CFA协会会议论文集季刊

时间:2014-10-13 13:41 作者:cfa 来源:cfa

Abstract

The reasons investors use exchange-traded funds (ETFs), rather than other alternative instruments, as part of their active equity strategies are increasing. In North America and Europe, ETFs are becoming popular in managed portfolios and for portfolio cash management. Institutional inertia is being replaced by an appreciation of the low-cost opportunities of using ETFs.

Editor’s Note: Noel C. Archard, CFA, works for an ETF provider.

This presentation comes from the 67th CFA Institute Annual Conference held in Seattle on 4–7 May 2014 in partnership with CFA Society Seattle.

Institutions were the early adopters of exchange-traded funds (ETFs). Traditionally, institutions used ETFs as trading tools, taking advantage of some of their early liquidity factors. That is still the reality today, but the base has widened. ETFs have crept into all facets of institutional life.

There are many ETF options currently available, including 5,100 products globally, representing over $2.5 trillion, and 1,500 products in the United States, representing more than $1.7 trillion. According to a Greenwich survey conducted in early 2013,1 almost everyone in the institutional world at least expects to use ETFs. Half of the institutional ETF users polled—insurers, asset managers, and institutional funds represented by pensions and endowments—expected to increase their use of ETFs by the end of 2013.

The factors that institutions consider when choosing ETFs have been evolving over time as well. Liquidity and trading volume have often been in the forefront, and although cost seems like it would be a big driver for institutions, it is overshadowed in importance; institutions continue to rate liquidity and benchmark over price for this type of exposure. In the early days, institutional access was perceived to be restricted, but liquidity has always been an important characteristic of ETFs. Many institutional investors and asset managers find it valuable to basically break their trades and hide in the liquidity afforded to them by some ETF products—whether that is in equity, emerging markets, international, small cap, or fixed income, such as Treasury bills and credit.

A lot of focus for investors remains on the average daily trading volume, and the fund’s expense ratio is factored in when ETFs are compared with competing instruments, such as futures and options. In some cases, even assets under management are a factor. Table 1 shows the most important criteria in ETF selection as reported in the Greenwich survey.

Table
Table 1. Most Important Factors in ETF Selection from a Survey of Institutional Investors
 
Table 1.Most Important Factors in ETF Selection from a Survey of Institutional Investors
Factor Percentage
Liquidity/trading volume 76%
Expense ratio of fund 68
Tracking error of fund 49
Benchmark used 39
Breadth of ETF offerings 22
NAIC ratinga 17
Fund company 16
Assets under management of ETFs 10
Servicing by sponsoring organization 6

aThe National Association of Insurance Commissioners (NAIC) does not endorse or recommend any securities or products. NAIC designations are issued for special regulatory purposes, and these designations are not equivalent to credit ratings issued by nationally recognized statistical rating organizations. NAIC designations are suitable only for NAIC members.

Note:Participants were asked to select their top three criteria; the results reflect which criteria were selected.

Source:Greenwich Associates, “Institutional Investors’ Relationship with ETFs Deepens” (2013).

The country and region of the investment and the accompanying regulatory choices have an impact on ETF choice as well, especially regarding the size of the ETF and how much an institution would own of that ETF if it invested. In some cases, regional external and internal restrictions based on firm policy dictate how much of an investment vehicle an institution can own. Such policy-based restrictions do not always make sense because at their heart, many ETFs are index funds. They are overseen by independent boards and are different from most financial products. The things institutions tend to focus on are an instrument’s liquidity, its overall cost—not surprisingly—and, as a close third, tracking error. How well the vehicle follows the mandate of the institution is particularly important from a cash equitization perspective or from a global macro perspective. With 1,500 products just in the United States, there is a significant range of exposures to select from and, frequently, multiple benchmarks to choose from.

ETF Uses in Asset Management

ETFs are becoming popular in a wide range of areas within the industry, including funds of funds, cash management, futures, and benchmark decomposition. In these areas, ETFs have started to replace more traditional products because of their unique benefits.

ETF Managed Portfolios.

ETF managed portfolios, which are investment strategies that typically have more than 50% of assets invested in ETFs, are one of the fastest-growing segments in the United States. Many funds of funds, which used to be composed of mutual funds, are now being constructed with ETFs. Sometimes, a broader palette and easier access are attractive characteristics for a fiduciary.

Other funds of funds, however, are finding themselves competing against the ETF model. There is about $100 billion in the ETF managed portfolio realm, which is growing at a faster pace than traditional funds of funds. As more and more sponsors enter the ETF market, the expenses of those products continue to be compressed. From a liquidity perspective, managers are able to use ETFs within secondary market trades to construct portfolios and avoid end-of-day net asset value issues, structured cash residuals, and other similar challenges.

The landscape of broad institutional ETF use, from portfolio management to delivery vehicles, is stretching out into the retail space; asset managers are using ETFs to construct the next generation of solutions-based products because of the diversity of these products, which I will discuss later. ETFs can be used to fine-tune a portfolio; for cash management, which is an advanced strategy deployed by institutional users; and as a substitution for futures.

Cash Management.

All investors usually have to deal with cash in their portfolios in some capacity or another. Obviously, cash buffers are necessary. Sometimes money comes in but cannot be deployed. Sometimes cash is needed for client redemptions. Sometimes cash is in the portfolio because of various asset management programs the manager is involved in, although the manager may have to hold it and release different portions to a subadviser.

At BlackRock, we have looked at portfolio performance over various time scales against US small cap, US large cap, emerging market, and EAFE. If cash has to be deployed in five days, what type of drag does this put on the portfolio? Again, I do not think this will come as a surprise to anybody. We are all in the market because we believe that over the long term, the returns are going to continue to increase. We are going to deliver growth for our clients. So, sitting on cash is generally a bad thing. Furthermore, the performance numbers vacillate depending on what the market is doing. A lot of studies have been conducted about the 10 or 12 best days in the market over the course of the year. That is, there are a handful of days each year that end up driving returns, and if an investor misses out on those days by not being fully invested or taking too long to deploy cash, there is a drag on the portfolio that ultimately affects the investor’s annual performance.

We looked at small-cap managers over a number of years, considering funds with at least three years of performance and multiple reported cash weightings. Funds with more than a 10% cash weighting were excluded; that much cash might have a strategic purpose within the portfolio. On average, US small-cap blend equity managers had nearly 3% cash consistently over the last three years, as Figure 1 shows. What does 3% cash mean? Tracking error over three years with a 3% cash position could basically average out to 50 bps of tracking error, as Figure 2 demonstrates. We created mock portfolios with a cash position and assumed no transaction costs for holding the cash. We included the iShares Russell Small Cap and the Russell 2000, and we included transaction costs of the holding. If a fund holds one of the most liquid ETFs in the US market instead of cash, the tracking error drifts down to 1 bp against the small-cap mandates. Over time, ETF spreads have narrowed and the costs have become lower as they have continued to compress over time. With today’s ETFs, investors have the choice of incredibly tight tracking to the benchmarks, which can be fairly significant when carrying a 50 bps tracking error over multiple years versus 1 bp by locking in the benchmark return and trying to deliver alpha over it.

figure
Figure 1. Distribution of Active Manager Cash Allocations

Notes: Cash levels shown for 108 funds with at least three years of performance and at least 12 reported cash weightings. Managers with an average cash level above 10% were excluded. Data are as of 31 March 2014.

figure
Figure 2. Rolling Three-Year Tracking Error with 3% Liquidity, March 2004–March 2014

Notes: The cash portfolio assumes an allocation of 97% to the Russell 2000 Index and 3% invested in a money market fund, rebalanced monthly. No additional transaction costs are assumed. The ETF portfolio assumes an allocation of 97% to the Russell 2000 Index and 3% invested in iShares Russell 2000 ETF (IWM) using ETF market price total returns, rebalanced monthly, and includes 2 bps of roundtrip transaction costs for the ETF investment per month. Data are as of 31 March 2014. This is for illustrative purposes only and is not indicative of any actual portfolio or asset allocation model.

The difference between first- and second-quartile performance can be slim. For example, it is 159 bps over three years and 160 bps over five years. Of course, long-term performance of the portfolio is most important, but no one is immune to the normal ratings and reviews and even, internally, how the portfolio behaves over short periods of time. The three-year cash drag with 3% cash is about 37 bps. If 37 bps were the arbitrary cutoff, that means 20 managers would have moved up an entire quartile in their ratings, from the second quartile to the first quartile, with the third quartile creeping up into the second quartile. It is not all about the rankings, but clearly, it is about leaving some money on the table—leaving performance on the table. If managers are doing a good job delivering alpha, the last thing they want is the cash they need to manage portfolio operations to erode a portion of that alpha component. This principle applies across the board, not just in small cap but also in other investment disciplines and asset classes.

Futures.

Managers who do not want to use ETFs for cash management in their portfolios have to select another option.

At a larger institution or asset manager, or in situations where portfolio management is separate from the trade desk, investors have often been more interested in using futures for cash equitization or even for asset management exposure than they are in using ETFs because, in general, futures have a much lower cost of execution. Over the last year or so, that trend has actually reversed.

Historically, to get exposure to the S&P 500 Index using futures or ETFs, futures were the clear winner, usually coming in somewhere around 7 bps of costs on average—that is, taking into consideration commissions going in and out of the products, the roll costs over the course of the year, and the tracking error embedded in the future. ETF prices in the early years were usually in the teens or a little bit higher after commissions and spreads were factored in. But over the last few quarters, it has started to be more cost-effective to use ETFs than to use futures. Futures have been trading rich, particularly in some of the developed countries. Some significant savings can be achieved by considering ETFs over futures.

ETF pricing is very simple. The price of an ETF is just the value of the underlying basket; the premium or discount relative to the underlying basket is fair value. Generally, premiums and discounts on ETFs get flushed out of the market because of the redemption mechanism and the constant arbitrage going on between the value of the basket and the market price of the ETFs. Under the efficient market hypothesis, any discrepancy quickly goes away.

In a situation of rolling futures contracts, futures may or may not accurately capture the market price. To get into the futures contract, the manager has to consider the spot price plus the financing costs. If the cost of financing is greater than the earnings, a negative cost of carry is priced into the future.

In this environment, with a lot of different forces at play, the manager is stuck with a persistent low interest rate. Dealers’ costs are changing for many different reasons, including requirements implemented by the Dodd–Frank Act and Basel III. Those costs are raising the cost of futures and causing some to question the role of futures in a portfolio. Futures are becoming more expensive quarter to quarter than ETFs, which are open-end vehicles. ETFs have no term date and do not have to be rolled. They are not risk-free products, but ETFs are starting to demonstrate many advantages over futures.

One of the factors contributing to the flagging investor interest in equity futures is the market funding spread. Money market rates have been below Libor, with a few short-term exceptions. This pattern will probably continue for some time. Although there are some shifts happening in yields and rates, no persistent jumps are expected in the next 6–12 months. Finding true risk-free cash vehicles is becoming challenging. In addition, investors continue to be willing to take on direct equity risk as the wider economic indicators remain positive.

In the equity futures markets, there is also a little bit of dividend forecasting error. More and more companies are going through buyback and special dividend processes. Special dividends are significant considering their impact on single-stock futures, but less so for index futures. It obviously takes a fairly sizable shift to cause a ripple through index futures, but it does happen—for example, with Apple’s special dividend. Such a dividend adds a little bit to the volatility of the index futures market and could occur more and more in the future as similar dividends increase in popularity.

Internationally, the dividend patterns are less consistent than in the United States, which creates more risk in pricing. There are more futures covering the various international markets, representing an opportunity. In both domestic and international markets, it is vital to compare futures with a simpler investment route, such as an ETF, to determine which is the most cost-effective.

Other factors lessening the popularity of futures include dividend risk, which depends on the tax withholding situation; interest rate risk, which is obviously embedded in the futures; and in some cases, the potential for a cash drag in ETFs. One of the benefits of an ETF is its transparency; managers can make sure that ETFs are fully invested all the time. Historically, futures often have the advantage on the equitization front. It is worthwhile to fight the inertia of just using the product that has always been used and instead take a look at the opportunities within the ETF market to try to save some money and ultimately enhance performance.

Benchmark Decomposition.

More and more institutions and asset managers are engaging in benchmark decomposition. In the past, ETFs rarely came up in conversations with institutional investors, primarily because of the cost or because of lack of comfort with the vehicle. In the very early days, some of the larger institutions viewed ETFs as more of a retail product and not something that they would want to be involved in. But some of the largest, most sophisticated institutions—particularly on the pension and endowment side—were the earliest adopters of ETFs because they could buy them in bulk. They could use slivers of a portfolio strategy to quickly increase or decrease their allocation exposure without using a swap or some other derivative. In many cases, the boards that these institutions were working with had a higher comfort level with ETFs. The regulatory framework around ETFs makes them slightly more approachable and understandable to some institutions than alternative approaches using derivatives.

Today, more and more ETF products are in the marketplace. Virtually any allocation that can be created by an institution can now be broken down and realized using ETFs.

Many institutional investors around the globe have been having problems deploying cash allocated for real estate and infrastructure over the last few years. Some of the traffic has either dried up or slowed down, and the capital has been harder to execute. So, managers have been looking at a combination of both liquid and illiquid strategies to bridge the implementation gap.

In some situations, an institutional investor might be investing client mandates through a third party that has limited ability to implement changes directly. It is possible to implement the desired exposure using ETFs very efficiently. With ETFs, changes in strategic allocations of client mandates can be implemented in virtually every asset class. For example, at BlackRock, we are seeing equity managers who might run balanced mandates of equity and fixed income using ETFs more and more by using some portion of their fixed-income exposure as a smoothing vehicle to protect the value of their equity mandates. This technique works just as well in the corporate bond arena, where investing in corporate bonds is sometimes better than investing in the underlying equity securities from the perspective of costs and liquidity.

We are also seeing a fundamental breakdown of regional investing or of broad asset classes like emerging markets, even within single countries. Investors in emerging markets—commodity producers versus commodity consumers—behaved very differently over the last few years and through this market cycle. ETFs give investors the ability to break exposure down to the country level and try to take advantage of any short-term inefficiencies, rather than just getting broad emerging market or developed market exposure.

A sad but very stark example is when the tsunami hit Japan a few years back. An immediate and sharp selloff of Japan occurred because no one knew how extreme the damage was going to be. The iShares MSCI Japan ETF (EWJ), which covers that space, traded very heavily; the activity happened overnight. When the markets opened the next day, the Japanese markets were closed. Many clients went long EWJ because they thought that the market reaction was actually too severe—a buy into what seemed like a suddenly depressed or potentially undervalued situation. Obviously, there is a risk there: If the damage had been more extreme, it could have snapped back the other way. But we have example after example of how a basic overweight/underweight position in a portfolio can be a fundamental way of enhancing returns to quickly allocate some capital or excess cash or of refining a portfolio to get a little more insight into the investment.

Table 2 shows some of the most liquid exchange-traded funds in the United States. Interestingly, ETF volumes on the NYSE can be anywhere from around 15% of daily volume up to 20%. On heavy volume days, spikes up into 30% of daily trading volume can occur. A lot of buyers and sellers are meeting each other in the middle. Compared with the value of the underlying basket, the costs of gaining small-cap exposure in a portfolio—or exposure to EAFE or emerging markets—are often reduced when using ETFs. In many cases, the ETF is trading at what comes to a 1–2 bp spread, and the alternative is buying the underlying securities, where the spreads could run anywhere from 10 to 30 bps. This happens in fixed income as well. People tend to get into ruts and look at equity exposures without thinking about fixed income; fixed income can be an even more interesting story in this space.

Table
Table 2. Bid–Offer Spreads of Liquid ETFs in the United States
 
Table 2.Bid–Offer Spreads of Liquid ETFs in the United States
Symbol iShares ETF Name iShares Bid–Offer Spread (bps) Estimated Basket Bid–Offer Spreada(bps) iShares Average Daily Volume (millions) iShares AUM (millions)
IWM Russell 2000 1 10 $3,971 $28,271
EFA MSCI EAFE 2 14 1,013 52,826
EEM MSCI Emerging Markets 2 31 2,478 40,125
EWZ MSCI Brazil Capped 2 45 574 4,378
FXI FTSE China 25 3 23 692 6,062
CSJ Barclays 1–3 Year Credit 2 20 64 11,809
LQD Markit iBoxx USD Investment Grade Corporate 2 30 192 15,680
HYG Markit iBoxx USD High Yield Corporate 1 50 354 15,088
TLT Barclays 20+-Year Treasury 1 5 718 2,209
EMB iShares JP Morgan USD EM Bond ETF 4 89 67 3,462

aBasket bid–offer spread refers to the underlying securities of the respective indices.

The benefit of basically letting the people who create and redeem the ETFs buy the underlying securities instead of packaging them together is why more asset managers are using these within their funds. It is not just for any type of equitization; ETFs are their core holdings and enable them to get the scale of both the expense ratio from big players and the underlying trading costs that have been compressed because of the volume that is trading out there in the marketplace.

Conclusion

Over the years, the ETF market has become much larger, and it is continuing to grow rapidly. In the same way, the uses of ETFs for investors large and small are also growing. The popularity of ETFs and the range of their uses are spreading from the United States into other countries and from managed portfolios into other sectors. ETFs provide low-cost solutions to many asset management and investment challenges.

Question and Answer Session

Noel C. Archard, CFA

Question: There are growing concerns that an increasing number of these ETF products are not fully backed by the underlying assets. What is your view?

Archard: A product with physical underlying assets has no counterparty risk or questions on pricing. It is a cleaner way to build the portfolio, and the overwhelming majority of ETF products are actually built with physical underlying assets.

Derivatives used for asset replication are not villains. In some regulatory schemes, the use of derivatives actually provides a tax benefit. Generally speaking, the issue with derivatives-based products is a combination of counterparty risk and cost.

Question: As the ETF market grows bigger, are there any big systemic risks unique to this market?

Archard: I do not think so. In Europe, there was a growing use of swaps for some ETF exposures. The ETF sponsor was using an internal dealer to do the swap, and there was not a lot of transparency in the pricing. If the subdealer went under, then there was huge risk to the portfolio. The swap-based ETF structure got trampled in Europe, which helped spur regulations.

About 90% of the assets in US ETFs are built using the traditional Investment Company Act of 1940,2 the rules that basically govern every mutual fund. A few things are different for ETFs, but they are essentially mutual funds that trade on an exchange, which means they require a custodian and a board that has oversight of the fund and the assets are segregated.

Question: With the increase in ETF popularity, are markets more volatile as a result of ETFs’ ability to move markets and trade easily? If volatility increases, do the low costs of ETFs come at the expense of increased risk?

Archard: It still seems to come down to the bigger triggers of volatility—macroeconomic conditions, not the structures around them—because if people want to buy the securities, they will.

Question: We have seen competition drive down the expense ratios to very low levels, in some cases under 10 bps. Where do you see the bottom line?

Archard: You are bumping up now against institutional pricing in some cases in this 5–10 bp range. You might get different servicing. An institutional product is going to have different reporting requirements that are going out to its underlying users, whereas the ETF will always be a little bit more hands-off because of the lack of look-through to the underlying client. Once you get down into the single digits, there is not much lower you can go and still have a robust business program where you can deliver on the services for ETF users.

Question: Why do you think there has been a reversal of cost in using futures versus ETFs for cash management?

Archard: It basically just comes down to the financing. It is getting harder to do business. I mentioned that Dodd–Frank and Basel II and III change the treatment of capital on balance sheets. Dodd–Frank has a lot of reporting requirements embedded within it. All of these things were good and logical results of 2008–2009, but there has been a little bit more focus not so much on the products but on their participants and the regulation of participants. Ensuring stability among the participants drives costs up and makes different products unattractive. In some cases, it is knocking people out of the business, which also drives prices up.

Question: What are your thoughts on actively managed ETFs?

Archard: For every ETF that gets constructed, the sponsor has to feel like it is a good idea and has a portfolio fit, and the product should be attractive to a client from a results perspective, meaning it is going to actually do what we say it is going to do.

There is a third group out there—the market makers, the people who own the inventory on the exchange. They need to have a pretty good idea of where the ETF trades so that they can come up with a market price. If there is any kind of uncertainty in it—and by definition, active management and alpha create some uncertainty—then they have to go a little bit wider with their spreads. Maybe you get a good product, but it has a wider spread and a higher cost base. It might be unattractive to the clients.

So, active ETFs have a few more considerations involved. I think the growth in active ETFs is and will be slower than in the mainstream, passive products.

Question: Is there any evidence to suggest that the increased popularity of ETFs and the ease with which you can move in and out of them may actually lead to excessive trading and higher transaction costs?

Archard: Does it lead to excessive trading? That is probably a personal issue for every firm and every client out there. Does it make them easier to access? Yes, but again, I would argue that if people want to access markets, they can.

The benefit of the ETF is its popularity; it has actually compressed the spread significantly in the secondary markets, so there are cost savings there. Economies of scale have brought the expense ratio down, so there are other savings there. If you are disciplined about how you use them, they can be great. If you go out there and trade them a lot, they can be damaging.

1 Greenwich Associates, “Institutional Investors’ Relationship with ETFs Deepens” (2013).

2 Most ETFs are registered under the Investment Company Act of 1940—often referred to as the ’40 Act—as open-end investment companies. As such, they are subject to the specific provisions of the ’40 Act as they pertain to open-end funds. However, several of the customary features of an ETF are not consistent with the requirements of the ’40 Act. Thus, a key part of the process in launching any ETF is to receive the required exemptions from these provisions of the ’40 Act. See//www.cfa.com.cn/cfajiaocai/166.html


分享: