There has been much talk in the financial press about exchange-traded funds (ETFs), which in their simplest
form simply mirror a particular index. That could be anything from the S&P/TSX 60, or a gold index made up of gold
mining companies that trade on a specific stock market.
But before you decide to invest in an ETF you might want to ask yourself a couple of questions.
The first is: "Who will my co-investors be?" Different investors have different timelines for how long they
want to hold an investment. A long-term investor might not want to own a security that is primarily held by
short-term investors. On the other hand, short-term investors who like to trade in and out of positions can affect
the volatility of long-term investments.
The second question is: "Does the index meet my risk tolerance?" You should look into the assets held by
the ETF and understand their liquidity and investment characteristics. The mix of assets held in the ETF might
not be suitable to your personal investment objectives.
There has been major growth in the number of investors buying ETFs as markets have evolved. At the same time,
how investors access markets is also constantly changing.
Times have changed and so have markets
When I started in the investment industry in the 1970s, pink (sell) and blue (buy) tickets were walked to a wire
clerk who transmitted orders to a trading desk, which would then call the orders to the floor of the stock exchange
where a trader would try to find a counter party. You might hear in an hour or so what happened. The trades were
reported on a stock ticker that spewed paper onto the floor of the office. Commissions would be as high as 2.5% of
the trade’s value. It was slow and expensive compared to today’s instant fills from online brokerages trading at a
minimal flat fee.
The growth of global capital markets has also been spectacular. The market capitalization of the MSCI World Index
is almost $38 trillion, and that of the Barclays Global Aggregate Bond Index is over $43 trillion. Before the
financial crisis, the MSCI World Index was valued at $31.1 trillion and the Barclays Global Aggregate Bond Index
at $23.1 trillion (January 31, 2007). Investment styles are also evolving. The average holding period for a stock
position in the 1960s was eight years. By the turn of the millennium it had shrunk to one year, and it is around
half a year now (NYSE).
In most cases the ability to meaningfully change the risk profile of a portfolio requires a liquid, portfolio-wide
trading vehicle and for more traditional accounts, the ETF is the instrument of choice. The desire to be in or out
of the market shows up in ETF holding periods. One way to measure holding periods is to look at the number of days
it takes to turn over the float of a security. In the case of the iShares Russell 2000 ETF: 1.7 days over the past
11 years; for the SPDR S&P 500 ETF: 4.3 days. To put this in context, the top 10 S&P 500 companies turn over their
floats on average in 187 days: the quickest turnover is Apple at 40 days, and the slowest is Berkshire B at 300 days.
Markets melt up and down
When markets get really rough, investors who typically have very different time horizons and investment styles act
together at the same time, and in the same direction. Normal market liquidity disappears, and markets can "melt up"
or "melt down". A classic example of a meltdown is 2008/2009, when sophisticated margin players and hedge funds were
forced to liquidate because of margin calls. At the same time, retail investors, pension and sovereign wealth funds
were "re-risking" their portfolios to get out of so-called riskier investments, such as equity and investment-grade
and high-yield bonds to buy into the safety of government debt and cash equivalents. There were some very savvy buyers
in this "re-risking" process; they did not buy everything, rather, they focused on quality.
In response to "Modern Portfolio Theory" and the belief in efficient markets, index alternatives began to evolve from
traditional mutual funds that offered end-of-day liquidity to a new form of mutual fund that could be traded throughout
the day: ETFs. The first ETF was the TSX 35 ETF, founded in March 1990. It was followed in 1993 by the S&P 500 ETF (SPY),
which was replaced in 2000 by the TSX 60 ETF, the XIU.
The increase in ETF volume traded since 2000 has coincided with a massive increase in assets for the hedge fund industry
and the adoption of passive strategies by large pension funds. According to Barclay Hedge, a provider of hedge fund
industry analytics, industry assets under management increased from $118 billion in 1997 to a peak of $2.1 trillion
in 2007. This peak has since been surpassed: hedge fund assets totalled $2.7 trillion as of the second quarter of 2015.
There are currently approximately 860 million SPDR units outstanding, and during the year ended August 31, 2015, 813
million units were created and 825 million were redeemed - in essence, 100% turnover in units. In contrast, the annual
average redemption rate for the Canadian mutual fund industry is 15%, which implies a very different holding period.
How to look at market activity when contemplating ETFs
You can look at the market activity of ETFs in a couple of ways. One way is to look at short interest rates as a
percentage of units outstanding. In both 2008 and 2011, there was a sharp spike in short interest rates leading into
a subsequent market decline. Another, similar spike formed in 2015.
A second way has been reported in the press: ETF trading is increasing as a percentage of total trading on global
markets. A recent Barron’s article, The Great ETF Debacle Explained (Sept 7, 2015) reported that "...fully
42% of every dollar traded on U.S. exchanges on Aug. 24 was in ETFs, according to Credit Suisse." Given the nature
of ETF liquidity, with certain designated large investors delivering units or securities for creation or redemption
of units, there must have been a great deal of arbitrage desk activity that day. According to the share count, 20.7
million units of the SPDR S&P 500 Index ETF with a value of $3.9 billion were redeemed at the close on August 24.
That is a lot of capital being used on the arbitrage desks.
On Aug. 24, 2015, circuit breakers had kicked in on a number of index components, stopping trading for a period of
time in response to substantial drops in value. In doing so, it rendered useless the algorithms used by the
broker-dealer ETF market-makers. By early afternoon the high for the day was established and the abnormal spreads
between the valuations of ETFs vis-ŕ-vis their underlying indexes had largely been eliminated. But the spreads began
to widen again in the afternoon when the downtrend continued and circuit breakers went off and on. Assets were no
longer priced correctly in higher volatility markets such as the NASDAQ and, oddly enough, in the PowerShares
S&P 500 Low Volatility ETF.
Disorderly markets cause most discrepancies
Since the 2008 financial crisis, there have been four trading days when the S&P 500 Index has been down more than
5% in the middle of the day. The first of these was the "Flash Crash" on May 6, 2010. Again it is clear that when
the market is most disorderly - usually at the lows for the day - certain programs are unable to maintain a tight
spread between an ETF and its underlying index. The rest of the time the relationship seems very orderly. In August
2015, the problem was circuit breakers and possibly capital constraints. In May 2010 the problem was flash or automated
trading combined with margin calls.
The next occurrence was on August 8, 2011, during the U.S. debt downgrade and European banking crisis. While there
was a mild dislocation in the pricing of the SPDR S&P 500 and iShares Russell 2000 ETFs at the opening, the rest of
the time seems absolutely normal.
The fourth occurrence of a 5% intraday fall was on August 18, 2011. On that day, there were pricing discrepancies at
the opening of the market that quickly settled down.
What has changed since 2010 and 2011? Why would a relatively small margin event in August - a reduction of $20 billion
or -4.1% - contribute to such a severe dislocation in the ETF market? Even during the Flash Crash, ETF tracking was
much less disorderly than on August 24, 2015. In response to the Flash Crash, the U.S. Securities and Exchange
Commission approved a 2010 pilot program for single-stock circuit breakers. It was intended to deal with a large
number of individual securities experiencing abnormal volatility while volatility was not broad enough to trigger
circuit breakers throughout the entire market. In 2012 this program was revised to use a new mechanism. The first
trading day where multiple circuit breakers came into play was August 24, 2015. According to a Wall Street Journal
article, circuit breakers on stocks and ETFs were triggered more than 1,200 times in early trading on that day.
Tracking errors can be good news or bad news
Tracking errors have also played a part in how ETFs are valued - sometimes that’s positive for the investor, sometimes
negative depending on the issue. But the difference can be as much as 2%, and on days of exceptional volatility, can
exceed 5%. The tracking error is always most extreme at opening and at the market lows. It is less extreme at the
highs and at the close.
Looking at tracking errors at the opening of the market, it is fairly clear that they are persistently more pronounced
in "active" markets with higher volumes: 1997 to 2003 and 2007 to 2013.
In recent years, at the high for the day, the tracking error has been much higher around the time the high for the day
has been reached. But by the time markets are ready to close, much of the tracking error has been squeezed out. Indeed
since the financial crisis it appears that the only time when ETFs are tight to their benchmark is when approaching the
The history on some more specialized ETFs is too short to allow for a detailed analysis. When examining some of the
more specialized ETFs with liquidity constraints on the underlying securities, it is difficult to determine whether
ETF activity is driving price or if it’s normal investment activity.
When you look at the annual averages of tracking error you find a similar outcome to the SPDR S&P 500 Index ETF.
The spread has stabilized at around 25 basis points. This is the undisclosed cost of buying or selling an ETF.
I would assume that the spread behaviour at open and at the highs and lows for the day will be wider.
This article has looked at some of the quantitative aspects of ETFs through the lens of the largest and most liquid
ETF. It has looked at the holding periods for ETFs (extremely short) and how ETFs are used to quickly add or eliminate
market exposure. Investors assume they are receiving pricing that accurately tracks the benchmark the ETF is designed
to follow. It looks at price movements at four points in time when you can accurately measure what has happened to
price relative to what happened to the price movements for the benchmark the ETF was designed to replicate.
It finds that there is a structural spread in the ETF that is an undisclosed cost of ownership. In fast-moving markets,
this spread can be much more than the cost of ownership of a traditional, actively managed portfolio. Entering into
sell at open in ETF land is clearly dangerous, as is selling in rapidly falling markets, because the cost of liquidity
increases meaningfully. The only time you have a relatively tight market relative to the underlying benchmark is
going into the close.
This article has not looked qualitatively into what you own within the ETF and whether the underlying investments
are suitable to your personal investment policy statement. That analysis will form the second instalment in my
look at ETFs. In the meantime ask yourself this question: "Do I really need immediate liquidity for core investments
in my portfolio?" There is always a cost to immediate liquidity.
On final thought: Do you really need second by second liquidity for long-term investments?
Sandy McIntyre is
Vice Chairman and Chief Investment Officer of Sentry
Notes and disclaimer
© 2015 by Sentry Investments. All rights reserved.
Commissions, trailing commissions, management fees and expenses all may be associated with mutual fund
investments. Please read the prospectus before investing. Mutual funds are not guaranteed, their values change
frequently and past performance may not be repeated.
Certain statements in this document are forward-looking. Forward-looking statements ("FLS") are statements that
are predictive in nature, depend upon or refer to future events or conditions, or that include words such as
"may," "will," "should," "could," "expect," "anticipate," "intend," "plan," "believe," or "estimate," or other
similar expressions. Statements that look forward in time or include anything other than historical information
are subject to risks and uncertainties, and actual results, actions or events could differ materially from those
set forth in the FLS. FLS are not guarantees of future performance and are by their nature based on numerous
assumptions. Although the FLS contained herein are based upon what Sentry Investments and the portfolio manager
believe to be reasonable assumptions, neither Sentry Investments nor the portfolio manager can assure that actual
results will be consistent with these FLS. The reader is cautioned to consider the FLS carefully and not to place
undue reliance on FLS. Unless required by applicable law, it is not undertaken, and specifically disclaimed that
there is any intention or obligation to update or revise FLS, whether as a result of new information, future events