Escolar Documentos
Profissional Documentos
Cultura Documentos
The heart of the trading system is our proprietary indicator, the Yates VXX
Indicator (YVI). It provides the entry and exit signals for the VXX Trading
System. Other indicators have also been developed by Len to give you more
information on the current conditions for VXX and XIV.
When trading this system you are either long the VXX, Long the XIV or in
cash (no position). It is that simple and very effective.
Len has built the proprietary indicators used in the VXX Trading System into
the OptionVue 8 software to enable OptionVue clients to easily trade the VXX
system on their own. However, the OptionVue 8 software is not required to
trade this system. Subscribers can see the information right on the VXX
Trading System website and you will receive emails for every trade alert.
1
www.vxxtradingsystem.com
OptionVue Systems International, Inc,
28100 N Ashley Blvd, Suite 102
Libertyville, IL 60048
Phone: 847-816-6610
Table of Contents
Introduction Page 3
2
Introduction
OptionVue has offered the VXX Trading system since April 2012. As we have
traded it and gained more experience, we have adjusted the rules to better
perform in all types of markets.
This eBook goes over the research and trading rules we have used and how
they have evolved over time. First is an article by Len Yates that details the
original system and the ideas and concepts behind it. The VXX Trading
System gained 102% over the first four years.
After getting significantly lower returns in 2014 and 2015 we began to look
at ways the system might be improved. The article A New Beginning
outlines the current trading rules we use and introduces our new indicator,
the Yates VXX Indicator ($YVI). Since these new trading rules were
introduced, the VXX Trading System has returned 100% over the period
March 6, 2016 through October 6, 2016.
Disclaimer
Volatility Products involve risk and are not suitable for all investors. In
addition, electronic trading poses unique risk to investors. System response
and access times may vary due to market conditions, system performance
and other factors. You should thoroughly research and understand any
security before investing in it. OptionVue provides neither investment nor
tax advice.
3
Trading the VXX: From the Beginning
By Len Yates. First Published in April 2012
In the Feb 23, 2012 issue of The Option Strategist, Larry McMillan
wrote about a system for trading the VXX and+ XIV which seemed
remarkable in its simplicity and its high rate of return a tenfold
increase in 5.5 years time. This only involved 14 trades, so it
doesnt even require you to become an active trader. Intrigued by
this trading approach, I decided to develop some tools and
information to assist in my own trading.
Lets cover the basics. This is not an option trading strategy per se;
but rather a strategy for trading two of Barclays Volatility ETNs the
VXX and the XIV. Note that the XIV is a mirror image of the VXX, so
the XIV is useful when you would like to be short the VXX but are not
allowed to short stocks in your account. You simply buy the XIV when
you want to be short the VXX.
What are these two ETNs and what is the advantage to trading them?
Quoting McMillan, Both VXX and XIV are constructed in such a way
that they actually own (or are short) the two $VIX nearest-month
futures. Each day, part of the position is rolled forward [so that a
constant 30-day forward position is simulated]. When the term
structure slopes upward (which it does most of the time), that daily
roll is a debit, and is a drag on the performance of VXX vis--vis
$VIX. However, since XIV is the opposite, it is a boon to XIV.
Conversely, when the term structure of the $VIX futures slopes
downward, the daily roll forward for VXX is a credit, allowing it to out-
perform $VIX (and consequently is a detriment to XIV).
4
Please Note: In the following pages you will read a lot about the $VXDIF,
which used to be our main indicator giving us the signals to be in the VXX
or XIV. The $VXDIF is still available in our system and updating in real time
but our main indicator is now the $YVI, discussed beginning on page 14
5
Figure 1: Historic Chart of VXX with $VXDIF 15-Day Moving Average in Lower Chart
This new indicator (the solid blue line) simply displays a moving
average of the value of the $VXDIF itself, except that the built-in bias
of 10 is removed so as to display values that swing above and below
zero. We initially included the Lentz Volatility indicator (LVI, an
indicator of market volatility). The details of LVI are unimportant
since we later replaced this with a new proprietary indicator, the
Yates VXX Danger Indicator ($YVD).
Browse back in time through an $SPX price chart and it will not take
long to notice that, in general, when the 15 day moving average of
$VXDIF is positive, the stock market is moving up and it is ok to be
short the VXX (or long the XIV). When the moving average of $VXDIF
is negative, the stock market is experiencing a rough time and it is ok
to be long the VXX. The trends in this period have been long lasting
and effective for this trading approach.
6
been:
The result was an amazing 76.68 times your money in the 3.8 year
period. Note that the final position was still on when I did this initial
research (11/13/12).
Note: The XIV hasnt been around long enough to be tradable during
this entire period, so for the sake of this study I augmented the
historical data in the XIV by assuming mirror image percentage daily
moves vis--vis the VXX, going back to the first day of the VXXs
existence. So please note that data prior to Nov 30, 2010 is
theoretical, while data from Nov 30, 2010 on is actual trade data.
Finally, the first trade was put on right when the VXX became
available for trading and the $VXDIF 15 was negative at that time. It
could be argued that this trade does not belong in the study because
it was not triggered by $VXDIF 15 crossing the zero line. However, it
was just a small loss and therefore it doesnt matter very much.
7
Figure 2: Historic Chart of VXX in 2010 with the LVI15 in Lower Chart
We are focusing on the period from May 1st until July 15th, in the
chart above. This was the flash crash and subsequent rough period
lasting 3-5 months. We would have liked to see the $VXDIF 15 line
(the solid blue line) swing more firmly into negative territory during
this period. Instead, it waffled close to, and mostly above, the zero
line. Following the system (as we did in the study) switched us out of
short VXX and into long VXX way too late, and switched us back into
being short the VXX earlier than it really should have.
That is why I looked for something else that could help us. We want
to be tipped off to the beginning of a rough period, even if that
beginning is as sudden as the flash crash. That is when I found the
Lentz Volatility indicator on a 15-day MA (denoted LVI 15 for short)
seemed to help by letting us include a measure of increasing or
decreasing volatility of the $SPX.
Incorporating the LVI 15, we came up with revised entry and exit
rules: Go long the XIV near the end of a rough period when the
8
$VXDIF 15 crosses from negative into positive territory and LVI 15 is
in positive territory (indicating that volatility seems to be coming
down). However, refrain from taking this entry any sooner than 2
months into the rough period. This rule was to prevent us from
entering prematurely, and helps deal with the kind of situation like
what happened in the rough period following the flash crash, where
our signals would have gotten us long the XIV sooner than it was safe
to do so. Stay long the XIV as long as $VXDIF 15 is positive.
Figure 3: Historic Chart of $SPX in 2010 with the LVI15 in Lower Chart
9
Note that the stronger the $VXDIF 15 is, the more confident we can
be in continuing to hold our long XIV position. Numbers higher than
+1.50 are strong and we must resist the temptation to jump out of
our position prematurely. The only exception would be, as mentioned,
the onset of volatility after a sustained market advance.
We will be long the XIV more of the time than being long the VXX,
and being long the XIV is where most of our gains will come from. We
must see our position through the smaller market corrections.
You see, the period of time when you want to be long the VXX is very
different in nature from the period of time when you are long the XIV.
The VXX typically shoots up quickly, reaches a peak and then
seesaws downward from there to the point where we get the signal to
close it. So our signal never seems to get us out at an optimal time.
For this reason it might be better to use the dilated pupils test.
When you open the quotes display and see that your VXX position has
gone very far very fast and for a moment your pupils dilate with
greed, then sell.
You certainly know when youre in one, but how can you recognize
when one is beginning? Well, a negative $VXDIF 15 tells you for sure.
10
Other than that, as I said, it is good to keep an eye on the LVI 15
after a sustained advance, watching for signs of increasing volatility.
The flash crash rough period, which was not well signaled by a
negative $VXDIF 15, was certainly prefaced by visible tremors in
terms of price action.
Long/Short Enter Enter Exit Date Exit Days Trade Percent Cum
Date Price Price Held Profit NAV
1.00
Long VXX 1/30/09 418.32 4/13/09 401.48 73 -16.84 -4.0% 0.96
Long XIV 4/13/09 0.72 5/4/10 4.49 386 +3.77 +523.6% 5.99
Long VXX 5/4/10 89.48 6/01/10 120.76 28 +31.28 +35.0% 8.08
Long XIV 7/13/10 4.67 8/5/11 11.71 388 +7.04 +150.7% 20.26
Long VXX 8/5/11 30.31 11/25/11 49.20 112 +18.89 +62.3% 32.88
Long XIV 11/25/11 4.91 (open) 16.65 354 +11.74 +239.1% 111.5
Two Footnotes:
1) In this study, we dealt with the flash crash and subsequent rough
period by assuming we saw trouble coming in late April 2010 and got
out at the end of the first bad day going into the crash, May 4, 2010.
It might have been possible to do better than this but we could have
done worse as well. Then, when the rough period seemed to be over I
picked the time to get short the VXX as being after the prescribed 2
months had passed and when the indicators looked strong. This was
7/13/10, which turned out not to be an optimal time to buy the XIV
because it went down for a few days after that, prior to turning up.
The VXX and XIV have a high degree of volatility, and as a result I
would not necessarily think of leveraging them up using options since
they are already, in a sense, leveraged up. (Note that options are
11
available on the VXX but not the XIV.) I suggest that we be content
to simply trade the ETNs. This volatility should also be considered as
you think of allocating some of your funds to this trading approach.
Just to give you an idea, the XIV recently went from 8.30 to 13.37
and back down to 11.00 in just 3 weeks.
A special note if you are going to try back testing this strategy. There
was a 1-for-4 reverse split in VXX on November 9, 2010. You would
need to adjust the size of your position accordingly when you
advance to that date.
Some might wonder if trailing stops would help this strategy. I dont
tend to think so, but you are welcome to test this on your own. If you
find something promising, please let me know! My sense is that the
use of trailing stops would only make returns worse. It also raises the
question of when it is ok to re-enter a position in the same direction.
The $VXXDRE and $XIVDRE indexes, like the $VXDIF index, have a
neutral bias value of 10 and Prev that is always 10, so you can simply
focus your attention on the Change value.
We also created the $VXXFV and $XIVFV indicators for the VXX and
VIV respectively. The FV stands for fair value and you can see what
the current fair value of the index is vs. what the market price is.
12
Warning: The VXX and XIV are volatile
Be prepared for some volatility. For example, after our clients originally
bought in at around 10.7, the XIV dipped down to 8.1 just 5 weeks later (It
recovered fairly quickly after that). Still, the use of stops is not
recommended. We do not recommend that all of your investment capital be
used in the VXX Trading System, but only a portion.
In contrast, no adjustments are needed when you go long the XIV. However,
the XIV, as we have discussed before, has its own drawback. Every time the
$VIX has a spike of any appreciable size (e.g. up 2 points and then back
down), the XIV will not return to the price level it was before. It comes up
just a little short. This places a drag on the XIV that amounts to a daily roll
yield of approximately -0.5 over time. As a result, when you see that the
XIV has a roll yield of +0.5, in practice you could very accurately interpret
this as zero roll yield.
It is not much of a stretch to imagine that this drag on the XIV and the costs
of adjusting the VXX position size (especially if adjusting frequently) are
counterparts of the same underlying thing, as both are triggered by frequent
(and significant enough) price moves in the foundational $VIX.
So both approaches have their drawbacks, but to me, being short the VXX
has more. There is the cost of active management, and there is also the risk
of total ruin in the event of a catastrophe that causes the $VIX to soar
beyond belief. If you are short the VXX you could owe additional money, but
if you are long the XIV, your capital simply goes down to nearly zero.
13
Trading the VXX: A New Beginning
After trading this system for four years, we thought it was a good time to
look back and assess the results of the VXX Trading System so far and
evaluate the system entry and exit rules to see whether they could be
improved. After researching this, we found they could be. More on this later.
First the assessment:
The VXX Trading System, if strictly followed as originally laid out, generated
a 102% gain during the period from April 2012, when the system was first
published through March 2016. This gain, which amounts to 25.5% annually,
is disappointing. The past two years (2014 and 2015) in particular were not
good and stand in stark contrast to the first two years. Market conditions
were different, as we have seen the bull market of 2009-2014 grow tired,
giving way to a long period of choppy, sideways action.
Knowing that the engine driving this whole approach is roll yield, it made
sense to question the validity of always being in a position. There are times
when the roll yield is small. Might it be better to be out of the market at
such times? We did some back-testing and found that, indeed it does seem
to be better to be long the XIV only when roll yield is significantly favorable.
We set out to create a new indicator for the VXX Trading System. The result
was the Yates VXX Indicator ($YVI), a proprietary indicator.
The new trading rules for the VXX Trading System are:
14
We have traded these rules since March 6, 2016 and through 10/6/2016 we
received a positive return of 100%! Double our money in just seven months!
The chart below shows the S&P 500 Index ($SPX) from 2016 with the new
$YVI indicator in the lower chart. The dotted green line corresponds to a
reading of 0.83. When the $YVI is above this line we are long the XIV (Short
VXX). The dotted red line corresponds to a reading of -0.60. When the $YVI
is below this line we are long the VXX. When the $YVI is between these
lines, we are in cash (no position).
15
You will notice the magenta bars through the lower chart. This is the Yates
VXX Danger Indicator ($YVD). It was developed to alert us when unusual
volatility moves are happening in the market, giving us a heads up that we
may want to close our position, or at least scale it back. You can learn all
about this indicator in a video recording on YouTube at:
www.youtube.com/watch?v=oOT347mjZzw.
We see these changes as being very beneficial and are encouraged about
the system going forward. Trying to second guess the system or introduce
other indicators can be detrimental. We need to simply let roll yield do the
job for us, and we can do so by following these new parameters.
16
Volatility Cones and Term Structure
There has been much discussion lately about volatility ETNs and ETFs and
systems for trading them. In particular, I have written about trading the VXX
and XIV and launched a VXX Trading Service built right into OptionVue. The
success of such trading systems is based to a large degree on the daily roll
yield - the day-to-day gain or loss from closing some of the fund's position in
the nearby VX futures and opening a corresponding position in the next out
VX futures. When a price difference exists between these two contracts, the
VXX and XIV (and all other ETNs like them) realize a gain/loss each day they
roll a portion of their position.
Price differences among the various VX futures contracts reflect what we call
the term structure of volatility. When farther out contracts have higher
prices than nearby contracts, it is what many
call "normal" term structure or "contango". This is the case whenever the
market is relatively calm and in an uptrend. The opposite situation occurs
when the market is extremely volatile. During such times the VX futures
exhibit a "backward" term structure, in which the nearby contracts have
higher prices than the farther out contracts.
When the VX futures term structure is "normal", the VXX is realizing a loss
from each day's roll while the XIV is realizing a gain from each day's roll.
When the term structure is "backward", the VXX is realizing a gain from each
day's roll while the XIV is realizing a loss from each day's roll. There is no
daily roll yield when the VX term structure is flat.
We should keep in mind that for VXX/XIV trading to succeed going forward,
we need the VX term structure behavior that has been observed in recent
years to persist. In order to get an idea whether this should happen, it would
be helpful to gain some understanding of why the VX futures' term structure
behaves the way it does.
Why do the VX futures' term structures behave the way they do?
17
over different time horizons such as 30, 60, 90, 120 days. The concept of a
volatility cone, as initially proposed by Burghardt and Lane, is based on
empirical data and depicts the mean reversion tendency of volatility. It puts
the current implied volatility into perspective. The chart below shows the
volatility cone for the SPX. While it is an older chart, it is an excellent
example for our purpose.
In the graph, the horizontal axis is time and the vertical axis is volatility.
What this shows is that an extremely high volatility reading today will, as
demonstrated historically, gradually make its way down towards average. By
the same token an extremely low volatility reading today will gradually make
its way higher. That volatility mean-reverts is practically an axiom, and this
mean-reversion tendency is exactly why the VX futures exhibit contango
when current volatility is low and backwardation when current volatility is
high.
Looking at the chart again, you can even imagine that either of the black
lines containing triangular or diamond shaped plot points represents the VX
futures prices at various durations; hence two possible term structure
scenarios. Conceptually this is entirely correct.
Since mean-reversion and the volatility cone are foundational, and the
cone's predictions relate directly to the term structure of volatility, we can be
18
confident that the VX futures' term structure behavior will continue.
Note that the term structure is unusually steep these days, as noted in
recent articles by Larry McMillan, suggesting that there exists an unusually
high demand for portfolio protection by means of buying the VX futures (or
indirectly buying them by buying the VXX or similar ETN). In the graph
above, it is perhaps like the solid blue line.
19