1 Setting the record straight: Truths about indexing Vanguard Research January 2018 James J. Rowley, Jr., CFA; Joshua M. Hirt; Haifeng Wang, Ph.D. ■ Indexed investments have grown substantially during the past several decades, leading to dramatic changes in the asset management industry and the way people invest, as well as to significant cost savings. ■ In this paper, we review the rationale for indexing’s efficacy, quantify the benefits of indexing to investors, clarify the definition of indexing, and explore the validity of claims that indexing has an adverse impact on the capital markets. We show that index fund assets account for only 10% of the global total investable ■ market and 5% of trading volume on U.S. exchanges. ■ Additionally, we find no evidence that indexing contributes to fewer market opportunities for active managers, causes greater market volatility, or leads to a declining number of public companies.
2 Advocating for indexing Introduction The growth of indexing is perhaps one of the most The case significant changes to asset management in the past Indexing’s efficacy is driven by the zero-sum game 1 30 years. Indexed assets under management have theory and the effect of costs. The core concept grown from almost zero in the 1980s to about 30% of underlying index fund investing, zero-sum game registered fund assets globally in 2017. Index investing theory (Sharpe, 1991) states that a market consists has offered investors new opportunities for lower-cost, of the cumulative holdings of all participants in that more diversified portfolios. Yet despite indexing’s market such that the weighted-average return of these benefits, its growth has sparked debates about its participants is the market return. For each position that potential negative impact on the capital markets. outperforms the market, there must be at least one position that underperforms the market by an equal The chief topic of the debate is the impact of indexing amount, before investment costs. Figure 1 shows New Yorker on the capital markets. Recently, the asked, that historical performance has been consistent with “Is Passive Investment Actively Hurting the Economy?” the theory, presenting as a bell curve of both positive (Ledbetter, 2016), and Bloomberg ran a multipart and negative manager returns centered on a value of 2 series entitled, “The Worst-Case Scenarios of Passive slightly below zero. Investing” (Gandel, 2017). Further, Sakoui and Kaminska (2017) argue that the rise of indexing has made it more The zero-sum game theory posits a hypothetically difficult for active managers to outperform. Subramanian cost-free market. However, in the real world, investors et al. (2017) believe that it heightens market volatility. are subject to costs, including management fees, Bleiberg, Priest, and Pearl (2017) think it could hinder administrative costs, and transaction costs, all of which price discovery, and Ledbetter (2016) questions its effect reduce returns over time. The aggregate result of these on the concentration of the equity market. Although costs plays a significant role in realized investment many of these individual claims have been found to performance (Sharpe, 1966; Bryan and Rawson, 2014). be misconstrued or misinformed, they have not been Expense ratio is a key consideration for investors, since examined as a group against the historical evidence. its value is more predictable than those of other costs. Figure 2 shows that higher expense ratios are associated In this paper, we reassert the benefits of indexing, with lower excess returns. This relationship exists for 3 review how investors use index funds, and explore the index funds as well. claims that indexing adversely affects capital markets. First, we examine the underlying reasons for indexing’s How then can investors mitigate the risk of under- efficacy and quantify its benefits for investors. Next, performing a benchmark index? Evidence supports we discuss the ways in which indexing represents the the view that seeking low-cost funds is one of the transformation of active management. Finally, we analyze best ways to increase the chances of outperforming. several assertions regarding indexing’s potential negative (Wallick, Wimmer, and Balsamo, 2015). impact on the capital markets. Notes about risk and performance data: All investing is subject to risk, including the possible loss of the money you invest. Past performance is no guarantee of future returns. The performance of an index is not representative of any particular investment, as you cannot invest directly in an index. Vanguard ETF Shares are not redeemable with the issuing Fund other than in very large aggregations worth millions of dollars. Instead, investors must buy and sell Vanguard ETF Shares in the secondary market and hold those shares in a brokerage account. In doing so, the investor may incur brokerage commissions and may pay more than net asset value when buying and receive less than net asset value when selling. See Harbron et al. (2017) for a further discussion and empirical analysis. 1 Technically, zero-sum game theory is predicated on asset-weighted measurement. However, our research suggests that equal-weighted measurement yields a similar 2 outcome. 2 Rowley (2015) analyzes the relationship specifically for index exchange-traded funds. 3
3 Figure 1. Relative investing is a zero-sum game Performance has been consistent with theory 1,20 0 1,000 800 600 400 Number of funds 200 0 Greater 6% 5% 4% 3% 2% 1% 0% –1% –2% –3% – 4% – 5% –6% –7% Less than to to to to to to to to to to to to to to than 7% 7% 6% 5% 4% 3% 2% 1% 0% –1% –2% –3% – 4% – 5% – 6% –7% Excess returns Index funds Active funds Notes: The chart displays the distribution of equity funds’ excess returns relative to their prospectus benchmarks for the 15 years ended December 31, 2016. Our survivor bias calculation treats all dead funds as underperformers. It’s possible that some of those funds outperformed their benchmark index before they died. If we splice fund category average returns onto the records of dead funds, we see a modest decline in the percentage of funds that trailed their index. However, the differences from our existing calculations are not material. Source: Vanguard calculations, using data from Morningstar, Inc. Figure 2. Higher expense ratios are associated with lower excess returns Small-cap-blend funds’ excess returns versus expense ratios 10% 8 6 4 2 0 –2 excess return –4 Ten-year annualized – 6 –8 –10 0.020 0.030 0.000 0.005 0.010 0.015 0.025 Expense ratio Index funds Active funds Each dot represents the relationship of a fund’s expense ratio to its ten-year annualized excess return compared to its stated benchmark. The straight line represents Notes: the linear regression or best-fit trend line—the general relationship of expenses to excess returns. Some funds’ expense ratios and returns go beyond the scale and are not shown. All data are as of December 31, 2016. Harbron et al. (2017) shows this relationship for several equity and fixed income categories. Vanguard calculations, using data from Morningstar, Inc. Source: 3
4 The benefits Another benefit of indexing is that, because it is generally lower-cost than active investing, investors Industry studies have shown that one of the most keep more of their assets working for them. By important decisions an investor can make is asset substantially reducing the costs of gaining access to allocation (Ibbotson, 2010; Donaldson, et al. 2017). high-quality investments, indexing offers a tremendous Allocation among stocks, bonds, and cash has opportunity for wealth creation. shows that Figure 4 been shown to explain a majority of the variability the average asset-weighted expense ratio for index in diversified portfolio returns. Appropriate asset funds has consistently been less than that of active allocation allows investors to form return and risk funds. As of 2016, the average expense ratios for expectations and is the cornerstone of a disciplined index funds and active funds stood at 0.11% and portfolio construction process. 0.63%, respectively. However, achieving a chosen asset class (or sub- Figure 4 also shows the annual savings realized by asset class) return in practice is not certain. Exposure index fund investors. In 2016, that amount reached through a market-cap-weighted index fund reduces $23 billion. According to our estimates, without index this uncertainty because the fund’s objective is to track funds, investors would have paid more than $150 billion the performance of the benchmark index that reflects in additional investment costs cumulatively since the the asset class. shows the greater relative Figure 3 early 1990s. Further, we believe that index funds have predictability associated with index funds. The distribution introduced competitive price pressure to the industry, of index fund excess returns stays within a very tight at least partially explaining the downward trend in range because the fund’s objective is to produce a active-fund expense ratios. The downward trend is return similar to that of the index. The range of active noteworthy because it reflects a falling revenue source fund excess returns is more dispersed because active for asset managers, whose business incentives run funds attempt to outperform a benchmark index, and counter to the trend. Regardless of the reasons, the attempt to outperform comes with the risk of falling expense ratios benefit all investors. underperforming. Figure 3. Index funds provide relative performance predictability Range of excess returns 10% 8 6 4 2 0 –2 –4 Excess return range – 6 –8 –10 2009 2012 2016 2006 2006 Index funds Active funds Shaded regions depict the range of returns between the 75th and 25th percentiles for U.S. equity active and index strategies. Returns are defined as rolling 12-month Notes: excess returns relative to a prospectus benchmark on a monthly basis. The sample of U.S. equity funds is defined by Morningstar U.S. category. Vanguard calculations, using data from Morningstar, Inc. Source: 4
5 Figure 4. Indexing has helped to drive down the cost of investing 1.0 0% $25 0.80 20 0.60 15 0.40 10 0.20 5 Asset-weighted expense ratio Aggregate investor savings (billions) 0.00 0 2016 1993 Investor savings Active expense ratio (right) Index expense ratio (right) Notes: Data reflect the difference between the cumulative expense ratio fees paid by investors in open-end funds versus what they hypothetically would have paid if index funds did not exist. Investor savings are calculated as (asset-weighted expense ratio of actively managed funds x industry assets) - (industry asset-weighted expense ratio x industry assets). Source: Vanguard calculations, based on data from Morningstar, Inc. a capitalization-weighted U.S. total stock market index Defining indexing fund. This essentially reflects the performance of the Index strategies are constructed to match or track index fund investors’ aggregate portfolio. the components of a market index. By this definition, they represent just a small part of the global market. Contrary to expectations, this portfolio hasn’t tracked Registered fund assets made up only about 10% of the the broad market. In fact, excess returns versus those value of global investable securities and about 15% of of a total market fund have fluctuated greatly, outper- 4 U.S. investable securities as of September 30, 2017. forming or underperforming the market by nearly 14% About 30% of global and 35% of U.S. open-end fund and 8%, respectively, at various intervals. Since a truly assets are in index strategies. passive approach would resemble the horizontal line slightly below zero (represented by Vanguard Total Stock Although investors use index funds to implement Market Index Fund) in Figure 5, this suggests that index an allocation to a specific market segment, in their fund investors have been building active portfolios. In aggregate portfolios investors seem to be using index other words, the use of products categorized as index 5 funds for active purposes. Figure 5 , on page 6, shows funds does not automatically lead to overall passive the asset-weighted, rolling five-year relative returns portfolio allocations. Index funds and ETFs are part of of U.S.-domiciled equity index mutual funds and the evolution of active management. ETFs that are invested in U.S. equities relative to 4 These figures represent indexed assets in registered funds. Assets can also be invested in index-based strategies that are not in registered-fund format, such as separately managed accounts (SMAs). 5 For example, an investor could use an index fund to implement allocations to U.S. large-, mid-, and small-cap stocks, but the allocations in aggregate might not match 5 that of the total U.S. stock market.
6 Figure 5. Index fund investors haven’t necessarily tracked the market 15% 10 5 0 –5 Rolling 60-month relative return –10 Feb. 2016 Feb. 2010 Feb. 2004 Feb. 1998 Aggregate index fund portfolio Vanguard Total Stock Market Index Fund Notes: “Average index fund” includes U.S.-domiciled index mutual funds and ETFs in the U.S. equity and sector equity categories; returns are asset-weighted. Average index fund returns and Vanguard Total Stock Market Index Fund returns are relative to a total market index represented by the Wilshire 5000 Index. Data are from 1993 to 2016. Source: Vanguard calculations, based on data from Morningstar, Inc. proportion of U.S.-domiciled equity funds. Dispersion Defending indexing levels have remained fairly stable over time, even as Market impact passive market share has grown. It has been argued that the rise of indexing raises correlations among a market’s securities, which in Some argue that the rise in indexing causes market turn leads to active management underperformance. volatility. We do not find that such a relationship exists. However, a better way of defining market opportunity Figure 7 displays the rolling 12-month standard deviation is through security dispersion (Rowley, 2017)—in this of the Russell 3000 Index as well as the trend of U.S. case, the percentage of stocks that either out- or equity index fund assets as a proportion of U.S.-domiciled underperformed the index by at least ten percentage equity funds. While the percentage of assets in indexed points. Figure 6 shows the active management strategies has grown, market volatility has risen and opportunity set as defined by dispersion along with fallen in a seemingly random pattern, with notable spikes the trend of U.S. equity index fund assets as a 6 around the tech bubble and the great financial crisis. Kinniry (2017) shows that market volatility predates exchange-traded funds and is driven by global macro events. 6 6
7 Figure 6. No connection between indexing and price dispersion 100% 50% 80 40 60 30 40 20 Dispersion Asset percentage 20 10 0 0 2016 1993 Index fund asset percentage Russell 3000 Index annual return dispersion Dispersion is defined as the percentage of stocks in the Russell 3000 Index that have either outperformed or underperformed the index by at least ten percentage Notes: points. Index fund asset percentage is the percentage of assets in U.S.-domiciled equity funds invested in index funds. Sector funds are included. Vanguard calculations, using data provided by FactSet and Morningstar, Inc. Source: Figure 7. The growth of indexing and market volatility 45% 2016 2001 40 1,13 8 97 35 equity equity ETFs ETFs 30 25 20 15 10 5 0 2014 2005 1999 2009 2 011 2003 2016 1997 1995 1993 2001 2007 Index fund asset percentage Russell 3000 Index 12-month rolling standard deviation Index fund asset percentage is the percentage of assets in U.S.-domiciled equity funds invested in index funds. Sector funds are included. Notes: Vanguard calculations, using data from FactSet and Morningstar, Inc. Source: 7
8 of this volume represents primary-market trading (trading Figure 8. Indexing accounts for just a fraction in the underlying securities market) (Vanguard, 2015). of trading activity shows the percentage of daily ETF trading Figure 9 volume conducted solely on the secondary market. 95% The median ratio for equity ETF trading volume active was 94%, suggesting that for every $1 in trading mandates volume, only 6 cents involved primary-market trading. Similarly, the median ratio for bond ETFs was 83%, suggesting that for every $1 in trading volume, only 17 cents involved primary-market trading. In other words, 83% of the trading volume had no portfolio management impact and involved no trading in underlying securities. Public companies and industry concentration 5% Concerns have recently been raised that the declining index mandates number of public companies and increased U.S. market concentration can be attributed to the growth of indexed Source: Vanguard and Bloomberg. assets. Rowley and Wang (2017) find little evidence to support this assertion. Figure 10a shows that the predominant number of public companies have been micro-caps, and that micro-caps have experienced Another argument that has been raised is that the greatest drop-off. shows that despite Figure 10b increased activity from indexing leads to a decrease micro-caps’ drop-off, their proportion of overall market in price discovery. However, because index strategies capitalization has stayed relatively stable, at around have low turnover and trade at the margin across a 1.5%. These firms are not considered investable by large list of securities, their impact on trading activity is most mutual funds and are not included in many minimal. Figure 8 shows that the portfolio management indexes because of their illiquidity and regulatory activity of indexing makes up about 5% of daily trading constraints on the amount of ownership that may volume on U.S. exchanges. Other market participants, be acquired. The shrinking number of publicly listed including—but not limited to—retail investors, high- companies consists almost entirely of securities that frequency traders, and pension funds, account for the active and passive funds would not invest in anyway. vast majority of trading volume. “Active” participants play the dominant role in security trading, thereby facilitating price discovery. A commonly held belief is that the creation/redemption mechanism of exchange-traded funds (ETFs) leads to too much trading in the underlying securities markets. However, the overwhelming majority of ETF trading volume reflects secondary-market transactions (trades between two market participants). Only a limited amount 8
9 Figure 9. Most ETF trading activity takes place on the secondary market a. Equity ETFs b. Fixed income EFTs 100% 100% 80 80 60 60 40 40 20 20 Secondary market ratio Secondary market ratio 0 0 January January January January January January January January 2012 2015 2013 2014 2014 2015 2013 2012 Data cover the period from July 1, 2012, through June 30, 2015. The ten largest equity ETFs and bond ETFs by assets are used as proxies. Primary market activity is Notes: computed as daily creations or redemptions for each ETF, estimated as the daily change in shares outstanding multiplied by net asset value. Vanguard calculations, based on daily data from Bloomberg Inc. Source: Figure 10. For micro-caps, number is not the same as proportion a. Number of public companies grouped by size b. Market-cap proportion of companies grouped by size 100% 8,000 80 6,000 60 4,000 40 2,000 20 of public companies Market-cap proportion Number of public companies 0 0 2016 1979 2016 1979 Small-cap Mid-cap Large-cap Micro-cap Capitalization levels are defined by CRSP. The first and second deciles are defined as large-cap; the third, fourth, and fifth as mid-cap; the sixth, seventh, and eighth as Notes: low-cap; and the ninth and tenth as micro-cap. Only securities that had a portfolio assignment at year-end are used. Source: Vanguard calculations, based on data from CRSP. 9
10 Figure 11. Degree of concentration of public equities in the investible market has no noticeable trend 200 150 100 (1984 = 100) 50 number of publicly listed companies Level of Gini coefcient, HHI, and the 0 2012 2004 2008 2000 1996 1992 1988 1984 2016 Number of public companies Herfindahl-Hirschman Index Gini coefficient The Gini coefficient is a statistical measure of the degree of variation in a set of values and represents wealth distribution. HHI is the sum of the square of the market Notes: share of each firm in the Russell 3000 Index. All levels were assumed to be 100 in 1984. Vanguard calculations, based on data from FactSet. Source: It does not appear that the investable market has Conclusion become more concentrated as a result of the smaller Index investing’s growth has led to criticisms about number of public companies, either. Rowley and Wang the market impact of that growth. However, we believe (2017) adopted two concepts from social and industrial many of these concerns are unfounded. Our analyses economics: the Gini coefficient and the Herfindahl- show that indexing offers substantial benefits to Hirschman Index (HHI). Applied to equity market investors by providing reasonably predictable relative concentration, the Gini coefficient and HHI would become asset-class investment performance and by allowing larger if the market were more concentrated. F igur e 11 investors to keep more of their money in the process. plots their year-on-year changes along with the change in Although the assets managed under indexed mutual the number of public companies. Although the number funds and ETFs are growing, investors are increasingly of public companies has been declining, neither the Gini using these products to construct active portfolios. coefficient nor the HHI shows a trend toward a higher Finally, we do not find that the growth of indexing level of market inequality or concentration. has had an adverse impact on capital markets. 10
11 Morningstar Analyst Rating: Exchange- Morningstar, 2016. References Traded Funds and Index Mutual Funds . Morningstar Manager Bleiberg, Steven D., William W. Priest, and David N. Research. Available at: http://corporate.morningstar.com/US/ The Impact of Passive Investing on Market Pearl, 2017. documents/PR/Methodology_Morningstar_Analyst_Rating_ Efficiency . Epoch Investment Partners, Inc. Available for_ETFs_and_Index_Mutual_Funds.pdf. at http://www.nylinvestments.com/polos/MSEP38ll- 051768130.pdf. Racanelli, Vito J., 2015. Why 2017 Could Be the Year of the Barron’s Stockpicker. . Available at: https://www.barrons.com/ The Cost of Owning Bryan, Alex, and Michael Rawson, 2014. articles/why-2017-could-be-the-year-of-the-stockpicker- . Morningstar Manager Research. ETFs and Index Mutual Funds 1482555870. Available at http://global.morningstar.com/us/documents/pr/ Cost-Of-Owning-Index-ETF-MFS.pdf. Rowley Jr., James J., 2017. No More, No Less Challenging for Active . Valley Forge, Pa.: The Vanguard Group. Available Donaldson, Scott J., David J. Walker, Kimberly Stockton, at: https://vanguardadvisorsblog.com/2017/08/08/no-more- James Balsamo, and Yan Zilbering, 2017. Vanguard’s no-less-challenging-for-active/. Framework for Constructing Globally Diversified Portfolios . Valley Forge, Pa.: The Vanguard Group. Available at Rowley Jr., James J., and David T. Kwon, 2015. The Ins https://personal.vanguard.com/pdf/ISGCGP.pdf. and Outs of Index Tracking. Journal of Portfolio Management 41(3): 35–45. Edwards, Tim, and Craig Lazzara, 2013. Dispersion: Measuring Market Opportunity . S&P Dow Jones Indices. Available at: What’s Behind Rowley Jr., James J., and Haifeng Wang, 2017. https://www.asiaindex.co.in/documents/research/research- Valley Forge, Pa.: the Falling Number of Public Companies? dispersion-measuring-market-opportunity.pdf. The Vanguard Group. The Growth of Passive Investing: Has FTSE Russell, 2017. Sakoui, Anousha, and Izabella Kaminska, 2017. The Rise Available at There Been an Impact on the US Equity Market? and Rise of Correlation. Financial Times . Available at: http://www.pionline.com/article/20171005/WHITE_PAPERS/ https://www.ft.com/content/5a40b6a8-d176-11df-96d1- 171009882/the-growth-of-passive-investing-has-there-been- 00144feabdc0. an-impact-on-the-us-equity-market. Sharpe, William F., 1966. Mutual Fund Performance. The Worst-Case Scenario for Gandel, Stephen, 2017. The Journal of Business 39(1): 119-138. Passive Investing (Parts I and II) . Bloomberg. Available at: Sharpe, William F., 1991. The Arithmetic of Active https://www.bloomberg.com/gadfly/articles/2017-08-28/ Financial Analysts Journal Management. 47(1): 7–9. the-worst-case-scenario-for-passive-investing-part-i and at https://www.bloomberg.com/gadfly/articles/2017-08-29/ Subramanian, Savita, Dan Suzuki, Alex Makedon, Jill Carey the-worst-case-scenario-for-passive-investing-part-ii. The Hall, Marc Pouey, Jimmy Bonilla, and James Yeo, 2017. ETF-ization of the S&P 500, Part I . New York City, NY: Bank Harbron, Garrett L., Daren R. Roberts, and James J. of America Merrill Lynch. The Case for Low-Cost Index-Fund Rowley, Jr., 2017. . Valley Forge, Pa.: The Vanguard Group. Available Investing Vanguard, 2015. Exchange-Traded Funds: Clarity Amid the at: https://personal.vanguard.com/pdf/ISGIDX.pdf. Clutter . Valley Forge, Pa.: The Vanguard Group. Available at https://personal.vanguard.com/pdf/ISGETFC.pdf. Ibboston, Roger G., 2010. The Importance of Asset Allocation. (66): 2, 18–20. Financial Analysts Journal Wallick, Daniel W., Brian R. Wimmer, and James J. Balsamo, Shopping for Alpha: You Get What You Don’t Pay For . 2015. .” Available Alternative Facts “ ETF Kinniry, Fran, 2017. Valley Forge, Pa.: The Vanguard Group. Available at https:// at: https://vanguardadvisorsblog.com/2017/02/24/ personal.vanguard.com/pdf/ISGSFA.pdf. etf-alternative-facts/. Ledbetter, James, 2016. Is Passive Investment Actively Hurting the Economy? The New Yorker . Available at: https://www.newyorker.com/business/currency/is-passive- investment-actively-hurting-the-economy. 11
12 ® Vanguard Research vanguard.com > Connect with Vanguard P.O. Box 2600 Valley Forge, PA 19482-2600 For more information about Vanguard funds, visit vanguard.com or call 800-662-2739 to obtain a prospectus or, if available, a summary prospectus. Investment objectives, risks, charges, © 2018 The Vanguard Group, Inc. expenses, and other important information about a fund are contained in the prospectus; read All rights reserved. and consider it carefully before investing. Vanguard Marketing Corporation, Distributor. ® CFA is a registered trademark owned by CFA Institute. ISGBEL 012018