Trim Size = 8.125 X 10.875 Tue Jun 4 14:30:00 EDT 2013

-----IGNORE BLEED FOR GUTTER-----

www.journalofindexes.com

SERIOUS IDEAS FOR SERIOUS INVESTORS

index construction

July / August 2013

The Index Is Dead. Long Live The Index. Lubos Pastor, John Heaton and Aaron Foss

blackyellowmagentacyan

Understanding Your Benchmark Konrad Sippel Demystifying Low-Volatility Strategies Frank Siu Of Commodities And Correlations Geetesh Bhardwaj and Adam Dunsby Plus an interview with ASRS’ Underwood, Razor Hedge’s Hashmi and Feng on target-risk indexes, FTSE’s Woods on classifying Korea, S&P DJI’s Blitzer on the virtues of simplicity, and more! .125 from Trim Trim Line Bleed Line

INDEX

JUL-AUG

INDEX_I.pdf

-----IGNORE BLEED FOR GUTTER-----

Trim Size = 8.125 X 10.875 Tue Jun 4 14:31:17 EDT 2013

As one of the groundbreaking ETFs in the marketplace, PowerShares QQQ invests in 100 of the world’s most revolutionary companies. Each is listed on the Nasdaq Stock Market® and includes household names such as Oracle, Starbucks, Microsoft, Intel, Cisco Systems, and Google. With all of these revolutionary companies in one fund, innovative investing is within your reach.

PowerShares QQQ is based on the Nasdaq-100 Index®. The Fund will, under most circumstances, consist of all stocks in the Index. The Index includes 100 of the largest domestic and international nonfinancial companies listed on the Nasdaq Stock Market based on market capitalization.

blackyellowmagentacyan

There are risks involved with investing in Exchange-Traded Funds (ETFs) including possible loss of money. The funds are not actively managed and are subject to risks similar to stocks, including those related to short selling and margin maintenance. Ordinary brokerage commissions apply. Shares are not FDIC insured, may lose value and have no bank guarantee. Holding Weights as of 4/11/13: Microsoft 7.6%, Google 6.4%, Oracle 4.8%, Cisco Systems 3.4%, Intel 3.3%, Starbucks 1.3%. Holdings are subject to change.

powershares.com/innovation |

Shares are not individually redeemable and owners of the shares may acquire those shares from the Funds and tender those shares for redemption to the funds in Creation Unit aggregations only, typically consisting of 50,000 shares. PowerShares® is a registered trademark of Invesco PowerShares Capital Management LLC. ALPS Distributors, Inc. is the distributor for QQQ. Invesco PowerShares Capital Management LLC is not affliated with ALPS Distributors, Inc.

An investor should consider the Fund’s Investment objective, risks, charges and expenses carefully before investing. To obtain a prospectus, which contains this and other information about the QQQ, a unit investment trust, please contact your broker, call 800.983.0903 or visit www.invescopowershares.com. Please read the prospectus carefully before investing.

@PowerShares .125 from Trim Trim Line Bleed Line

INDEX

JUL-AUG

INDEX_II.pdf

Trim Size = 8.125 X 10.875 Tue Jun 4 14:07:45 EDT 2013

RIGHT HAND PAGE-------->

www.journalofndexes.com

Vol. 16 No. 4

features The Index Is Dead. Long Live The Index. By Lubos Pastor, John Heaton and Aaron Foss . . . . . . . . 16 How to construct a sound and useful benchmark

Understanding Your Benchmark By Konrad Sippel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Considering the impact of seemingly minor decisions

ASRS Takes An Unusual Step By Heather Bell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Asst CIO Underwood discusses ETF seeding

Reclassifying South Korea By Christopher Woods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Behind FTSE’s promotion of an emerging markets favorite

16

Keep It Simple … By David Blitzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Complexity is not necessarily better for investors

Demystifying Low-Volatility Strategies By Frank Siu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 An exploration of a popular—but oft-misunderstood—strategy

Of Commodities And Correlations By Geetesh Bhardwaj and Adam Dunsby . . . . . . . . . . . . . 40 Correlation trends appear to be shifting

Prudent Target-Risk Indexes By Quratulain Hashmi and Junchen Feng . . . . . . . . . . . . 46 Building an index for a globally diversified portfolio

Under Construction By Bruce Greig. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Test the foundations of your index knowledge!

data

Global Index Data Index Funds Morningstar U S Style Overview S&P Dow Jones Indices U S Industry Review Exchange-Traded Funds Corner

56 57 58 59 60

46

July / August 2013

www.journalofindexes.com

INDEX

10 10 10 11 11 11 12

JUL-AUG

INDEX_1.pdf

1 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

news

Bond Managers Earn High SPIVA Marks Stoxx Creates Libor Alternative US ETF Assets Break $1 5T S&P DJI Debuts Credit Spread Indexes MSCI Announces May Review Results Case-Shiller Indexes Show Continuing Rally iShares Debuts ASRS-Seeded Factor ETFs

32

Trim Size = 8.125 X 10.875 Tue Jun 4 14:07:57 EDT 2013



Okay, I’ll invent the wheel, you invent the index.

Think alongside a pioneer of index innovation to advance original investment ideas.

S&P Dow Jones Indices LLC is a subsidiary of McGraw Hill Financial. Standard & Poor’s and S&P are registered trademarks of Standard & Poor’s Financial Services LLC, a subsidiary of McGraw Hill Financial. Dow Jones is a registered trademark of Dow Jones Trademark Holdings LLC (“Dow Jones”). All Trademarks have been licensed to S&P Dow Jones Indices LLC. It is not possible to invest directly in an index. S&P Dow Jones Indices LLC, Dow Jones, S&P and their respective affiliates (collectively “S&P Dow Jones Indices”) do not sponsor, endorse, sell, or promote any investment fund or investment vehicle that seeks to provide an investment return based on the performance of an index. This document does not constitute an offer of services in jurisdictions where S&P Dow Jones Indices does not have the necessary licenses. S&P Dow Jones Indices receives compensation in connection with licensing its indices to third parties. Copyright © 2013 by S&P Dow Jones Indices LLC, a subsidiary of McGraw Hill Financial, and/or its affiliates. All rights reserved.

INDEX

JUL-AUG

INDEX_3.pdf

.125 from Trim Trim line Bleed line

blackyellowmagentacyan

We’re serious about you. Imagine what’s next at spdji.com/get-ideas

Trim Size = 8.125 X 10.875 Wed Jun 5 11:19:22 EDT 2013



VOO Vanguard S&P 500 ETF For an expense ratio 44% lower than the industry average*

®

Every client’s portfolio could use some Vanguarding. ®

The Vanguard S&P 500 ETF is one of nine Vanguard S&P ETFs available to your clients. With an expense ratio of only 0.06%, it can be a great building block for any portfolio. Give your clients a low-cost way to gain diversifed exposure in a wide range of sectors with the Vanguard S&P lineup. Take a closer look at advisors.vanguard.com/VOO 800 257-4333

All investments are subject to risk. Diversifcation does not ensure a proft or protect against a loss in a declining market. Vanguard ETFs are not redeemable with the issuing Fund other than in Creation Unit aggregations. Instead, investors must buy or sell Vanguard ETF Shares in the secondary market with the assistance of a stockbroker. In doing so, the investor will incur brokerage commissions and may pay more than net asset value when buying and receive less than net asset value when selling. For more information about Vanguard ETF Shares, visit advisors.vanguard.com/VOO, call 800 257-4333, or contact your broker to obtain a prospectus. Investment objectives, risks, charges, expenses, and other important information are contained in the prospectus; read and consider it carefully before investing.

*Source: Morningstar as of 5/1/2012. Based on 2012 ETF industry average expense ratio for S&P 500 Index Objective Fund ETFs of 0.09% and Vanguard S&P 500 ETF expense ratio of 0.05%. Morningstar data © 2012 Morningstar, Inc. All Rights Reserved. The information contained herein: (1) is proprietary to Morningstar and/or its content providers; (2) may not be copied or distributed; and (3) is not warranted to be accurate, complete or timely. Neither Morningstar nor its content providers are responsible for any damages or losses arising from any use of this information. Past performance is no guarantee of future results. © 2013 The Vanguard Group, Inc. All rights reserved. U.S. Pat. No. 6,879,964 B2; 7,337,138; 7,720,749; 7,925,573; 8,090,646. Vanguard Marketing Corporation, Distributor

Follow us @Vanguard_FA for important insights, news and education.

INDEX

JUL-AUG

INDEX_5.pdf

.125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Standard & Poor’s® S&P 500®, are registered trademarks of Standard & Poor’s Financial Services LLC (“S&P”) and have been licensed for use by The Vanguard Group, Inc. The Vanguard ETFs are not sponsored, endorsed, sold or promoted by S&P or its Affliates, and S&P and its Affliates make no representation, warranty, or condition regarding the advisability of buying, selling, or holding units/shares in the ETFs.

Trim Size = 8.125 X 10.875 Wed Jun 5 11:19:26 EDT 2013



The PowerShares DB ThePowerShares PowerSharesDB DB

Currency ETFs Currency ETFs ETFs Currency US Dollar Bullish Fund

UDN

US Dollar Bearish Fund

DBV

G10 Currency Harverst Fund

http://pwr.sh/CUp tp:// | http://pwr.sh/DBVp tp:// The funds are not mutual funds or any other type of Investment Company within the meaning of the Investment Company Act of 1940 and are not subject to its regulation. DB Commodity Services LLC, a wholly owned subsidiary of Deutsche Bank AG, is the managing owner of the funds. Certain marketing services may be provided to the funds by Invesco Distributors, Inc. or its affliate, Invesco PowerShares Capital Management LLC (together, “Invesco”). Invesco will be compensated by Deutsche Bank or its affliates. ALPS Distributors, Inc. is the distributor of the funds. Invesco, Deutsche Bank and ALPS Distributors, Inc. are not affliated. Commodity futures contracts generally are volatile and are not suitable for all investors. An investor may lose all or substantially all of an investment in the funds.

INDEX

JUL-AUG

INDEX_7.pdf

.125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

UUP

Trim Size = 8.125 X 10.875 Tue Jun 4 14:12:52 EDT 2013



upcoming

conferences

2013

THE WORLD’S BEST ETF CONFERENCES

SAVE THE DATE!

For Each Conference, Earn up to 8 CE Credits for CFP, CFA & CIMA Organizations*

Take advantage of our early-bird rates!

*Pending CFP, CFA & CIMA approval

6th annual

September 23, 2013

CONVENE CONFERENCE CENTER • NY

Complimentary attendance for INSTITUTIONAL INVESTORS Special registration fee for FINANCIAL ADVISORS & HEDGE FUNDS With the recent free fall of gold, Brent dropping below $100 and the general decline in commodity prices, what’s next? Is now a good buying opportunity for commodities, or time to reposition your portfolio? Join Keynote speaker Nouriel Roubini at this one-day event that will take a closer look at the future prospects for commodities including precious metals, energy and how the effects of currency debasing are impacting the commodity markets.

2nd annual

October 18, 2013 GRAND HYATT • SAN FRANCISCO Designed for FINANCIAL ADVISORS and INSTITUTIONAL INVESTORS Learn about nontraditional fixed-income products, best practices for a rising rate environment and what fixed-income opportunities lie outside the U.S. At Inside Fixed Income, uncover how to tactically tilt and strategically incorporate fixed-income investments into your portfolio and find out what risks are involved in trading fixed-income ETFs.

2nd annual

November 4, 2013 NEW YORK STOCK EXCHANGE • NY Designed for FINANCIAL ADVISORS and INSTITUTIONAL INVESTORS

TO REGISTER OR VIEW AGENDAS FOR ALL OUR CONFERENCES, VISIT:

Presented by:

www.IndexUniverse.com/conferences INDEX

JUL-AUG

INDEX_9.pdf

.125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Keynote speaker Tom Dorsey, along with other leading financial experts, will share with you tactical and actionable trading strategies to benefit any portfolio. See how the top traders achieve the best execution through a live trading session, and find out what macro-driven strategies will be steering the market in 2014 and beyond.



S&P DJI Debuts Credit Spread Indexes In April, S&P Dow Jones Indices launched new credit spread indexes based on U.S. corporate bond issues. The index firm has also given a license to trueEX, a U.S.-based electronic exchange for interest-rate swaps, to create futures contracts on the new indexes. The new indexes are the S&P/ ISDA US 150 Credit Spread Index, the S&P/ISDA U.S. Corporate 120 Credit Spread Index and the S&P/ISDA U.S. Financial 30 Credit Spread Index. The indexes’ constituents are all taken from the S&P 500 equity index and are the issuers of the largest and most liquid investment-grade long-term debt from the 500-stock parent index. The move signifies the intentions of S&P Dow Jones, which has traditionally been seen primarily as a provider of equity indexes, to step further into the fixed-income sector. The credit derivatives index sector of the fixed-income markets was caught up in the J.P. Morgan “London whale” trading scandal in the middle of last year. J.P. Morgan suffered an estimated $6 billion of losses from trades undertaken by traders in its Londonbased Chief Investment office, notably in Markit’s CDX North America Investment Grade Series 9 10-Year Index, CDX IG 9, which is based on credit default swaps from 125 investment-grade U.S. corporate issuers.

U.S.-domiciled Zoetis A, Germany’s Telefónica Deutschland and U.S.based Biomarin Pharmaceutical. The MSCI Emerging Markets Index’s largest additions include Thailand’s Shin Corp. PCL, Brazil’s Suzano Papel e Celulose and India’s Oil India. The MSCI Global Small Cap Indexes will see 372 additions and 334 deletions in the family’s broadest index, the MSCI ACWI Small Cap Index, while the MSCI Global Investable Market Indices’ headline index, the MSCI ACWI IMI, will see 348 additions and 311 deletions. MSCI’s global style indexes are also reviewed for style consistency during the reviews. Among the largest companies switching from a growth to a value designation are United Technologies Corp., Germany’s BASF and Altria Group, the press release said. Meanwhile, Berkshire Hathaway B, Walt Disney and Bank of America Corp. have been reclas-

sified as growth stocks and removed from the value category. Among other changes, the MSCI Frontier Markets Index saw five additions and 5 deletions, with Flour Mills of Nigeria and Union Bank of Nigeria and Argentina’s Irsa ADR representing the largest additions. MSCI’s main China A-Shares index saw 18 additions and 57 companies marked for deletion.

Case-Shiller Indexes Show Continuing Rally U.S. home prices rose again in March, tagging on double-digit yearon-year gains in what amounts to the market’s best price improvement since 2006, according to the latest S&P/Case-Shiller Home Price report. For the third-straight month, all 20 cities surveyed saw home values tick up year-on-year, with average home prices up 10.3 percent and 10.9 percent, respectively, for the 10-City and 20-City Composites in the past year

MSCI Inc. announced the results of its semiannual equity index review in mid-May. The changes were to be effective May 31, the press release said. The MSCI Global Standard Indices’ MSCI ACWI Index will lose 61 securities and gain 60, according to the press release, with the three largest additions to the MSCI World Index including

The MSCI Global Standard Indices’ MSCI ACWI Index will lose 61 securities and gain 60.

www.journalofindexes.com

INDEX

July / August 2013

JUL-AUG

INDEX_11.pdf

11 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

MSCI Announces May Review Results

Trim Size = 8.125 X 10.875 Tue Jun 4 14:14:46 EDT 2013



In May, Stoxx Ltd. rolled out an equally weighted version of its Stoxx China A 50 Index, which launched in February, a press release said. The Stoxx China A 50 Equal Weight Index implements the same methodology as its sister index, but instead of weighting components by free-float market capitalization, it resets its component weights back to 2 percent each on a quarterly basis. The very basic change in weighting approach results in some notable changes, with the top 10 components representing a total weight within the index of 20 percent at each rebalancing. The top 10 components in the cap-weighted Stoxx China A 50 Index represent roughly 46 percent of the index, the press release said.

Markit Rolls Out Non-Agency RMBS Indexes Markit in April launched a new range of fixed-income indexes, giving investors access to residential mortgage-backed securities (RMBS) from nongovernmental issuers such as trusts and other special purpose entities. The Markit iBoxx U.S. non-agency RMBS indexes comprise 27 subindexes, including 18 subsector indexes, that cover a total of approximately 350 senior bonds screened from a universe of 22,000 RMBSs issued between 2005 and 2007, a press release said. The family can also be broken down by vintage or by category. Those categories include prime, subprime, Alt-A and Option ARM, according to the press release. Components are selected based on deal size and pricing date, and on the type and quality of the mortgages referenced in each deal, the press release said. The index prices are determined via Markit’s bond pricing service, which sources quotes from a panel of dealers.

S&P DJI Launches Buyback Index In early May, S&P Dow Jones Indices launched the S&P 500 Buyback Index, which is designed to measure the performance of the top 100 stocks

with the highest buyback ratio in the S&P 500, offering an indexed way for investors to benefit from companies returning money to shareholders. The index’s constituents include the 100 companies within the S&P 500 that have bought back the highest ratio of their own stock within the past 12 months. The index is equal-weighted and rebalances on a quarterly basis. An S&P DJI representative said in a press release that stock buybacks can be an indicator of corporate health and noted that companies bought back nearly $400 billion in shares in 2012, but still hold record-high levels of cash.

Stoxx Rolls Out 3D Printer Index Stoxx Ltd. launched an equal-weighted global equities index in April that is focused on developed-market companies involved in 3D printing, a nascent technology that many analysts think may change the way societies and countries organize manufacturing capacity. Companies included in the Stoxx Global 3D Printing Tradable Index must derive at least 1 percent of their revenues from business operations related to 3D printing, the company said in a press release, which also noted that the index is designed to underlie investable products. Eligible components must also be listed in developed markets and meet minimum size and liquidity requirements. Although the index can include up to 30 components, only 10 companies had met all of the criteria for inclusion when it launched, the press release said.

FTSE, BOCHK To Launch Bond Series FTSE announced in late April that it is partnering with the Bank of China (Hong Kong) to develop the FTSE-BOCHK Offshore RMB Bond Index Series. According to the press release, BOCHK Asset Management Ltd. will serve as the advisor to the index series. BOCHK already operates an existing family of RMB indexes, and FTSE’s role will involve the management and calculation of those indexes and pre-

www.journalofindexes.com

INDEX

sumably any indexes launched in the future. FTSE will also assume responsibility for the licensing of any new indexes, the press release said. The press release also noted there has been increasing demand from investors for RMB fixed-income products and pointed out that FTSE has been expanding its lineup of fixedincome indexes in recent months.

S&P DJI Debuts RiskWeighted S&P GSCI S&P Dow Jones Indices launched a risk-weighted version of the S&P GSCI Index in late April that essentially attempts to ensure that each sector contributes the same amount of risk to the index. The S&P GSCI Risk Weight Index evaluates the risk contribution of each of five commodity sectors—agriculture, energy, industrial metals, livestock and precious metals—based on each sector’s covariance and uses that to weight the sectors. However, it also is designed to ensure that it offers up diversified exposure to the commodity spectrum by allocating no more than 33 percent to one commodity sector. The original S&P GSCI has been criticized for allocating too heavily to the energy sector—sometimes in the vicinity of 70 percent of its total weight. When a sector’s risk level exceeds the 33 percent cap, the overage is allocated among July / August 2013

JUL-AUG

INDEX_13.pdf

13 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Stoxx Debuts Equally Weighted China Index

Trim Size = 8.125 X 10.875 Tue Jun 4 14:14:46 EDT 2013



Van Eck To Self-Index Africa, Mideast ETFs Van Eck filed regulatory paperwork in May detailing index shifts for two of its ETFs to in-house indexes created by the company’s own indexing subsidiary. The planned changes, which should take effect by early summer, cover the Market Vectors Africa ETF (NYSE Arca: AFK) and the Market Vectors Gulf States ETF (NYSE Arca: MES), and are just the latest of a series of similar index shifts the company has carried out since establishing its own Frankfurtbased indexing unit last year. The Africa ETF “AFK” will shift to the Market Vectors GDP Africa Index from its current benchmark, the Dow Jones Africa Titans 50 Index; the Mideast ETF, “MES,” will switch to the Market Vectors GDP GCC Index from the Dow Jones GCC Titans 40 Index, according to the paperwork and the company’s website. The existing indexes are rulesbased, modified capitalizationweighted, float-adjusted benchmarks, but for each of the two new Market Vectors indexes, the weighting of each country is determined by the size of its gross domestic product relative to the GDPs of the other countries represented in the respective indexes.

Global X Debuts First Mongolia, Nigeria ETFs Global X Funds kicked off April with the rollout of two frontier-

market-focused ETFs targeting the markets of Nigeria, Mongolia and Central Asia. The Global X Nigeria Index ETF (NYSE Arca: NGE) focuses exclusively on the most populous country in Africa. The fund tracks a 25-stock Solactive index designed by Frankfurt-based Structured Solutions AG. It comes with an annual expense ratio of 0.68 percent. The Global X Central Asia & Mongolia Index ETF (NYSE Arca: AZIA) also tracks a Solactive index and invests in some 25 securities from Mongolia as well as Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan. AZIA comes with an annual expense ratio of 0.69 percent. Both Global X ETFs are what the company called “first to market.”

KNOW YOUR OPTIONS CBOE Sees April Options Fall Slightly CBOE Holdings saw its average daily trading volume in options fall just 1 percent year-over-year for the month of April. On its main exchange, the CBOE, average daily volume was down 6 percent, dragged down by a 31 percent decline in equity options. However, index options trading on the CBOE saw their average daily volume increase by 26 percent yearover-year, while the exchange-traded-product options’ average daily volume for April grew by 10 percent. The five most actively traded options contracts based on indexes or exchange-traded products included those tied to the S&P 500 Index, the CBOE Volatility Index (VIX), the SPDR S&P 500 Index Fund, the iShares Russell 2000 Index Fund and the iPath S&P 500 VIX Short-Term Futures ETN, the press release said.

CBOE Victorious In Index Lawsuit The Chicago Board Options Exchange said in May that the Supreme Court of the United States had declined to review the decision of the Illinois Appellate Court that not only blocked the International

www.journalofindexes.com

INDEX

Securities Exchange from basing options products on the S&P 500 Index and the Dow Jones industrial average but also prevented the Options Clearing Corp. from clearing or settling products of that nature created by the ISE. The press release noted that the Supreme Court’s decision ended a lawsuit that had been winding through the courts for six years.

BACK TO THE FUTURES CME Group Volumes Increase In April CME Group said in a press release that its average daily volume for the month of April increased by 8 percent from the prior year. Average daily volumes for contracts tied to indexes were up 16 percent year-over-year. In particular, the exchange’s most popular index futures contract, the S&P 500 E-Mini, saw its monthly volume increase by 14.7 percent year-over-year, while the E-mini Nasdaq 100 and Mini $5 Dow contracts saw their monthly volumes grow by 51.3 percent and 42.5 percent, respectively, from April 2012.

ON THE MOVE Fulton Exits PowerShares Ben Fulton, managing director of global ETFs at Invesco PowerShares, stepped down from his post at the end of April, ending an eight-year tenure during which the Wheaton, Ill.-based ETF company grew from an upstart into the fourth-biggest ETF issuer in the world. Andrew Schlossberg, head of U.S. retail distribution and global ETFs, will assume Fulton’s duties. Schlossberg has co-headed Invesco PowerShares’ management team since Bruce Bond resigned as president and chief executive officer in 2009. Fulton, who was previously president and CEO of the former Claymore Securities, joined PowerShares in late 2004 as a consultant and became a full-time executive in 2005. PowerShares was founded in 2002, and was acquired continued on page 45 July / August 2013

JUL-AUG

INDEX_15.pdf

15 .125 from Trim Trim line Bleed line

blackmagentacyan

costing 0.10 percent each versus BulletShares’ 0.24 percent price tag. The ETFs and tickers are as follows: • The iShares 2016 Investment Grade Corporate Bond ETF (NYSE Arca: IBCB) • The iShares 2018 Investment Grade Corporate Bond ETF (NYSE Arca: IBCC) • The iShares 2020 Investment Grade Corporate Bond ETF (NYSE Arca: IBCD) • The iShares 2023 Investment Grade Corporate Bond ETF (NYSE Arca: IBCE)

Trim Size = 8.125 X 10.875 Tue Jun 4 14:14:22 EDT 2013



Figure 1a

Counts Of U.S. Stocks The number of listed stocks changes signifcantly over time, falling by almost half over the past 12 years. Concurrently, the market risk represented by a portfolio with a fxed number of stocks changes.

The Approach

8,000

CRSP’s approach to index construction directly combines theory and empiricism:

7,000

INDEX

Seccurities

4,000 3,000

INDEX_17.pdf

Jan ‘12

Jan ‘10

Jan ‘08

Jan ‘06

Jan ‘04

Jan ‘02

Source: CRSP

Figure 1b

Market Capitalization Of Common Stock Counts A large-cap index of the top 500 stocks jumped from around 75% of cumulative market capitalization in the mid-1990s to almost 85% fve years later.

100% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% Jan 80

n2501-3000 n1001-2500 n751-1000 n501-750 n301-500 n201-300

n1-200

Jan 84

Jan 88

Jan 92

Jan Jan 96 00

Jan 04

Jan 08

Jan 12

Source: CRSP

July / August 2013

JUL-AUG

Jan ‘00

Jan ‘98

Jan ‘96

Jan ‘94

Jan ‘92

Jan ‘90

Jan ‘88

Jan ‘80

2,000

17 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

www.journalofindexes.com

5,000

% Of Cumulative Capitalization

Theoretically And Logically Guided The CRSP Indexes aim to be “current.” Company performance and valuation fluctuate with economic conditions, firm decisions and investor expectations. The result: Companies that looked cheap/expensive/small/big/U.S.-domiciled/liquid at one point likely will not remain so indefinitely. We could simply state that an index should be as close to current as possible, but that would have clear drawbacks in terms of turnover. A common industry compromise between being current and limiting turnover is to reconstitute indexes semiannually or annually. The CRSP Indexes’ quarterly reconstitution places a relative premium on being current; we married reconstitution to a novel migration strategy that limits turnover. The resulting indexes reflect changes to the investment opportunity set quickly while keeping turnover low. Free float is another example of a constraint derived theoretically/logically rather than empirically. It has become widely adopted by index providers because it makes sense. Shares that are not available for trading cannot possibly lie in the investment opportunity set. While a departure from pure cap-weighted indexes, freefloat-adjusted indexes are a more appropriate representation of those assets that investment managers should consider in their decisions. A robust way to set breakpoints is another theoretical

6,000

Jan ‘86

The Center for Research in Security Prices (CRSP) at the University of Chicago Booth School of Business sought, with its new index products, to combine current academic thought and practice with Sauter’s cardinal rule, while paying considerable attention to the material constraints faced by investors. The result is a family of indexes that are both theoretically justifiable and a practical representation of those securities in which a manager, subject to a related mandate, could invest. Other authors have done a good job of describing important features of benchmarks: completeness, objectivity, investability, etc. The CRSP Indexes have all of these features; however, this article delves deeper. CRSP seeks to explain our index design process and present the indexes’ mechanics in the context in which the solutions arose. We believe our process can be most easily digested by understanding our theoretically guided and empirically validated approach and the balances struck that make CRSP’s indexes valuable.

Jan ‘84

Building the Ideal Index

problem. Breakpoints reflect the ability to discriminate between the opportunity set contained in one index and the next. Ideally, breakpoints would arise naturally along some dimension of investor concern. However, the dimension of investor concern itself may not be obvious. We begin with a simple example: market cap. Academics and practitioners have long noted that equities with different market capitalizations display significant differences in average returns. In addition, small stocks and large stocks, as groups, have tended to move together. Domestic index providers have historically made the decision on cap-index membership in an ad hoc fashion by using counts of securities as proxies for market capitalization. For example, the S&P 500 Large-Cap Index contains 500 stocks, while the Russell 1000 Large Cap Index contains 1,000 stocks when reconstituted. However, most of us scratch our heads when asked why an investor should care whether a security is ranked 999 or 1,001 by size; it would seem hard to argue that such arbitrary cutoffs reflect genuine investor concerns. Furthermore, count-based indexes remain anchored at a point in time and always reflect the relationship that count had with

Jan ‘82

G

us Sauter, Vanguard’s recently retired CIO, argues that theory provides little guidance as to how investing, beyond holding the total market, is actually done (Sauter, 2002). His point, that managers define the opportunity set, is best articulated in his cardinal rule of indexing: An index must reflect the way that money managers actually invest.

Trim Size = 8.125 X 10.875 Wed Jun 5 11:20:53 EDT 2013

RIGHT HAND PAGE-------->

INDEX

The Latest Styles CRSP's multifactor model is designed to be easily intelligible to investors. The model acknowledges that all factors are not equally important. GROWTH

VALUE

EP = (2/3) FEP + (1/3) HEP FG = (1/3) FLGE + (1/3) FSGE + (1/6) INV + (1/6) ROA V1 = (2/3) EP + (1/3) BP

HG = (2/3) HGS + (1/3) HGE

V2 = (2/3) SP + (1/3) DP

G = (2/3) FG + (1/3) HG

V = (2/3) V1 + (1/3) V2 Source: CRSP

Figure 4b

Decomposition Of CRSP Style Model While the actual model is hierarchical, the fow-through weights expressed in our factor model are visible in the chart. Value

FEP HEP SP BP DP FLGE FSGE 3Y HGS 3Y HGE ROA INV

Growth

0%

20%

40%

60%

INDEX_19.pdf

100%

Source: CRSP

Figure 5

Thinking Inside The Box The CRSP Style Box is an intuitive representation of our style assignment routine. Our two-dimensional style model allows us to plot any frm in a value/growth plane. The 0.5 AR breakpoint separates value from growth, but it is not a hard line, as the investment styles themselves dictate. After initial placement, a security’s style characteristics may change. Given the inherent ambiguity in style, migration requires a signifcant change in characteristics; one that takes a frm from its current location through the band region (white) to the opposite band threshold. If the security remains beyond the threshold at rank, a 50% packet is moved. This model can be easily extended to form indexes containing characteristics that are value-only, growth-only, a combination of value and growth, or neither. Low Growth 1 AR = 0.5

RV

Value Stock AR = 0.8

Growth Stock AR = 0.2 0.3 Band Region

RG High Growth 0 Low Value

1

High Value

Value Percentile Rank Source: CRSP

July / August 2013

JUL-AUG

80%

% Of Composite Score

19 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

www.journalofindexes.com

Figure 4a

Growth Percentile Rank

us would consider American are incorporated overseas and recognize revenue abroad for tax purposes. Others, clearly foreign, go public through reverse mergers with U.S.-listed shell companies. We turned to data in an attempt to develop a transparent methodology. The domicile scheme we rely upon is a product of testing thousands of models that examine hundreds of company-level variables. Other areas are less clear still—cap breakpoints, for example. As mentioned before, large and small stocks behave differently, but even after significant research into potential breakpoints, no clear statistical discontinuities emerge. CRSP’s conclusion: Cap segments are a matter of convention. In an effort to stay practical, we adopted cumulative cap breakpoints at levels that should look very familiar to practitioners. A comparison with other major indexes is provided in Figure 3. Importantly, though, CRSP also included “bands” around these breakpoints and a migration plan called “packeting.” We studied banding and migration in depth and let the data provide guidance as to scale and mechanics. We require that a security pass a threshold beyond the breakpoint before a 50 percent “packet” of the security’s weight is moved to the adjacent index. It follows that a small-cap stock must move beyond the mid/small breakpoint and the mid/small band threshold before triggering the first 50 percent move to the midcap index; the migration of the remaining 50 percent depends on remaining beyond the threshold next quarter. Among all index providers that rely on banding, turnover reduction is touted as a benefit. We agree, but we believe there is also a deeper purpose. As the indexes age, banding and packeting capture something important: There is no unanimity of opinion among managers as to which marginal companies qualify as large or small. Beyond the reduction in turnover, our migration strategy improves the fit of the index to manager behavior. If there is a lack of unanimity on cap breakpoints, value and growth styles are downright fuzzy. CRSP’s work suggests that, in contrast to the model in Figure 2, value and growth managers do something other than hold portfolios of single-dimension “value” or “growth” stocks. Value managers typically describe their process as one that involves trying to buy assets or cash flows at inexpensive prices. Growth managers, on the other hand, look for fast-growing firms. These managers follow separate, though related, processes. Accordingly, we treat value and growth separately. This two-dimensional method allows us to generate a richer description of the style-specific investment opportunity set. Recent empirical research provides useful insights into the factors managers consider. CRSP was the first to introduce investment rate (“INV”) and return on assets (“ROA”) as growth factors. Academic studies show that firms that invest more tend to grow faster, as do firms that are more profitable. Additionally, economic theory links both INV and ROA to expected stock returns. CRSP studied its factors in typical empirical fashion: portfolio sorts along factor dimensions, cross-sectional Fama-MacBeth regressions, cross-sectional and predic-

Trim Size = 8.125 X 10.875 Wed Jun 5 11:20:30 EDT 2013



predictable process, and several leading indexes see front-running by Annualized Turnover Vs. Annualized Monthly Tracking Error active managers looking 45% to trade securities tranIdeal sitioning into or out of the index ahead of pas40% sive managers who must make these trades. We randomize the pricing 35% date during reconstituPacketing tion, which alleviates front-running by making 2 30% the inclusion/exclusion of marginal securities in 4 5 a specific index harder 6 8 to predict. The random25% ization itself follows a 10 15 transparent algorithm— it should preclude 20 20% manipulation of index membership without 33 introducing any meth15% odological opacity. Tracking error, too, comes at a cost, albeit 10% in a slightly less salient 2.85% 2.90% 2.95% 3.00% 3.05% 2.70% 2.75% 2.80% “risk” dimension. This Annualized Monthly Tracking Error can be thought of as the potential for variance in Source: CRSP, periods 6/30/01-3/31/11 returns versus the instanNote: Tracking error, as presented above, is the simple average of each CRSP style index’s tracking error with respect to the appropriate Lipper active manager benchmark for a given banding/packeting conformation. taneous opportunity set. Active, benchmarked contradictions and tensions. One tries to provide a good investors call this “active” risk, as it represents a decision measure of the opportunity set for an asset class or style at a to deviate from a naive position in the asset. Measuring given point in time, but must balance the abstract desire for the quality of investing decisions then becomes a quesa perfect measure against practical considerations of fund tion of the return-to-risk ratio. operations. All major index providers tackle operational In some sense, index providers make similar decisions issues with features such as free-float adjustments and to those made by active managers. The index obviously banding/migration schemes. CRSP is no different in this deviates from the instantaneous “true” opportunity set; respect, though the attention we pay to costs borne by the the goal of the index provider is to strike a balance investor separates the CRSP Indexes from others. between the cost of turnover and risk from tracking The two most important costs of indexing are turn- error. Developing transparent, mechanical rules for over and tracking error. Turnover is obvious; irrespective banding and migration precludes a simple functional of cause, turnover has a direct dollar cost and eats into approach, but the idea pays homage to standard meanreturns. The bulk of turnover occurs at reconstitution, variance optimization. To determine the appropriate when constituents move in or out of an index. To track the mechanics, CRSP ran a large number of experiments that index, the indexer must access capital markets. Capital- studied problems ranging from the symmetry of bands at market access comes with price impact, especially when breakpoints to the ideal band width to the use of threshthe volume of a transaction is relatively large. olds versus continuous transitions to transitional packet As one could imagine, turnover and price impact are size. We evaluated more than 40 different approaches different. An added dimension of our analysis looked at, for before settling on our breakpoints, threshold bands lack of a better name, “bad turnover.” Bad turnover is that and 50 percent packeting. We compared all approaches which demands transacting in volumes that are difficult for on several metrics: tracking errors versus appropriate the market to meet and are thus expected to have greater Lipper and Morningstar manager indexes, tracking error price impact cost. versus “pure” (i.e., no migration strategy) indexes, aggreThe CRSP Indexes also address turnover costs asso- gate turnover and bad turnover (Figures 6a and 6b). ciated with front-running. Reconstitution is a fairly continued on page 55 Figure 6b

www.journalofindexes.com

INDEX

July / August 2013

JUL-AUG

INDEX_21.pdf

21 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Annualized Turnover

What Trade-Off?: Lipper

Trim Size = 8.125 X 10.875 Tue Jun 4 14:17:02 EDT 2013



The Treatment Of Corporate Actions When looking at an index, investors often focus on its selection methodology (i.e., which components are in the index) and its weighting scheme (i.e., which component or components influence the index most, and to what extent). From the perspective of these two dimensions only, the majority of indexes, and certainly most major market benchmarks, are actually quite simple. Selecting the largest and most liquid stocks from a country, region or market segment and weighting them by their size does not hold any major surprises. However, even for “plain vanilla” indexes, the calculation of the index often involves additional complexities that may be overlooked by those evaluating an index’s methodology. This is especially true in the area of the treatment of corporate actions. In other words, the rules governing how capital raisings and other corporate events are reflected in an index can have a significant impact on its behavior and performance. To illustrate this, below we examine different types of corporate events and their treatment in indexes. Regular Dividends Regular dividends are a relatively simple event. The holder of the index portfolio receives a dividend payment. In price indexes, this leads to a drop in the index value. In total-return indexes, this price drop is adjusted for in the index calculation. So far, so simple; however, there are three different ways in which this reinvestment can be calculated: Option 1: Reinvestment in the Index Portfolio – In this case, the received cash is redistributed across all components of the index. This means that an investor tracking the index will need to purchase additional shares in all components of the index according to their current weighting, up to a total amount equal to the dividend payment. This method has the advantage that the portfolio represents the true market capitalization weightings of the underlying market correctly at all times. On the downside, this adjustment generates relatively large amounts of trading activity and therefore costs, as a trade needs to be made in all the components of the index. Most international index benchmarks, including indexes from Stoxx, MSCI and FTSE, follow this methodology. Option 2: Reinvestment in the Single Stock – In this case, the cash received is reinvested only into the stock paywww.journalofindexes.com

INDEX

ing the dividend. An investor tracking the index would therefore purchase shares of that stock for the amount of the payment received. This artificially increases the market capitalization of the affected stock to a synthetic level, equivalent to that observed prior to the payment. Reinvestment into the entire portfolio still happens, but not until the next regular rebalancing of the index. This method reduces portfolio adjustment costs, as transactions only take place in one share in the portfolio. This methodology is used by the indexes of Deutsche Börse AG, including the DAX index. Option 3: No Reinvestment – The third option is of a more theoretical nature only and assumes that the amount paid out is kept as cash until the next rebalancing of the index, where it will be reinvested in the entire portfolio. While this method theoretically minimizes transaction costs, it also skews the returns of the index, as during dividend seasons, a significant part of the portfolio may not actually be invested in the equity market. This methodology is not commonly used in any major indexes. Spinoffs From an index calculator’s perspective, one of the most complex corporate actions is the event of a spinoff within the index. In this case, a member company of the index spins off parts of its business into a separate corporation, which is itself typically stock market listed. To ensure a smooth adjustment, the effects of such corporate actions need to be reflected in the index. Such adjustments need to be performed on an ex ante basis; in other words, the adjustment needs to take place at the market close on the day prior to the ex-date of the event in order to ensure that no jumps in the index occur. In theory, it is fairly easy to determine the theoretical post-spinoff price for the share in the index using a simple formula, as shown below: Formula 1

Q PA = PA - B * PB ’ QA Where PA = Price for old company in index PA’ = Adjusted price for mother company in index after spinoff PB = Theoretical price for daughter company prior to trading on ex-date QA = Number of shares in mother company QB = Number of shares in daughter company As for all index adjustments, this calculation assumes a ceteris paribus environment post spinoff; in other words, the corporate action has no effect on valuations. So index compilers assume that the valuation of the “old” company is equal to the sum of the valuations of the two companies after the spinoff takes place. July / August 2013

JUL-AUG

INDEX_23.pdf

23 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

I

ndexes are often praised as being fully transparent and rules-based, with published rule books spanning hundreds of pages. Within those rule books, many smaller details can also have a significant influence on the performance of a benchmark and its characteristics. In this article, we highlight three areas that can easily be overlooked, but which may be of interest for investors seeking a specific exposure or wanting to invest in a derivative product based on such an index. We look at the treatment of corporate actions, interest rates and country exposures to illustrate why it is necessary to consult and study index rule books and compositions with great care in order to fully understand their effects on the investment.

Trim Size = 8.125 X 10.875 Tue Jun 4 14:17:43 EDT 2013

24

How Refinancing Rates Affect Leveraged Indexes Leveraged indexes apply a leverage factor to the daily return of a chosen base index. Leveraged indexes are available for all major benchmarks and from all major global index providers, with the leverage factor typically ranging from plus eight to minus eight. For simplicity, we will consider a two-times leveraged index for this part of the article. A leveraged index with a factor of two aims to provide the investor with twice the daily return of the base index. However, to ensure that investors can actually replicate the index, some additional factors need to be considered. To calculate the leveraged index, index calculators assume that for every investment made into the base index that a loan of the same amount is taken out and also invested into the base index for one day. At the end of the day, the loan is theoretically repaid, and the long position in the index is liquidated; hence, generating twice the return of the underlying index for that day. The index calculation must therefore reflect the costs of taking out that loan. The index formula for the leverage index with a factor of two is: Formula 2

Leveraged_Indext = Leveraged_Indext -1 * Base_Indext - daily_interest_rate) (2 * Base_Indext -1 When leveraged indexes were first introduced, the interest rate used in Formula 2 was the overnight financing rate. In the case of euro-denominated indexes, the daily EONIA rates were used. In practice, however, products tracking these indexes would not reset their positions on a daily basis and Figure 1

Euribor, EONIA And The Liquidity Spread 0.025

6

0.02

5

0.015 4 0.01 3 0.005 2 0

■ Euribor

■ Eonia 1-Year Fixing

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

-0.01

2003

0

2002

-0.005

2001

1

2000

The potential problem when this calculation is performed prior to the ex-date is that the price for the daughter company PB cannot be evaluated objectively or quantitatively. The valuations forecast for spinoffs by market analysts and participants often differ widely, even for very large and well-covered constituents of major market indexes. In addition, such valuation estimates may imply a higher total valuation of the two parts following the spinoff. But any such potential increase (or decrease) in value cannot be reflected in the index, as it would violate the ceteris paribus adjustment principle and not accurately capture the return made by the investor in the index, leading to tracking errors. Many international index providers do, however, calculate the adjustment based on estimated or synthetically constructed hypothetical prices for the daughter company (in the case where the daughter company does not become an index component itself, as is the case in most blue chip indexes with a fixed number of components). Technically, this requires a tracking portfolio to sell the daughter shares at this hypothetical price at the close prior to the event in order to maintain zero tracking error. In practice, of course, this transaction can only occur once trading in the new instrument actually starts: typically, the morning of the next trading day. Any difference between the price actually realized in that transaction and the hypothetical price used in the index calculation does generate some tracking error against the index. An alternative to this adjustment method was first introduced by Deutsche Börse AG for the DAX family of indexes in 2003, and has since been adopted by other index providers. The new method includes the daughter company as an index component for one extra trading day, irrespective of whether the company becomes eligible for the index, meaning the index’s component count also increases by one for this extra day. The inclusion price is then mathematically irrelevant, since the formula for the adjusted price ensures that the sum of the market capitalization of the two parts equals the market capitalization of the pre-spinoff company. As the trading starts in both components, both will receive current market prices and the index will update accordingly. This method ensures all price movements due to a potential new valuation are actually recognized as part of the index’s performance, in the same way an investor would experience them in his or her portfolio. At the end of the first post-spinoff trading day, the new component is then excluded at its closing price, and the reinvestment of the resulting cash distribution is handled according to the regular process specified in the index rules. This method effectively ensures that even a complex corporate action such as a spinoff can be implemented by all index investors without experiencing any tracking error against the index calculation. As a side effect, this leads to the special case of an index with a fixed number of components carrying more than that number for a short time. This was observable in the DAX index on Jan. 31, 2005, when the rule was first applied to the Bayer’s spinoff of Lanxess, leading to 31 components for the first, and so far only, time in 25 years of existence of the DAX index.

1999

blackyellowmagentacyan



Figure 2

reference rate can have a significant impact on an index’s performance. Using a reference rate that is close to the actual replication costs facing those tracking the index ensures that products with low tracking errors can be offered to investors.

The Efect Of Diferent Leveraged Index Methodologies 2500

Country Exposures In Standard Indexes Index Level

2000

1500

1000

■ Old Method

12/12

12/11

12/10

12/09

12/08

12/07

12/06

12/05

12/04

12/03

12/02

12/01

12/00

12/99

0

12/98

500

■ New Method

Source: Stoxx

instead decided to provide financing at a longer-term rate. Since the outbreak of the financial crisis, the financing spread between a regular one-year Euribor rate and the annualized EONIA swap rate has significantly increased, as shown in Figure 1, making it increasingly difficult to replicate the performance of leveraged indexes accurately. As can be seen from the chart, until the middle of 2007, this financing spread was negligible in size; hence, it did not impact the tracking of leveraged products. But since then, in response to the appearance of the spread, many index providers have decided to adjust the refinancing methodology to incorporate the liquidity spread in the index calculation, helping index-tracking funds to track more accurately. The adjusted formula now reads: Formula 3

Another topic of interest in benchmark construction, and one that is becoming increasingly important for investors, is how companies are allocated to a home market. In the past, this allocation has been relatively easy, as for most companies, the country of listing is strongly correlated with the location of the company’s business activities. However, with increasing globalization in corporate activity and in financial markets, the process of allocating companies to countries is becoming much less straightforward. Increasingly, stock exchanges compete for the primary listings of companies from less-developed markets and are more and more successful in doing so. Hence, we see many more listings of corporations outside of their home markets. Index providers typically use three main criteria to determine the country classification of companies: • Country of incorporation • Primary listing • Point of major trading turnover However, economic exposure does not commonly feature in the index criteria. In “normal” cases, this works sufficiently well not to cause too many problems, but with the increasing internationalization of listings, certain scenarios can lead to undesirable results, as the two hypothetical examples below highlight. Example 1: A corporation operating exclusively in an emerging market country seeks a stock market listing in a developed market. The company’s risk exposure is exclusively in its original home market. However, the standard practice employed by index providers will typically classify the stock according to its primary listing. In many cases, the company will also continued on page 61

Leveraged_Indext = Figure 3

INDEX

0.16

0.12

0.08

INDEX_25.pdf

12/12

12/11

12/10

12/09

12/08

Source: Stoxx

July / August 2013

JUL-AUG

12/07

12/06

12/05

12/04

12/03

12/02

12/01

12/00

0.00

12/99

0.04

25 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

www.journalofindexes.com

0.20

12/98

In the new formula, the daily financing spread between the one-year Euribor rate and the annualized EONIA swap rate is an additional deduction from the index’s performance. Figure 2 illustrates the relatively minor difference in performance between the two leveraged index methodologies for the Euro Stoxx 50 Net Return Index over a 15-year period. On a monthly basis, however, in times of high liquidity spreads, as we have observed from 2007 onward, the monthly performance differential between the two index methodologies may exceed 10 basis points, as Figure 3 illustrates. While differences in financing costs made little impact until the financial crisis, Figure 3 shows that the choice of

Monthly Performance Drag In New Leveraged Index Version

Monthly Drag (%)

Leveraged_Indext -1 * Base_Indext - daily_interest_rate (2 * Base_Indext -1 - daily_financing_spread)

Trim Size = 8.125 X 10.875 Wed Jun 5 11:19:00 EDT 2013



ASRS Asset Allocation 4/30/2013

■ U.S. Fixed Income ■ EM Fixed Income ■ Private/Opportunistic Debt ■ U.S. Equities ■ Non-U.S. Equities ■ Private/Opportunistic Equities ■ Infation-Linked Assets

Source: ASRS

risk-factor ETFs, and in early 2012, began to develop a program. Russell announced later in 2012 that it would exit the factor ETF space. Due to our relationship with BlackRock, we approached them with the conceptual program we had developed thus far. BlackRock queried their institutional clients about potential similar interest in access to risk-factor-specific ETFs, and found definite interest. MSCI, another of our strategic partners, likewise solicited opinions from its institutional clients. They also found broad interest. Concurrently, we began working with Joe Mezrich of Instinet/Nomura, about developing a signaling mechanism for identifying riskfactor regimes. Nomura is one of the few investment banks in the forefront of alternative beta research. These relationships allowed us to put together essentially a collaborative partnership and design what we call the “Equity Asset Risk Factor Paradigm.” We repurposed a portfolio already with BlackRock that had been originally envisioned for the now-defunct Russell ETFs, or similar vehicles. Funding the iShares ETFs wasn’t something we did out of the goodness of our hearts: Everyone involved needed something, and it just worked—all the planets were aligned, so to speak. We’re employing the risk-factor ETFs as an overlay strategy on our total equity construct. We intend to adjust the proportions in a given factor-specific ETF or ETFs to address our active risk and the directional regimes of the risk factors. We believe we now have a vehicle with which we can easily accomplish that. You should note that there’s also a quality-factor ETF in registration right now that will track the MSCI USA Quality Index from MSCI’s Risk Premia index family. The portfolio has been set up, and will convert to the ETF over the summer. We funded that portfolio at the same time as the other iShares ETFs. JOI: How did ETFs fit into your portfolios before you seeded the funds? Underwood: We have not used ETFs in any meaningful fashion previously, although our external GTAA (global tactical asset allocation) managers do employ them. We do manage some alternative beta strategies internally and more cheaply www.journalofindexes.com

INDEX

than what can be done externally, certainly far less expensively than heretofore-existing ETFs. Also, with ETFs in and of themselves, we were just a little too big to really use them in-house. We had the latitude in our in-house portfolios to use them to equitize cash, but we found index futures were more fluid and involved lower tracking error. Very much to their credit, BlackRock, in the course of developing the risk-factor iShares ETFs, listened to our and other institutional investors’ suggestions and priced these ETFs at 15 basis points. JOI: How much of the ASRS portfolio is indexed or passively managed? Underwood: Excluding our GTAA investments, we’re in about the low to mid-sixties in terms of overall percentage of what I would call beta or passive—essentially pure index—across the asset classes in which we use passive management. Just to give you some hard numbers, we’ve got about $16.5 billion of the total plan in what we categorize as public equities. And we’re at $30.8 billion in total assets, with roughly 55 percent in public equities. In the domestic large-cap equity subclass, we’re roughly 80 percent indexed, but in small-cap, we’re about 30 percent indexed. Across all public equities, on a combined basis, we are about 62 percent indexed. We’re a hybrid, in a sense, in our public equity classes. We manage domestic equity assets internally, essentially replicating pieces of the S&P 1500. We have an S&P 500 portfolio, midcap growth and value portfolios, and a small-cap portfolio. A prior S&P 500 sampling portfolio was split last year into two pieces and repurposed 60 percent to replicate the MSCI High-Dividend Yield Index and 40 percent to replicate the MSCI Minimum Volatility Index; they operate together in a systematic fashion. The relative proportion between the two portfolios can be adjusted within 20 to 80 percent ranges, depending upon the volatility of the equity markets. We can also move assets from the other part of the beta portfolio set, the S&P portfolios, to increase this component, or shrink it in more traditional market environments. Externally is where we use active management, and that’s a mixture of quantitative and fundamental disciplines. However, we don’t manage indexed non-U.S. portfolios inhouse—BlackRock does that for us and replicates the MSCI EAFE, EAFE small-cap and emerging markets indexes. In general, we start with a big beta component across the system, although its relative size shrinks within various asset subclasses. JOI: Do you use more active management when you feel there are more opportunities or inefficiencies in the market? Underwood: That is the mode we’re in now. Every three years, the plan redoes its Strategic Asset Allocation policy, or SAA. It’s essentially the blueprint of all the asset classes in which the plan can invest. Around each target allocation are bands providing some latitude for continued on page 62 July / August 2013

JUL-AUG

INDEX_27.pdf

27 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Figure 1

Trim Size = 8.125 X 10.875 Tue Jun 4 14:18:29 EDT 2013



Figure 1

South Korean Gross National Income Per Capita

Current International $

45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5000 0 ‘80

‘84

‘88

‘92

‘96

‘00

‘04

‘08

‘12

■ High Income Countries ■ South Korea ■ All Countries Source: World Bank Note: South Korean GNI growth closely tracks other high-income and developed markets

www.journalofindexes.com

INDEX

South Korea: • is the 15th-largest economy measured by GDP, industrial output and services • is the 12th-largest economy in terms of purchasing power parity • is the 7th-largest exporter and 10th-largest importer in the world • is the world’s 13th-largest stock market and the 3rdlargest in the Pacific Rim (after Japan and Australia). • is home to some of the world’s biggest and most successful corporations, such as Samsung and Hyundai Motors, and it is the world’s 4th-largest car producer and world’s largest shipbuilder • has a thriving high-tech industry and a high-tech population, being the first country to achieve more than 50 percent Internet broadband adoption among the population, and is the most wired country on the planet in terms of Internet use and speed • is a member of the OECD, and the World Bank classifies South Korea as a high-income developed country with a developed market The strength and size of its economy rank it among the G-20 major economies. Sources: The World Bank, The Economist

evolved and foreign participation increased. However, the simultaneously ongoing 2008-09 financial crisis made many governments around the world more cautious about opening their markets to foreign investors. Although this has delayed the expected final improvements in South Korea’s market, the FTSE policy group continues to believe South Korea is a developed market. This is because, while the use of detailed technical market criteria to assess the development of markets has helped to give additional objectivity to the creation of indexes, such criteria are not the only factors in FTSE’s decision-making process. Unlike some other indexers, FTSE does not regard country classification as merely an exercise in box-ticking. Broader considerations of a country’s overall standing in the global economy, and the perceptions of investors, are also essential to the final decision. The manifest economic strength of South Korea is a key consideration to asset owners and investors in forming their opinions of how the country should be classified, and FTSE believes that the majority of such users no longer classify the country alongside the likes of China, India and Brazil. This perception, together with the objective economic and market factors, informs the decision to classify South Korea as a developed market. When investors think of investing in emerging markets, FTSE believes it is developing countries they mean to target. An emerging market index that includes an economy as advanced as South Korea would simply create, in FTSE’s view, unwanted distortions that fail to serve investors’ purposes. It is no surprise, therefore, that the classification of South continued on page 31 July / August 2013

JUL-AUG

INDEX_29.pdf

29 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

for classification changes, including a “watch list” of countries whose status might change) • Stability (a country’s change in status would only reflect permanent changes in market status and global acceptance) • Market access (no discrimination against nondomestic investors) FTSE established a new Country Classification committee to assess the information and adopted a transparent quality of markets matrix to objectively judge and compare markets and to provide consistency and add predictability to the classification change process. In addition, FTSE adopted a “watch list” for countries potentially approaching reclassification as well as a policy of engagement with the markets in such countries, to help them understand the steps they would have to take to make themselves eligible for promotion. On these bases, the Country Classification committee makes recommendations to FTSE’s policy group regarding country classification changes, and the FTSE policy group makes the final decision on classifications. In 2009, employing this governance structure, FTSE’s policy group, which comprises index experts and users drawn from senior levels of practitioner firms, promoted South Korea from emerging to developed status. The decision reflected the size and sophistication of the South Korean market at the time and what the policy group believed to be the general market perception: that it was now a developed market (see box). The FTSE policy group recognized at the time that the South Korea market had some trading and implementation limitations. For example, in the FTSE quality of markets evaluation process, while South Korea achieved a “pass” rating on 18 of the 21 criteria, on the following three criteria it remained “restricted”: the convertibility of the currency (i.e., the won) outside domestic banking hours; difficulties with free-delivery transfers of stock, particularly in pooled funds; and a facility for foreigners to trade among themselves in securities that have reached a foreign-ownership limit. Nevertheless, in line with the expectations of international market professionals, the FTSE policy group took the view that these outstanding issues would be addressed as the South Korean market

Trim Size = 8.125 X 10.875 Tue Jun 4 14:18:39 EDT 2013



References Bhardwaj, G. and A. Dunsby (2013), “The Business Cycle and the Correlation between Stocks and Commodities,” SummerHaven Investment Management, available at: https://www.summerhavenim.com/guest/our-research.html. Bernanke B., M. Gertler and S. Gilchrist (1996), “The Financial Accelerator and Flight to Quality,” Review of Economics and Statistics, 78: 1-15. Erb, C. and C. Harvey (2006), “The Strategic and Tactical Value of Commodity Futures,” Financial Analysts Journal 62(2): 69-97. Gorton, G. and G. Rouwenhorst (2006), “Facts and Fantasies about Commodity Futures,” Financial Analysts Journal 62: 47-68. Kiyotaki N. and J. Moore (1997), “Credit Cycles,” Journal of Political Economy, 105(2): 211-248. Masters M. (2008), Testimony before the Committee on Homeland Security and Governmental Affairs, U.S. Senate, May 20. Tang, K. and W. Xiong (2010), “Index Investment and Financialization of Commodities,” NBER Working Paper 16385.

Endnotes 1

See Gorton and Rouwenhorst (2006).

2

We use weekly asset returns to generate realized correlations over annual windows. For a detailed discussion of the methodology, see Bhardwaj and Dunsby (2012), available at https://www.summerhavenim.com/guest/our-research.html.

3

Estimates of half-life are based on the regression results reported below.

4

For a complete list of commodities and the period of data availability of individual commodities, see Bhardwaj and Dunsby (2012). In 1962, when this study starts, there are nine commodities in the sample. This number grows to 25 with the addition of the natural gas in May 1990.

5

For stocks, we use the percent change in the S&P 500 price index. Dividends are not included.

6

A 12-month period is identified as recessionary if it contains at least one recessionary month. This is why the entire period January 1980 through November 1982 is grayed, even though it contains two distinct recessions. Default spread is defined as the average of difference of Moody’s Seasoned Baa Corporate Bond Yield (BAA), and Moody’s Seasoned Aaa Corporate Bond Yield (AAA).

7

Figure 5 retains its same basic shape if only the nine commodities that exist for the entire sample are included, though the overall level of correlation is higher since most of the full sample commodities are agriculturals. GDP growth is measured over the same period as the dependent variable.

8

See the models of Bernanke et al. (1996), and Kiyotaki and Moore (1997).

9

We don’t include a similar analysis for the bond-commodity correlation since this series does not exhibit a business cycle component.

European ETF range expanding from 10 to 45 in number.

by Invesco in January 2006. Fulton ascended into his current role as global head of ETFs in 2009, when Bruce Bond resigned. Fulton declined to discuss what exactly he planned to do next, but he did say he would remain in the ETF industry, though not as a sponsor.

Costandinides Joins Markit

Ebner To Lead SSgA Product Development Scott Ebner, State Street Global Advisors’ current global head of ETF product development, is to become the fund manager’s head of global product development and research. Ebner will start in the newly created role later this year and will relocate to Boston. His role will continue to include ETF product development, says State Street. Previously at NYSE Euronext, Ebner moved to London in October 2010 as State Street’s European head of ETFs. Since the first half of 2011, the firm’s European ETF market share has grown from 0.3 percent to 1.6 percent, according to data from db X-trackers, with State Street’s

It was announced in April that Christos Costandinides, former ETF strategist at Deutsche Bank, was to join Markit as a director in the firm’s index business in May. In his new role, Costandinides will be publishing research on the company’s indexes, which include the iBoxx family of bond indexes and the iTraxx and CDX families of credit derivative indexes. He reports to Sophia Dancygier, head of Markit’s iBoxx index range. Costandinides left his previous job at Deutsche Bank in October last year. Previously, he worked in the fund ratings business at Moody’s Investors Service.

iShares’ Bednall Departs iShares’ product development director has left the firm, according to sources. Managing director Martin Bednall left ETF provider iShares last month, according to sources who wished not to be identified. His future plans are unclear.

www.journalofindexes.com

INDEX

Prior to joining iShares, Bednall was at Barclays Global Investors from 2000 to 2006 as a transition strategist, before taking on the role of integration director until 2008. BlackRock acquired BGI in 2009, when Bednall became product development director for Europe, the Middle East and Africa. Before joining BGI, Bednall worked at State Street. A representative for the firm declined to comment.

SocGen Index Trading Head Leaves Benjamin Fussien, head of portfolio, ETF and index execution at French bank Société Générale, left the firm in April, according to market sources. An employee of the equity trading desk at Société Générale confirmed that Fussien had left the bank. Fussien had held the position for two years, prior to which he was global head of ETF sales at Société Générale Securities, the bank’s brokerage arm. Société Générale owns Lyxor, the third-largest ETF issuer in Europe by funds under management. July / August 2013

JUL-AUG

INDEX_45.pdf

45 .125 from Trim Trim line Bleed line

blackmagentacyan

News continued from page 15

Trim Size = 8.125 X 10.875 Wed Jun 5 11:21:37 EDT 2013



W

hat is the appropriate benchmark for a globally diversified portfolio? Unfortunately, this simple question doesn’t have a simple answer. We address a number of issues that make this question hard to answer and propose a solution: the Prudent Target Risk Index (PTRI) portfolios. We start by analyzing the available index solutions. There are two classes of global asset allocation indexes: the prescriptive and descriptive. The prescriptive indexes provide model-based global allocations that reflect the views of the model producer. Such prescriptive indexes may not be representative of the views of the broader investment industry. The descriptive indexes seek to reflect some particular characteristics of the financial markets or investment industry. One style of descriptive indexes uses the relative capitalization of global financial markets to derive the market portfolio of all asset classes. Such capitalization-weighted indexes exist for world equity markets and the world bond markets. However, there is no theoretical consensus on how alternative investments should be reflected in the world total market portfolio, or exactly how the intermediate-risk-level portfolios should be constructed. Accordingly, a complete index portfolio of liquid capital-weighted investable world markets does not exist. Another style of descriptive method is to aggregate the asset allocation of global managers to create target-risk indexes. Some index providers analyze the underlying holdings of the global asset managers to produce their set of target-risk index portfolios. These portfolios have just a few fixed equity levels and are updated infrequently. The PTRI is a descriptive portfolio index and uses the publically available information from both holdings and returns of the U.S.-based global mutual fund managers. PTRI seeks to provide a diversified, adaptive, consistent and comprehensive portfolio allocation solution that reflects the changing nature of global investment opportunities and allocation of the institutional advisors. PTRI is a diversified solution because it uses the broadest global asset classes as reflected by investment choices of the top institutional mutual funds that combined have more than $2 trillion in AUM. The PTRI is an adaptive solution because it dynamically changes to reflect updated institutional allocations. The PTRI is consistent because the out-of-sample performance of its portfolio tracks the average return profile of peer mutual funds with comparable

allocation to risky and capital preservation assets. Lastly, the PTRI is comprehensive because it provides a set of global asset allocation benchmarks for portfolios that span the entire risk frontier from 0 to 100 percent allocation to risky assets, and covers more than 95 percent of the liquid investable long-only global financial markets. Asset managers, financial advisors and investors can use the PTRI for two purposes. First, they can benchmark their own investment strategy using the appropriate PTRI risk-level portfolio. For instance, investors with a 67 percent portfolio allocation to risky assets can benchmark their portfolios against the performance of the PTRI 67 percent risky asset portfolio. Second, investors and advisors can use the appropriate PTRI risk-level portfolio as an open-architecture building block for their own custom portfolios. If investors have specific economic views about the future performance of specific asset classes, then they can tweak the asset class allocation relative to the PTRI benchmark. Once they decide on the asset allocation, the investors can choose the appropriate securities to invest in each asset class. The PTRI results in this article reflect performance using passive-asset-class index ETFs. However, the investors could alternatively use a combination of active fund ETFs, mutual funds or separately managed accounts managers—or even select a set of individual securities to invest in specific asset classes. The rest of the article discusses the choice criteria of specific funds selected from the mutual fund industry to perform the analysis. It describes the choices of the asset classes and indexes used. It further details the methodology of style analysis for each mutual fund, and explains how the style allocation of the industry is then aggregated to construct the PTRI. We conclude with an empirical test of the external validity of the PTRI.

Mutual Funds Selection We begin the process of constructing the PTRI by first specifying the criteria for selecting the mutual funds upon which the benchmark is based. For our analysis, we use institutional share class mutual funds having at least $50 million in assets under management, and at least two years of return history. The selected mutual funds need to have between 20 and 90 percent of their assets invested in the U.S., and no more than 20 percent of their net capital exposure to derivatives.

Figure 1

Category AUM (Million Dollars)

Mean

Standard Deviation

25th Percentile

50th Percentile

75th Percentile

1,425.16

4,676.47

64.64

222.33

697.91

Turnover Ratio

68.80

96.37

18.00

36.00

78.00

Years Since Inception

11.23

10.46

4.84

8.01

15.31

Risk Level

57.88%

29.24%

39.14%

62.81%

81.27%

Foreign Asset Exposure

45.7%

16.45%

36.46%

44.74%

54.82%

Sources: Razor Hedge, SEC Edgar, Yahoo Finance

www.journalofindexes.com

INDEX

July / August 2013

JUL-AUG

INDEX_47.pdf

47 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Summary Statistics

Trim Size = 8.125 X 10.875 Wed Jun 5 11:22:26 EDT 2013



asset-class indexes. The solution at which one arrives depends on the type of investment funds one is trying to classify. Since we are trying to analyze global asset allocation portfolios, we want to use asset-class indexes that drive global risk premium. Accordingly, we divide the assets into three broad asset classes: bond; equity; and alternatives. Within each broad class, we solve a complex puzzle of finding the best set of subindexes to use and impose a number of constraints. The first constraint is practical: We limit the choice to asset-class indexes that have liquid tracking index ETFs and mutual funds. The reason to limit the liquid investable indexes in this way is that we want to make PTRI an investable portfolio. The second constraint is to construct an exhaustive yet mutually exclusive set of risk factors. Here we try to strike a balance between the completeness of the coverage and the law of parsimony. On the one hand, a large-enough set of asset classes is needed to minimize the bias of missing risk factors; on the other hand, the set should be as small as possible to improve the efficiency of the identification and to maintain model stability out of sample. As an example, we will use U.S. equity indexes to demonstrate the type of analysis we performed. There are two major index families that can represent the U.S. equity asset classes and have the largest following: the S&P and

the Russell indexes (Figure 2). The two index families offer two alternative perspectives through which to view the U.S. equity market. The Russell U.S. indexes have a broader capitalization coverage of 3,000 securities, which is split into the Russell 1000 and 2000. The S&P benchmark family’s flagship index is the S&P 500, which is the most popular index by number of managers, and also has the highest AUM tracking any single index. The broad S&P U.S. market index covers the top 1,500 securities. Both of these index families provide many different combinations of possible growth/value and large/mid/small-cap subindexes. We first perform multicollinearity analysis to find just the right set of subindexes to use. We use the adjusted R-square and variance inflation factor to measure the multicollinearity effects of introducing each additional asset-class index given a set of asset-class indexes. Our analysis confirms that the Fama-French size and value distinction matters in explaining investment behavior of investment managers. What is a little more surprising is that midcap indexes have distinct returns behavior that cannot be represented as a simple linear combination of large-cap or small-cap indexes. Given the multicollinearity analysis, we review five different ways to view the asset-class indexes, and find the mix that represents the best trade-offs.

Figure 3

Indexes Used In PTRI Index

Commodity REIT

Proxy

DBIQ Optimum Yield Diversified Commodity Index Excess Return

DBC

MSCI US REIT Index

VNQ

Emerging Market Equity

MSCI Emerging Markets Index

VWO

Developed Market Equity

MSCI EAFE Index

EFA

S&P 500 Growth Index

IVW

S&P 500 Value Index

IVE

S&P MidCap 400 Growth Index

IJK

US Large Growth Equity US Large Value Equity US Middle Growth Equity US Middle Value Equity

S&P MidCap 400 Value Index

US Small Growth Equity

Russell 2000 Growth Index

US Small Value Equity

IJJ IWO

Russell 2000 Value Index

IWN

Developed Market Bond

Barclays Global Treasury ex-US Capped Index

BWX

Emerging Market Bonds

JPMorgan EMBI Global Core Index

EMB

iBoxx $ Liquid Investment Grade Index

LQD

US Corporate Bond US High Yield US Short-Term Treasury Bonds

Barclays High Yield Very Liquid Index

JNK

Barclays U.S. 1-3 Year Treasury Bond Index

SHY

US Medium-Term Treasury Bonds

Barclays U.S. 7-10 Year Treasury Bond Index

IEF

US Long-Term Treasury Bonds

Barclays U.S. 20+ Year Treasury Bond Index

TLT

US Agency Bonds US Mortgage-Backed Securities Bonds US Municipal Tax-Free Bonds Cash

Barclays Agency Bond Index

AGZ

Barclays U.S. MBS Index

MBB

S&P National AMT-Free Municipal Bond Index

MUB

Barclays U.S. Short Treasury Bond Index

SHV

Source: ETFDb.com

www.journalofindexes.com

INDEX

July / August 2013

JUL-AUG

INDEX_49.pdf

49 .125 from Trim Trim Line Bleed Line

blackyellowmagentacyan

Asset Class

Trim Size = 8.125 X 10.875 Wed Jun 5 11:21:45 EDT 2013