invested 2.57 * * * 6.02 * * * 2.74 * * * 6.10 * * * 2.72 * * * 6.01 * * * 2.80 * * * 6.07 * * *
capital (0.46) (0.52) (0.46) (0.52) (0.46) (0.52) (0.46) (0.52)
Free cash flow growth (0.04) 0.01 (0.04) 0.01 (0.04) 0.01 (0.04) 0.01
(0.03) (0.03) (0.03) (0.29) (0.03) (0.03) (0.03) (0.03)
Expected growth rate 14.85 * * * 14.68 * * * 14.68 * * * 14.65 * * * 14.84 * * * 14.69 * * * 14.70 * * * 14.66 * * *
of dividend stream (0.78) (0.71) (0.78) (0.71) (0.78) (0.71) (0.78) (0.71)
Aggregate net cash 0.14 * * * 0.08 * * * 0.12 * * * 0.10 * * *
flow (0.02) (0.02) (0.03) (0.03)
Lagged aggregate net 0.09 * * * 0.05 * * 0.07 * * * 0.06 * *
cash flow (0.02) (0.02) (0.03) (0.03)
S&P 500 index fund 0.37 * * * (0.03) 0.22 * * (0.15) *
net cash flow (0.07) (0.07) (0.08) (0.09)
Lagged S&P 500 index 0.36 * * * (0.03) 0.25 * * * (0.12)
fund net cash flow (0.07) (0.07) (0.08) (0.08)
Observations 622 1,864 622 1,864 622 1,864 622 1,864
Adjusted R 2 0.14 0.30 0.14 0.30 0.14 0.30 0.14 0.30
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars; coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
columns (7) and (8), respectively. After controlling for aggregate US equity fund NCF, the The impact
negative impact of S&P 500 index fund NCF on the PE ratio of nonconstituents becomes of passive
statistically significant at the 10 percent level. The difference in the regression coefficients
between the constituent and nonconstituent samples are quantitatively unchanged investing
relative to the models that do not control for aggregate US equity fund NCF. It is also
interesting to note that fund flows do not explain much of the total variation in corporate
valuations, as the introduction of fund flow variables into the regression models 1075
(columns (5) through (8)) does not improve adjusted R 2 values or substantially alter the
value of the regression coefficients associated with the valuation control variables.
The results from these regressions fail to show that NCF into S&P 500 index funds
explains a substantial amount of the total variation of corporate valuations. However,
they do show that money flow into S&P 500 index funds leads to considerable changes
in the value of companies in the index relative to those outside of the index. This
central finding is consistent with the hypothesis. A related question is whether or not
price-pressure effects dissipate over time. To address this, we examine the coefficients
on lagged S&P 500 index fund NCF. If market participants correct the flow induced
price-pressure effects, then we would observe a reversal or at least a reduction in the
difference in the coefficients on lagged S&P 500 index fund NCF between the
constituent and nonconstituent samples relative to the difference in the coefficients on
contemporaneous S&P 500 index fund NCF. However, this is not apparent in the data.
The differences in the coefficients on the lagged and contemporaneous values in
the regressions that do not control for aggregate US equity fund NCF ((5) and (6))
are 0.39 and 0.40, respectively. The difference in the coefficients on the lagged and
contemporaneous values for S&P 500 index fund NCF in the regressions that do
control for aggregate NCF ((7) and (8)) are both 0.37.
The difference in the coefficients on lagged S&P 500 index fund NCF between the
constituent and nonconstituent samples are similar to the difference in the coefficients
on contemporaneous S&P 500 index fund NCF. Moreover, the standard errors of the
coefficients on the lagged values are not greater than those associated with the
contemporaneous values. The results from this analysis do not show that an adjustment
process exists.

B. Description of results from the analysis of price-to-book ratio


Table III presents the results from equations that relate PB ratio to the explanatory
variables. The results pertaining to this alternative valuation metric are similar to
those obtained through the use of price-to-earnings ratio. While S&P 500 index fund
NCF does not explain much of the total variation in corporate valuations, it does
positively impact on the valuation of constituents relative to nonconstituents. When
evaluated at the overall average value of PB ratio (4.25), the PB ratio of constituents
increased by a statistically significant (at the 1 percent level) 2.1 percent due to S&P
500 index fund NCF when evaluated at its time-series mean ($0.97 billion). The PB ratio
of nonconstituents also increased but only by 0.6 percent when evaluated at mean
levels. We attribute the slight increase in the value of nonconstituents to aggregate US
equity fund NCF, which were omitted from this set of regressions. After controlling for
aggregate US equity fund NCF, when evaluated at the overall average of PB ratio,
the PB ratio of constituents increased by a statistically significant (at the 1 percent
level) 1.6 percent while the value of nonconstituents was quantitatively unchanged.
MF
38,11

1076

Table III.
Price-to-book ratio

1993-February 2007)
regression results (March
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 9.82 * * * 6.01 * * * 9.66 * * * 5.83 * * 9.87 * * * 6.06 * * * 9.74 * * * 5.81 * *


(0.86) (2.29) (0.85) (2.29) (0.86) (2.29) (0.85) (2.29)
Return on invested 0.06 * 0.60 * * * 0.10 * * * 0.62 * * * 0.10 * * * 0.61 * * * 0.12 * * * 0.62 * * *
capital (0.03) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04)
Free cash flow growth (0.00) 0.01 * * * (0.00) 0.01 * * * (0.00) 0.01 * * * (0.00) 0.01 * * *
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Expected growth rate 5.95 * * * 2.11 * * * 5.91 * * * 2.10 * * * 5.95 * * * 2.11 * * * 5.92 * * * 2.10 * * *
of dividend stream (0.07) (0.06) (0.07) (0.06) (0.07) (0.06) (0.07) (0.06)
Aggregate net cash 0.03 * * * 0.02 * * * 0.02 * * * 0.02 * * *
flow (0.00) (0.00) (0.00) (0.00)
Lagged aggregate net 0.02 * * * 0.01 * * * 0.02 * * * 0.01 * * *
cash flow (0.00) (0.00) (0.00) (0.00)
S&P 500 index fund 0.09 * * * 0.03 * * * 0.07 * * * 0.01
net cash flow (0.01) (0.01) (0.01) (0.01)
Lagged S&P 500 index 0.09 * * * 0.01 0.06 * * * (0.02) * *
fund net cash flow (0.01) (0.01) (0.01) (0.01)
Observations 619 1,851 619 1,851 619 1,851 619 1,851
Adjusted R 2 0.57 0.62 0.57 0.62 0.57 0.62 0.57 0.62
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars; coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
To examine the possibility that flow induced price-pressure effects dissipate over time, The impact
we turn to the coefficient on the lagged value of S&P 500 index fund NCF. Interestingly, of passive
the difference in the coefficient on lagged S&P 500 index fund NCF between the
constituent and nonconstituents samples ((5) and (6)) is 0.08 yet the difference in investing
the coefficient on the contemporaneous values between the two samples is 0.06. The
same phenomena is observed among the regressions that control for aggregate US
equity fund NCF ((7) and (8)). While the slightly greater impact of lagged S&P 500 index 1077
fund NCF than contemporaneous S&P 500 index fund NCF on valuations is somewhat
perplexing, it clearly provides no support for the idea that the price-pressure effects
dissipate over time.
Lastly, we acknowledge the concern that our results may be partially an artifact
of the price effects associated with changes to the S&P 500 index. Evidence of
asymmetric price responses upon changes to index constituency status, detailed in
Chen et al. (2004), and competing explanations for the price effect validate this concern.
To examine how sensitive our results are to the event effects, we run an additional set
of regressions in which we exclude companies from our samples if they were either
added to or deleted from the S&P 500 index over the prior two months. The results
obtained from these regressions were quantitatively unchanged.

C. Subperiod analysis
Tables IV and VI provide subperiod results for March 1993 through February 2000 for
price-to-earnings ratio and price-to-book ratio, respectively. Tables V and VII provide
subperiod results for March 2000 through February 2007 for price-to-earnings ratio
and price-to-book ratio, respectively. Although the impact of NCF on valuation is
statistically significant and positive among constituent firms in both sub-periods, the
effect is more pronounced between 1993 and 2000. This may be due to heavier inflows
into index funds during this period (the fraction of total US equity mutual fund assets
tracking the S&P 500 increased by 151 percent between 1993 and 2000 versus
29 percent between 2000 and 2006). The more pronounced effect in the first subperiod
may also be due to the increased popularity of day trading over that period (Tables VI
and VII).
As in the full period analysis, the subperiod analysis also does not provide much
evidence to show that fund money flow explains much of the total variation in
corporate valuations but the analysis also does not show that flow induced valuation
effects dissipate over time.

V. Concluding remarks
A. Summary of research findings
This study is the first to explore the long-run relationship between S&P 500 index fund
money flow and corporate valuations. Through a series of panel regressions, we
examine whether money flow into S&P 500 index funds impacts the price-to-earnings
ratio and price-to-book ratio of firms within and outside the index. The results are
consistent with the hypothesis that money flow into S&P 500 index funds positively
impacts the price multiples of companies that are in the index relative to those that are
not. The impact on valuations persists the month after the flow occurs, suggesting that
the impact does not dissipate over time. These results are consistent with the theory
MF
38,11

1078

Table IV.

1993-February 2000)
Price-to-earnings ratio
regression results (March
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 15.15 (46.79) * 9.04 (51.42) * 12.84 (46.47) * 9.92 (51.41) *


(24.98) (26.68) (24.98) (26.69) (24.97) (26.69) (24.98) (26.69)
Return on invested 0.20 0.74 * * * 0.36 * 0.83 * * * 0.39 * 0.74 * * * 0.41 * 0.80 * * *
capital (0.21) (0.21) (0.22) (0.21) (0.22) (0.21) (0.22) (0.22)
Free cash flow Growth (0.02) 0.12 * * * (0.01) 0.12 * * * (0.02) 0.12 * * * (0.01) 0.12 * * *
(0.05) (0.04) (0.05) (0.04) (0.05) (0.04) (0.05) (0.04)
Expected growth rate 18.60 * * * 19.61 * * * 18.28 * * * 19.61 * * * 18.54 * * * 19.61 * * * 18.37 * * * 19.56 * * *
of dividend stream (1.54) (1.15) (1.54) (1.15) (1.54) (1.15) (1.54) (1.15)
Aggregate net cash 0.29 * * * 0.15 * * * 0.22 * * * 0.18 * * *
flow (0.05) (0.05) (0.06) (0.05)
Lagged aggregate net 0.04 0.11 * * (0.05) 0.18 * * *
cash flow (0.06) (0.05) (0.06) (0.06)
S&P 500 index fund 0.47 * 0.15 0.18 0.04
net cash flow (0.24) (0.22) (0.25) (0.23)
Lagged S&P 500 index 0.95 * * * (0.12) 0.81 * * * (0.74) * * *
fund net cash flow (0.25) (0.24) (0.29) (0.27)
Observations 493 1,335 493 1,335 493 1,335 493 1,335
Adjusted R 2 0.21 0.40 0.21 0.40 0.21 0.40 0.21 0.40
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars; coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 2.44 12.33 * * * 0.79 12.24 * * * 1.84 12.81 * * * 1.45 12.23 * * *


(1.62) (1.97) (1.60) (1.97) (1.58) (1.97) (1.58) (1.96)
Return on invested 0.00 0.04 * * * 0.05 * * * 0.04 * * * 0.07 * * * 0.03 0.07 * * * 0.03 * *
capital (0.01) (0.02) (0.01) (0.02) (0.01) (0.02) (0.01) (0.02)
Free cash flow growth (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Expected growth rate 3.30 * * * 2.28 * * * 3.20 * * * 2.29 * * * 3.26 * * * 2.26 * * * 3.24 * * * 2.25 * * *
of dividend stream (0.10) (0.09) (0.10) (0.09) (0.10) (0.09) (0.10) (0.09)
Aggregate net cash 0.05 * * * 0.01 * * 0.02 * * * 0.02 * * *
flow (0.00) (0.00) (0.00) (0.00)
Lagged aggregate net 0.04 * * * (0.01) 0.01 * * * 0.02 * * *
cash flow (0.00) (0.00) (0.00) (0.00)
S&P 500 index fund 0.25 * * * 0.01 0.23 * * * (0.01)
net cash flow (0.02) (0.02) (0.02) (0.02)
Lagged S&P 500 index 0.25 * * * (0.18) * * * 0.20 * * * (0.25) * * *
fund net cash flow (0.02) (0.02) (0.02) (0.02)
Observations 491 1,324 491 1,324 491 1,324 491 1,324
Adjusted R 2 0.70 0.73 0.71 0.73 0.71 0.73 0.71 0.73
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars; coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
investing
of passive
The impact

1993-February 2000)
regression results (March
Table V.
Price-to-book ratio
1079
MF
38,11

1080

Table VI.

2000-February 2007)
Price-to-earnings ratio
regression results (March
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 38.70 * * * 33.83 * * 38.31 * * * 33.72 * * 38.72 * * * 33.85 * * 38.36 * * * 33.73 * *


(10.77) (15.87) (10.76) (15.87) (10.76) (15.87) (10.76) (15.87)
Return on invested 39.95 * * * 43.43 * * * 39.68 * * * 43.36 * * * 40.18 * * * 43.34 * * * 39.82 * * * 43.17 * * *
capital (2.64) (2.44) (2.64) (2.44) (2.64) (2.44) (2.64) (2.44)
Free cash flow growth (0.06) (0.01) (0.06) (0.01) (0.06) (0.01) (0.06) (0.01)
(0.05) (0.04) (0.05) (0.04) (0.05) (0.04) (0.05) (0.04)
Expected growth rate 2.75 * * 1.25 2.74 * * 1.27 2.73 * * 1.25 2.73 * * 1.29
of dividend stream (1.19) (1.00) (1.19) (1.00) (1.19) (1.00) (1.19) (1.00)
Aggregate net cash 0.08 * * * 0.05 * * 0.06 * * 0.08 * * *
flow (0.02) (0.03) (0.03) (0.03)
Lagged aggregate net 0.07 * * * 0.03 0.07 * * * 0.03
cash flow (0.02) (0.03) (0.03) (0.03)
S&P 500 index fund 0.20 * * (0.09) 0.14 (0.18) * *
net cash flow (0.08) (0.08) (0.09) (0.09)
Lagged S&P 500 index 0.13 (0.05) 0.03 (0.07)
fund net cash flow (0.08) (0.08) (0.09) (0.09)
Observations 476 1,164 476 1,164 476 1,164 476 1,164
Adjusted R 2 0.18 0.33 0.18 0.33 0.18 0.33 0.18 0.33
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars; coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents Constituents Nonconstituents
(1) (2) (3) (4) (5) (6) (7) (8)

Intercept 9.59 * * * 6.01 * * * 9.50 * * * 5.89 * * * 9.59 * * * 6.00 * * * 9.51 * * * 5.83 * * *


(0.85) (2.02) (0.84) (2.01) (0.84) (2.02) (0.84) (2.01)
Return on invested 2.35 * * * 2.26 * * * 2.28 * * * 2.23 * * * 2.43 * * * 2.26 * * * 2.33 * * * 2.21 * * *
capital (0.23) (0.19) (0.23) (0.19) (0.23) (0.19) (0.23) (0.19)
Free cash flow growth (0.00) 0.02 * * * (0.00) 0.02 * * * (0.00) 0.02 * * * (0.00) 0.02 * * *
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Expected growth rate 4.41 * * * 1.00 * * * 4.42 * * * 1.01 * * * 4.40 * * * 1.00 * * * 4.41 * * * 1.01 * * *
of dividend stream (0.12) (0.09) (0.12) (0.09) (0.12) (0.09) (0.12) (0.09)
Aggregate net cash 0.02 * * * 0.01 * * * 0.01 * * * 0.01 * * *
flow (0.00) (0.00) (0.00) (0.00)
Lagged aggregate net 0.02 * * * 0.01 * * * 0.02 * * * 0.02 * * *
cash flow (0.00) (0.00) (0.00) (0.00)
S&P 500 index fund 0.04 * * * 0.01 * * 0.03 * * * 0.01
net cash flow (0.01) (0.01) (0.01) (0.01)
Lagged S&P 500 index 0.04 * * * 0.01 * * 0.01 * (0.03) * * *
fund net cash flow (0.01) (0.01) (0.01) (0.01)
Observations 477 1,173 477 1,173 477 1,173 477 1,173
Adjusted R 2 0.59 0.66 0.59 0.67 0.59 0.66 0.59 0.67
Notes: Significant at: *p , 0.10, * *p , 0.05 and * * *p , 0.01; odd-numbered columns report results from the sample of stocks in the S&P 500 index;
even-numbered columns report results from the sample of stocks outside of the S&P 500 index with greater than median market capitalizations; net cash
flow variables are reported in billions of dollars, coefficients are estimated using the fixed effects model; standard errors are given below the regression
coefficients
investing
of passive
The impact

2000-February 2007)
regression results (March
Price-to-book ratio
Table VII.
1081
MF of downward-sloping (i.e. not horizontal) demand curves for stocks and suggest that
38,11 money flow from index funds can distort stock prices.

B. Practical implications of research findings


Based on the results from our empirical study, it appears that the preference shift
towards index fund investing is reducing the informational efficiency of stock prices.
1082 Informed investors may recognize the oversupply of capital allocated to stocks in
indices and then place trades which counteract the effect (Grossman and Stiglitz,
1980)[6]. However, the speed of adjustment back to equilibrium valuations will be slow
in the presence of inattentive investors (Duffie, 2010). By their nature, index fund
investors are inattentive to asset valuations and, as described in De Long et al. (1990)
and Shleifer and Vishny (1997), arbitragers (and perhaps most importantly those who
provide them with capital) are rather impatient[7].
Prior theoretical work implies that, in the specific setting examined in this study, the
preference shift towards index fund investing is an endogenous determinant of the
speed of adjustment and as a result intertemporal arbitrage opportunities stemming
from the shift will be unattractive. Moreover, until the preference shift abates,
attempting to arbitrage the mispricing away may drown those informed traders
swimming against the current of passive investment. This is because, as discussed in
De Long et al. (1990), the impact of noise trader risk on asset prices is increasing in the
proportion of noise traders in the market. This implies that the distortive effect of index
fund investing on stock prices may be intensified by the perverse tendency of many
active fund managers to closet index (Cremers and Petajisto, 2009).
Mispricing among equities within the index may have adverse implications,
including a reduction of the allocative efficiency of the stock market and investors
performance evaluations of actively managed funds. The flow of non-valuation
motivated passive funds may also reduce oversight among firms within the S&P 500,
even if index inclusion increases the number of analysts providing information to
investors about firm governance (Yu, 2008). These and other economic consequences
are discussed in Wurgler (2010), Woolley and Bird (2003), Wermers and Yao (2010) and
Ye (2011). This study adds to the literature on the importance of passive investing on
valuations by linking flows into index funds with the valuations of companies that are
constituents of the index relative to those that are not.

Notes
1. See, for example, Shleifer (2000) for a discussion of this hypothesis.
2. Some advantages of PB ratio are that it provides more meaningful valuation estimates in the
event that earnings are negative and book values tend to be less volatile than earnings.
However, it is important to bear in mind that one downside of book value is that it does not
consider human capital as an operating factor. Additionally, it is difficult to compare
companies in different industries based on PB ratios due to differing levels of hard asset
intensity across industries.
3. Details on the methods employed by Morningstar to estimate net cash flows can
be found at: http://corporate.morningstar.com/us/documents/Direct/INS_MDT_
EstimatedNetFlowsMethodology.pdf
4. Non-operating establishments are detected through industry name. Following the method
of Petersen and Rajan (1997), financial services companies are considered to be those with
a Standard Industrial Classification code ranging from 6,000 to 7,000. ADRs, ADSs, AMs, The impact
and GDRs are detected through the company name (i.e. ADR in a company name indicates
that the security is an ADR). of passive
5. For a discussion of how these factors impact valuations, see Koller et al. (2005) and investing
Damodaran (2007).
6. Alexander et al. (2007) find that valuation-motivated trades by mutual fund managers
outperform non-valuation motivated trades, and that this outperformance was greater in the 1083
more recent period (1992-2003) than it had been between 1980-1991. The increased
magnitude of the effect over a period characterized by a preference shift towards passive
investing lends some credence to Grossman and Stiglitzs (1980) theory.
7. De Long et al. (1990) suggest that this is because investors evaluations of professional
money managers often occur over short time intervals. Gromb and Vayanos (2010) provide a
detailed review of the literature on the limits of arbitrage.

References
Alexander, G.J., Cici, G. and Gibson, S. (2007), Does motivation matter when assessing trade
performance? An analysis of mutual funds, Review of Financial Studies, Vol. 20 No. 1,
pp. 125-50.
Baberdis, N., Shleifer, A. and Wurgler, J. (2005), Comovement, Journal of Financial Economics,
Vol. 75 No. 2, pp. 283-317.
Brealey, R. (2000), Stock prices, stock indexes and index funds, Bank of England Quarterly
Bulletin, February, pp. 61-8.
Brealey, R., Allen, F. and Myers, S. (2006), Principles of Corporate Finance, 8th ed.,
McGraw-Hill/Irwin, New York, NY.
Chen, H.H., Noronha, G. and Singal, V. (2004), The price response to S&P 500 index additions
and deletions: evidence of asymmetry and a new explanation, Journal of Finance, Vol. 59
No. 4, pp. 1901-29.
Damodaran, A. (2007), Return on capital (ROC), return on invested capital (ROIC) and return on
equity (ROE): measurement and implications, New York University Stern School of
Business Working Paper, available at: http://ssrn.com/abstract1105499
De Long, J.B., Shleifer, A., Summers, L.H. and Waldmann, R. (1990), Noise trader risk in
financial markets, Journal of Political Economy, Vol. 98 No. 4, pp. 703-38.
Duffie, D. (2010), Presidential address: asset price dynamics with slow-moving capital,
The Journal of Finance, Vol. 65 No. 4, pp. 1237-67.
Fama, E. (1970), Efficient capital markets: a review of theory and empirical work, The Journal
of Finance, Vol. 25 No. 1, pp. 383-417.
French, K.R. (2008), Presidential address: the cost of active investing, Journal of Finance, Vol. 63
No. 4, pp. 1537-73.
Geweke, J., Meese, R. and Dent, W. (1982), Comparing alternative tests of causality in temporal
systems: analytic results and experimental evidence, Journal of Econometrics, Vol. 21
No. 2, pp. 161-94.
Goetzmann, W.N. and Massa, M. (2003), Index funds and stock market growth, Journal of
Business, Vol. 76 No. 1, pp. 1-28.
Gromb, D. and Vayanos, D. (2010), Limits of arbitrage, in Lo, A.W. and Merton, R.C. (Eds),
Annual Review of Financial Economics, Vol. 2, Annual Reviews, Palo Alto, CA, pp. 251-75.
MF Grossman, S. and Stiglitz, J. (1980), On the impossibility of informationally efficient markets,
The American Economic Review, Vol. 70 No. 3, pp. 393-408.
38,11 Investment Company Institute (2011), 2011 Investment Company Fact Book, 51st ed., The
Investment Company Institute, Washington, DC, pp. 1-236, available at: www.ici.org/pdf/
2011_factbook.pdf
Morck, R. and Yang, F. (2002), The mysterious growing value of S&P 500 membership,
1084 National Bureau of Economic Research Working Paper 8654.
Petajisto, A. (2011), The index premium and its hidden cost for index funds, Journal of
Empirical Finance, Vol. 18 No. 2, pp. 271-88.
Petersen, M. and Rajan, R. (1997), Trade credit: theories and evidence, Review of Financial
Studies, Vol. 10 No. 3, pp. 661-91.
Shleifer, A. (1986), Do demand curves for stocks slope down?, Journal of Finance, Vol. 41 No. 3,
pp. 579-90.
Shleifer, A. (2000), Inefficient Markets: An Introduction to Behavioral Finance, Oxford University
Press, New York, NY, p. 2.
Shleifer, A. and Vishny, R.W. (1997), The limits of arbitrage, Journal of Finance, Vol. 52 No. 1,
pp. 35-55.
Urban Institute and the Brookings Institution (2010), Historical Top Bracket and Rate Tax,
available at: www.taxpolicycenter.org/taxfacts/displayafact.cfm?Docid65&
Topic2id70 (accessed January 15, 2010).
Wermers, R. and Yao, T. (2010), Active vs passive investing and the efficiency of individual
stock prices, working paper.
Woolley, P. and Bird, R. (2003), Economic implications of passive investing, Journal of Asset
Management, Vol. 3 No. 4, pp. 303-12.
Wurgler, J. (2010), On the economic consequences of index-linked investing, in Allen, W.T.,
Khurana, R., Lorsch, J. and Rosenfeld, G. (Eds), Challenges to Business in the Twenty-first
Century: The Way Forward, American Academy of Arts and Sciences, Cambridge.
Ye, P. (2011), The value of active investing: can active institutional investors remove excess
comovement of stock returns?, Journal of Financial and Quantitative Analysis, May 1.
Yu, F. (2008), Analyst coverage and earnings management, Journal of Financial Economics,
Vol. 88, pp. 245-71.

Corresponding author
David Nanigian can be contacted at: david.nanigian@theamericancollege.edu

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com


Or visit our web site for further details: www.emeraldinsight.com/reprints

You might also like