ETFs and technical analysis, used separately, are useful tools
for investors. And their virtues compound when technicals are
specifically used for trading ETFs. That said, sometimes there's a
fly in the ointment, as a closer look at State Street Global
Advisor's SPDR S&P Retail ETF (NYSEArca:XRT) shows.
First, let's look at the percentage change chart below, that begins
with XRT's inception in June 2006. Everything seems normal when
compared to the S&P 500. There's an obvious high correlation in
movements between the two, as we would expect. It's also clear that
from a price base only (no dividends), XRT has had strong relative
strength outperforming the S&P 500 by approximately 13 percent
over this time span.
Our next chart is daily, and covers a period of about one year,
showing classic backward-looking technical analysis. Here we can
see that the price trended upward (A-B) in a relatively uniform
channel and then based (or ranged), as we normally see, in a
prolonged upward movement from points (B-C).
With no pronounced selling or major bearish technical breaches
in this basing area, the price resumed its upward climb (C-D), the
same distance as the previous run, as measured from points (A-B).
From that high, the trend breaks (
), falling back to a previous resistance area at the $37 price
level, where it is trying to stabilize in forming what appears to
be a triangle formation.
From a pure price analysis we seem to be in an area of
indecision, so standard procedure dictates that we look to other
information for further clues to predict future probable price
In this case, we'll look for clues in volume data, which is
displayed at the bottom of our next chart. What immediately stands
out is what appears to be a
Bullish Volume Pattern
between the price and volume (magenta lines). One of the rules of
technical analysis in Dow theory is that "volume should expand in
the direction of the trend," which means that volume should support
the current momentum and movement of the established trend.
On this chart, prices have clearly fallen since May until now,
volume is not expanding, but rather has fallen. In technical
analysis, that would be interpreted as a bullish signal insofar as
the declining rate of daily "selling transactions" as time
progresses, suggesting that the bearishness is diminishing in
So what's the catch? Where is that potential pitfall that I
alluded to above that sometimes makes combining the value of ETFs
and technical analysis a bit problematic?
Data compiled by IndexUniverse.com might just provide the
answers we're looking for. From the period of the end of May to the
end of July, XRT's total assets fell $747 million-from $1.2 billion
to just $453 million-an incredible 60 percent. There could be many
reasons for the drop, from falling prices (less than 11 percent in
this example), to sector rotation of funds, to investment dollars
switching to similar ETFs or other investments with lower expenses
or other perceived benefits.
A more likely scenario in this case might be that of a simple
"pair's trade" that is reducing the total number of outstanding
shares. This might involve a large institution shorting XRT, and
going long a volatility-adjusted amount of another retail-focused
investment that, relatively speaking, has underperformed XRT over a
specific period of time. From my technician's point of view, the
reasons behind XRT's drop in assets can be somewhat irrelevant.
What is relevant are any changes in the number outstanding shares,
and the effect that this can have on any technical
In our case, this bullish volume pattern-the reduction in volume
with falling prices that appeared on our charts-is an illusion. The
falling volume that we initially interpreted as a reduction of the
selling intensity that had bullish implications is much more likely
due to the fact that far fewer shares of the ETF are now being
traded than in the month of May.
In fact, some quick statistics support this phenomenon, in that
from January to April of this year, the average daily volume for
XRT was 24.6 million, and in July, the average daily volume dropped
to only 13.3 million shares, a 46 percent reduction.
Obviously, changes in volume to a stable number of shares being
traded are classically noteworthy, but changes in volume compared
to a changing number of outstanding shares are potentially
misleading. Unfortunately for the technician, this is camouflaged
in the charts as the transient number of outstanding shares is not
visible, and additional research for proper and complete analysis
The implications here are clear:It's important for the ETF
investor using technicals to be aware that the number of ETF shares
is changing every day due to the creation and redemption process
that is central to how exchange-traded funds work. The key is to
realize that when the number of shares outstanding of an ETF
changes dramatically, especially over a short period of time, it
can greatly distort classic technicals-based volume analysis
techniques and thus lead one astray.
Have a profitable week and remember always, "Keep it in
All data and information provided in this column are for
informational purposes only, and should not be regarded as
recommendations to buy or sell securities.
All charts created with TradeStation. ©TradeStation
Technologies, Inc. All rights reserved.
Ray Rondeau is president of the Boston Chapter of The American
Association of Individual Investors
When he's not trading using technicals, he presents to various
groups on technicals and trading in New England and beyond. He can
be contacted for presentation information at
Don't forget to check IndexUniverse.com's ETF Data
2010 Index Publications LLC
. All Rights Reserved.