VIX Strategy &

Analysis Collection
 Contents

3 
Tracking VIX swings 27 
The VIX and market capitulations
The VIX has been a widely discussed stock It takes more than a VIX spike to identify
market barometer, but how reliably does it identify a market bottom.
market turning points? This study turned up a few By Howard L. Simons, Active Trader, May 2009
surprises in analyzing how the S&P 500 tracking
stock responded to VIX highs and lows. 32 
Fair value and the VIX futures
By David Bukey, Active Trader, January 2006 The relationship between the VIX futures and
the underlying index can shed light on the
9 
VIX-based system market’s expectations for future volatility and
Futures Trading System Lab. market direction.
By José Cruset, Active Trader, January 2006 By Russell Rhoads, Active Trader, December 2011

12 
The VIX fix 36 
Trading with the VIX Put/Call ratio
A “synthetic” VIX calculation can be used in There are several put/call ratios, and they all
any market to reproduce the performance have different characteristics. Find out how the
of the well-known volatility index. VIX Put/Call Ratio can signal trade opportunities
By Larry Williams, Active Trader, December 2007 in the stock market.

By Russell Rhoads, Active Trader, January 2012

19 
VIX options
Despite appearances, VIX options behave
40 
The VIX-neutral option trade
differently from other options.
When trading VIX options, it’s the VIX
By Marc Allaire, Futures & Options Trader, futures price that counts.
December 2007
By Russell Rhoads, Active Trader, August 2012

23 
Hedging with VIX options 45 
The VIX basis trade
The VIX’s unique characteristics give its
The unique nature of VIX futures offers a simple
options certain advantages in constructing a
but significant trade opportunity.
stock-market hedge.
By Euan Sinclair, Active Trader, January 2013
By Steve Papale, Active Trader, July 2008

VIX Strategy and Analysis Collection

 MARKET Pulse

Tracking VIX swings

M
The VIX has been a widely discussed stock market because they consider it an important
any traders follow the
Chicago Board Options
Exchange’s

market sentiment indicator — a sup-

barometer, but how reliably does it identify market posed “fear gauge” that measures the
(CBOE)
volatility index (VIX)

market’s overall anxiety or complacen-

turning points? This study turned up a few surprises in cy, extremes of which are generally con-
sidered bullish and bearish signals,
analyzing how the S&P 500 tracking stock (SPY) respectively.
The VIX measures the S&P 500’s
responded to VIX highs and lows. implied volatility using the price of its
near-term options — the market’s cur-
rent estimate of expected future volatili-
ty, reflected in option prices. In general,
BY DAVID BUKEY the index tends to rise as the S&P 500
falls, and drop as the market rallies. (For
more information
about the VIX and its
FIGURE 1 RISING VOLATILITY — 1998 calculation, see “The
volatility index”).
SPY tends to drop as the VIX rises, but SPY’s short-term response to these extremes is less Over longer-term
straightforward. The market rebounded the day after VIX highs, although this strength didn’t periods, this inverse
last. However, 60- and 120-day VIX highs pinpointed better entry points than its shorter-term relationship is hard to
extremes. miss. For example, the
VIX rose from a multi-
year low (16.23) to an
all-time high (49.53)
between mid-July and
early October 1998 as
the S&P 500 fell 17.1
percent during this
period (see Figure 1).
Both instruments then
retraced their moves
within six weeks.
Similarly, as the S&P
500 began to rally in
March 2003, the VIX
slid to yearly lows and
continued to drop
throughout the year.
However, there’s lit-
tle proof that buying
the market as the VIX
drops and selling short
when it spikes is
always profitable. The
following study takes
Source: MetaStock
a different approach: It

www.activetradermag.com • January 2006 • ACTIVE TRADER

 The volatility index
(VIX)

analyzes large VIX moves (up and

down) that also exceed recent highs or
lows and measures how the S&P 500
just three times a year, on average.
Large jumps in volatility were more
common than big drops. For instance,
S tatistical and implied volatili-
ty are two ways to measure a
security’s degree of price
movement. While statistical (“his-
torical”) volatility tracks actual
tracking stock (SPY) responded in the 10 we identified sixty-seven 40-day VIX daily price fluctuations, many
days after these volatility changes. highs, but no same-length lows met our traders focus on implied volatility,
While some type of inverse relationship criteria. Also, there were twenty-eight which estimates future moves based
between the market and the VIX certain- 120-day VIX highs, but no matching on the prices of its options.
ly exists, we also found several unex- lows. The VIX measures the implied
pected short-term market patterns fol- Figure 1 shows a daily chart of the volatility of S&P 500 index options
lowing extreme volatility levels. VIX (top) and SPY (bottom) from May traded on the Chicago Board Option
1998 to February 1999. The market Exchange (CBOE). The index reflects
Tracking large volatility swings tanked as the VIX rose in July, but SPY the market expectation of near-
We measured VIX moves over several bounced off lows when volatility term (i.e., 30-day) volatility. The
look-back periods (10, 20, 40, 60, and 120 reached an all-time high on Oct. 8. VIX has been around since 1990, but
days) to identify significant volatility Figure 1 shows shorter-term VIX underwent a major transformation
changes. We then ranked them against a highs leading to further market sell-offs, in late 2003. It is a commonly refer-
certain number of previous moves of the but longer-term (60- and 120-day) highs enced gauge of the stock market’s
same length and direction. For example, preceding market rallies. The market “fear level.”
we compared each 10-day move to the tended to sink as the VIX hit new highs, When the CBOE overhauled the
past twenty 10-day moves. Each 20-day but SPY rebounded from these declines VIX in September 2003, it changed it
move was also ranked against the prior the next day. from a volatility measurement
forty 20-day moves, and 40-, 60-, and Studying the daily relationship based on the S&P 100 (OEX) to one
120-day moves were compared to the between the VIX and the market showed based on the S&P 500 (SPX).
past 80, 120, and 240 same-length SPY tended to fall as the VIX reached its The old VIX formula used the
moves, respectively. highs over the past 12 years, but it Black-Scholes pricing model that
This process ranked each move on a regained ground the next day. However, looked at eight near-term at-the-
scale of 0 to 100, which shows its size no clear pattern appeared during the money OEX options (calls and puts).
relative to past moves. A percentile rank nine days following these events. The new VIX is derived from the
of 100 means the move is larger than all Figure 2 highlights the oppo- prices of options themselves — near-
previous moves, while a value of zero site scenario surrounding the market term at-the-money SPX options as
indicates it’s smaller than prior ones. We recovery of 2003. At first glance, VIX and well as out-of-the-money puts and
then focused on VIX moves with a per- SPY seem to mirror each other. Volatility calls — so the index reflects the full
centile rank of at least 95. surged as the market dropped in mid- range of volatility.
Finally, for each move, the VIX itself January, and the VIX peaked as the S&P The CBOE also applied the new
also had to exceed its recent highs or 500 reached a four-month low in mid- calculation method to the CBOE
lows. For example, a 10-day increase February. Also, larger VIX decreases Nasdaq Volatility Index (VXN), which
with a percentile rank of at least 95 must didn’t appear until after the market bot- reflects the volatility of the Nasdaq
also exceed the highs of the past 10 days. tomed on March 12. 100 index. The exchange still pub-
Similarly, a 60-day low with the same While the market’s shorter-term pat- lishes the original VIX calculation,
percentile ranking must also drop below terns are hard to discern in Figure 2, SPY which can be found under the ticker
the lows of the past 60 days. tended to climb the day the VIX hit its symbol VXO.
lows as well as the day after these Our study uses historical VIX val-
Overall patterns events. But the market’s behavior was ues based on the new CBOE’s formu-
The analysis measured SPY’s reaction to mixed throughout the remaining nine la because they correspond to the
these extreme VIX changes, which were days of our analysis window. S&P 500 more accurately than the
relatively rare, from Jan. 29, 1993 to Oct. original one.
10, 2005. Although 10-day volatility highs VIX highs lead to brief market For more information about the
and lows appeared roughly once a rallies VIX and its calculation, visit
month, longer-term (20-, 40-, and 60-day) The S&P 500’s tendency to climb on the www.cboe.com/vix.
VIX extremes occurred only about six day following large VIX declines makes
times a year, and 120-day moves emerged

ACTIVE TRADER • January 2006 • www.activetradermag.com

FIGURE 2 DROPPING VOLATILITY — 2003
SPY rallied significantly as the VIX fell to its lows in spring 2003. Although this behavior
implies an inverse relationship between volatility and the market, SPY actually tended to sense, but the market’s
underperform in the first two weeks after VIX drops (see Figure 5). one-day rally after big
VIX increases seems
counterintuitive.
However, SPY dropped
S&P 500 Depositary Receipts (SPY), daily
at least 0.77 percent, on
average, as volatility
climbed to highs on all
time frames (e.g., 10, 20,
40, 60, and 120 days).
This implies the market
bounced back after sharp
sell-offs, a pattern that
has appeared in prior
Market Pulse articles (see
“Related reading,” p. 8).
Figure 3 shows SPY’s
average gain on the first
day after large, 95th-per-
centile VIX increases of
various lengths also hit
same-length highs. The
further the VIX soared,
the stronger the market’s
rally the next day.

Source: MetaStock
Rising volatility:
60- and 120-day
VIX highs
FIGURE 3 SPY RESPONSE TO VIX HIGHS Figure 4 shows SPY’s average daily
gains and losses in the 10 days following
The market jumped the day after volatility hit new highs, a tendency that 60- and 120-day VIX highs. The figure
intensified as the VIX hit longer-term extremes. But SPY fell at least 0.77 also shows the market’s cumulative
percent, on average, as the VIX reached these respective highs (not shown), move from the close the day the VIX hit
so it’s likely this pattern merely represents a short-term rebound. a 120-day high to each subsequent day’s
close. This cumulative behavior is then
compared to its benchmark, or typical
same-length move, since 1993.
If you compare SPY’s response to 60-
and 120-day VIX highs, you’ll notice the
market rose much more following 120-
day extremes than after the VIX’s 60-day
moves. For example, SPY climbed an
average 1.14 percent the first day after
the VIX rose sharply to a 120-day high
before rallying another 0.69 and 0.35 per-
cent on the second and third days. In
contrast, the market posted an average
0.70-percent gain following the VIX’s
steep rise to 60-day highs, but it gave up
0.22 percent in the subsequent two days.
Although the market’s initial rally is
impressive, volatility’s effect is clearly
short-lived. SPY either traded sideways
or drifted lower during six of the remain-

www.activetradermag.com • January 2006 • ACTIVE TRADER

 FIGURE 4 60- AND 120-DAY VIX HIGHS
While SPY surged an average 2.18 percent in the first three days after the VIX
rose strongly to a 120-day high, this bullishness didn’t last, and the market
fell slightly during most of the second week after this event.
ing seven days. (Although Figure 4 does-
n’t show how the market reacted to 10-,
20-, and 40-day VIX highs, those
responses are nearly identical to the lack-
luster behavior after 60-day VIX highs.)
Table 1 lists the daily statistics for 120-
day VIX highs and proves that Figure 4’s
performance is fairly reliable. The aver-
age, median, maximum, minimum, and
standard deviations are shown for each
day after these events. The largest intra-
day up and down moves (LUM and
LDM), benchmark price moves, and per-
centage of gains (“%>0”) are also shown.
All 10 of Table 1’s average values
point in the same direction as their
medians, and the majority have only
small differences between both values.
For example, the market’s 2.18-percent
average gain from days 1 to 3 is only
slightly below its median 2.45-percent
jump during that same period. (For a
more detailed explanation of average
and median, see “Key Concepts and
Definitions,” p. 89.) er than the LDMs, or close-to-low percent on the day the VIX slid sharply
The market also had a greater percent- moves. In contrast, LDMs are bigger and reached either 10-, 20-, or 60-day
age of gains on the first two days — than their corresponding LUMs during VIX lows. But instead of retracing those
71.43 percent and 67.86 percent, respec- seven of the other eight days. moves the next day — as SPY did fol-
tively — than on the remaining eight lowing VIX highs, the market tended to
days. Another sign of a bullish market Dropping volatility: rally again on the first day after falling to
on days 1 and 2 is the average LUMs, or 60-day VIX lows VIX lows. However, SPY’s overall reac-
close-to-high moves, are noticeably larg- The market rose at least an average 0.52

TABLE 1 120-DAY VIX HIGH STATS

The market rallied the first three days after the VIX rose to 120-day highs, but patterns are less clear over the remaining eight days.

28 instances Day 1 LUM LDM Day 2 LUM LDM Day 3 LUM LDM Day 4 LUM LDM Day 5 LUM LDM
Avg: 1.14% 2.27% -1.58% 0.69% 1.79% -1.37% 0.35% 1.24% -1.30% 0.03% 1.09% -1.21% 0.12% 1.32% -1.64%
Med: 1.22% 1.84% -1.24% 0.99% 1.47% -0.75% 0.24% 0.79% -1.00% 0.25% 0.63% -0.87% 0.05% 0.61% -1.17%
Max: 5.97% 6.52% 0.00% 5.97% 6.47% 0.00% 4.87% 5.54% 0.00% 4.87% 5.54% 0.00% 5.38% 6.65% 0.00%
Min: -7.13% 0.00% -8.10% -7.13% 0.00% -8.10% -3.24% 0.00% -4.31% -3.52% 0.00% -3.99% -3.24% 0.00% -5.52%
Standard deviation: -2.69% 1.65% 1.75% 2.31% 1.41% 1.79% 1.76% 1.33% 1.17% 1.74% 1.29% 1.08% 2.28% 1.80% 1.55%
Pct. > 0: 71.43% 67.86% 50.00% 53.57% 50.00%

Day 6 LUM LDM Day 7 LUM LDM Day 8 LUM LDM Day 9 LUM LDM Day 10 LUM LDM
Avg: -0.25% 1.06% -1.60% -0.08% 1.26% -1.59% -0.35% 1.06% -1.75% -0.14% 1.30% -1.51% 0.68% 1.66% -1.13%
Med: -0.03% 0.77% -1.21% -0.17% 0.75% -1.11% -0.47% 0.47% -1.50% -0.27% 0.78% -1.03% 0.39% 0.99% -0.96%
Max: 5.37% 5.37% 0.00% 5.97% 6.47% 0.00% 4.80% 5.58% 0.00% 5.97% 6.47% 0.00% 5.97% 6.47% 0.00%
Min: -3.52% 0.00% -5.52% -3.52% 0.00% -3.99% -3.48% 0.00% -3.85% -4.71% 0.00% -4.97% -3.48% 0.00% -3.73%
Standard deviation: 2.06% 1.20% 1.43% 2.33% 1.70% 1.35% 2.10% 1.36% 1.37% 2.38% 1.63% 1.48% 2.17% 1.74% 1.23%
Pct. > 0: 46.43% 42.86% 35.71% 42.86% 60.71%

ACTIVE TRADER • January 2006 • www.activetradermag.com

 FIGURE 5 60-DAY VIX LOWS
The market lagged its benchmark in the two weeks following both short- and
long-term VIX lows. Here, SPY lost ground on seven of the 10 days after
volatility fell to 60-day lows, a pattern that roughly matched the behavior
after short-term, and 10- and 20-day VIX decreases (not shown).
formance of 0.36 percent.
Table 2 breaks down the details
behind SPY’s response to those extreme
drops in VIX to 60-day lows and sug-
gests that post-VIX-low market patterns
are even less clear than Figure 5 implies.
Four of the 10 average values point in
the opposite direction as their medians
and percentage of gains. For example,
the first day’s 0.12-percent average gain
seems pretty shaky considering its
median performance is -0.03 percent,
and it gained ground just 46.67 percent
of the time.
The market’s response to 60-day VIX
lows is nearly identical its reaction to 10-
and 20-day VIX lows (not shown). SPY
traded sideways in the first two weeks
after 10-day lows and merely climbed
in-line with its benchmark following 20-
day lows.

Longer-term price reaction

tion to dropping volatility was less clear on day 1, but it lost ground and lagged We also measured how the market
than its response to VIX highs. its benchmark (0.04 percent) on seven of behaved during longer-term 15-, 20-,
Figure 5 shows SPY’s average the remaining nine days. For whatever and 40-day periods after VIX extremes.
performance in the 10 days after large reason, SPY rebounded strongly on the The S&P 500 gained ground in nearly all
60-day VIX declines also corresponded ninth day and regained the prior week’s instances, regardless of whether volatili-
with their 60-day lows. The market post- losses. By the 10th day, the market had ty had risen or dropped. But there was
ed an average 0.12-percent average gain lost 0.12 percent vs. a benchmark per- one clear difference between VIX highs

TABLE 2 60-DAY VIX LOW DETAILS

Although SPY climbed an average 0.59 percent on the ninth day following four consecutive daily sell-offs, no clear pattern
emerged immediately after the VIX fell to 60-day lows.
30 instances Day 1 LUM LDM Day 2 LUM LDM Day 3 LUM LDM Day 4 LUM LDM Day 5 LUM LDM
Avg: 0.12% 0.67% -0.54% -0.29% 0.30% -0.79% -0.14% 0.58% -0.67% 0.11% 0.75% -0.53% -0.19% 0.56% -0.78%
Med: -0.03% 0.40% -0.41% -0.08% 0.21% -0.65% 0.14% 0.43% -0.37% 0.24% 0.59% -0.29% -0.21% 0.42% -0.72%
Max: 2.30% 2.61% 0.00% 1.38% 1.48% 0.00% 2.81% 3.22% 0.00% 2.81% 3.22% 0.00% 1.61% 2.20% 0.00%
Min: -2.34% 0.00% -2.68% -1.95% 0.00% -2.36% -2.25% 0.00% -2.40% -2.51% 0.00% -2.67% -2.51% 0.00% -2.68%
Standard deviation: 1.02% 0.66% 0.59% 0.79% 0.33% 0.66% 1.12% 0.67% 0.74% 1.09% 0.80% 0.65% 0.95% 0.60% 0.64%
Pct. > 0: 46.67% 43.33% 56.67% 56.67% 46.67%

Day 6 LUM LDM Day 7 LUM LDM Day 8 LUM LDM Day 9 LUM LDM Day 10 LUM LDM
Avg: -0.09% 0.61% -0.79% -0.02% 0.62% -0.69% -0.12% 0.70% -0.80% 0.59% 0.95% -0.55% -0.11% 0.59% -0.77%
Med: -0.03% 0.41% -0.61% 0.13% 0.41% -0.47% 0.20% 0.56% -0.39% 0.60% 0.69% -0.25% -0.22% 0.43% -0.63%
Max: 1.75% 2.22% 0.00% 2.37% 2.61% 0.00% 2.52% 2.61% 0.00% 5.77% 6.52% 0.00% 1.75% 2.22% 0.00%
Min: 2.34% 0.00% -2.68% -2.34% 0.00% -2.55% -7.25% 0.00% -7.61% -1.30% 0.00% -3.23% -2.75% 0.00% -2.77%
Standard deviation: 0.99% 0.67% 0.73% 0.92% 0.58% 0.72% 1.66% 0.64% 1.42% 1.30% 1.20% 0.75% 0.97% 0.59% 0.68%
Pct. > 0: 50.00% 56.67% 56.67% 73.33% 43.33%

www.activetradermag.com • January 2006 • ACTIVE TRADER

and lows: SPY was more likely to out- which shows SPY climbed on the first
perform its benchmark price moves after day after VIX highs, a tendency that
VIX highs than it was after lows. intensified as volatility reached succes-
This pattern makes sense if you con- sively higher extremes. However, this
sider the market drops as volatility study’s remaining figures and tables
reaches extremely high levels. SPY’s failed to find other signs the market sig-
bullish moves following these extremes nificantly reacted on a longer-term basis
highlight its tendency to bounce back to those VIX extreme price swings that
after sharp down moves, a pattern evi- also led to highs or lows.
dent in other studies of price gaps, run This analysis is limited because it
days, and spike days. focused solely on volatility moves with-
In contrast, the market tends to rally out combining them with specific S&P
the day the VIX reaches extreme lows, price action, a tactic that may have
but its subsequent market behavior is uncovered more interesting patterns.
less predictable. The study is also constrained by its
definition of extreme VIX moves. A
Bottom line large, 95th-percentile VIX climb or drop
The clearest relationship between the had to also reach a relative high or low,
VIX and the S&P 500 appears in Figure 3, which filtered out many examples.

Related reading
“Filling in the gap picture,” Active Trader, November 2005.
Are gaps really meant to be filled? Discover what objective analysis in the
S&P 500 shows about these patterns.

“Can ‘reversal days’ catch market turning points?” Active Trader, October
2005.
Find out how significant reversal bars have been in the S&P 500 since 1993.

“Spikes and spiders,” Active Trader, August 2005.

This analysis shows how the popular S&P 500 tracking stock (SPY) reacted to
one-day price spikes over the past 12 years.

“Forecasting the VIX,” Active Trader, June 2005.

A novel approach to analyzing VIX results in a volatility forecasting technique
and countertrend volatility trading method.

“Trading system lab: VIX system,” Active Trader, December 2004.

This system test is designed to see how a basket of stocks perform following a
short-term price drop combined with an uptick in the VIX index.

“Trading the VIX,” Active Trader, April 2001.

Learn about a strategy that trades off extremes in the VIX.

“Putting volatility to work,” Active Trader, April 2001.

Get a handle on the essential concepts and learn how to improve your trading
with practical volatility analysis and trading techniques.

You can purchase past Active Trader articles at

www.activetradermag.com/purchase_articles.htm and download them to
your computer.

ACTIVE TRADER • January 2006 • www.activetradermag.com

FUTURES
Trading System Lab
FIGURE 1 TRADE EXAMPLES
These two representative trades show how the system takes advantage of
VIX-based system temporary market overreactions.

Market: Stock index futures.

System concept: In 1993, the Chicago

Board Options Exchange (CBOE)
introduced the Volatility Index (VIX),
which reflected the volatility of at-the-
money S&P 100 index options. In 2003,
the VIX was revised to reflect S&P 500
options. Since its inception, the VIX
has been an industry benchmark for
stock market volatility.
Usually, during periods of financial
stress or panic accompanied by market
declines, option prices and the VIX
tend to rise. As a result, the VIX is
often referred to as an “investor fear
gauge” — the more fear there is in the
market, the higher the VIX level. As
fear decreases, option prices tend to
decline, which in turn causes the VIX
to decline. Overall, the VIX tends to
increase during stock market declines
and decrease when the market
advances. For more information about
the VIX, see “Tracking VIX swings” (p. Source for all figures: Wealth-Lab Inc. (www.wealth-lab.com)
3) or visit the CBOE Web site
(www.cboe.com/vix). FIGURE 2 EQUITY CURVE
The following test uses a system described There was a smooth increase in account equity over the test period.
by Larry Connors in the TradingMarkets.com The system’s exposure is relatively low.
blog on Sept. 16
(www.tradingmarkets.com/site/blog). The
system tries to find oversold situations in the
S&P by identifying VIX spikes, as follows:

1. The S&P 500 is trading above its

200-day simple moving average.
2. Today the S&P 500 makes a 10-day low.
3. Today the VIX closes five percent or
more above its 10-day simple moving
average.
4. Buy the S&P 500 on the close and exit
when it closes above its 10-period
simple moving average.

Notice this is a long-only approach. Our

test creates rules for short trades by inverting
the long-trade rules.

System rules:
Go long at close if:
• the S&P 500 is above its 200-day simple
moving average; and
• today made a 10-day low; and
• today the VIX closes five percent or
more above its 10-day simple moving
average.

www.activetradermag.com • January 2006 • ACTIVE TRADER

 Exit long at close if the S&P 500 is above its 10- FIGURE 3 DRAWDOWNS
day simple moving average.
The system had small drawdowns and recovered from them quickly.
Go short at close if:
• the S&P 500 is below its 200-day simple
moving average; and
• today it made a 10-day high; and
• today the VIX closes five percent or more
below its 10-day simple moving average.

Exit short at close if the S&P is below its 10-

day simple moving average.

Figure 1 shows two sample trades, one long

and one short. The S&P is in the upper pane and
the VIX is in the lower pane. With the VIX, the
SMA +/-5 percent difference is used as the entry signal. Because of the very low exposure and the low drawdown,
The short trade occurred on Oct. 6, 1999 when the S&P was we set the maximum risk per position at 10 percent of total
below its 200-day SMA, the S&P made a 10-day high, and the account equity to calculate the number of contracts (we usual-
VIX was more than 5 percent below its 10-day SMA. Five days ly use 1-2 percent).
later, the S&P dropped below its 10 day-SMA and the system The number of contracts per position is calculated using the
covered the position with 49.4 points of profit. basis price (the closing price of the entry bar), the stop-loss
The long trade occurred on Nov. 30. The S&P was above its level, the contract’s point value (i.e., the dollar value of a one-
200-day SMA and made a 10-day low, and the VIX moved 5 point move), and the portfolio’s total equity.
percent above its 10-day SMA. The system exited shortly on For example, the S&P futures contract has a point value of
Dec. 3 when the S&P traded above its 10-day SMA — a 51.4- $250. Assume the system goes long at 1,000 (the basis price)
point profit in three days. and the stop-loss is 900. To determine the trade’s dollar risk,
multiply the point value ($250) by the difference between the
Risk control and money management: basis price and the risk-stop (1,000 - 900 = 100). Therefore, a
1. Set a stop-loss at three times the 10-day average true single contract’s dollar risk is $25,000.
range (ATR). If the portfolio’s total equity before entering the position was
2. Risk 10 percent of account equity per position. 

STRATEGY SUMMARY
PERIODIC RETURNS
Profitability Trade statistics
Net profit ($): 1,671,657.75 No. trades: 80 Avg. Sharpe Best Worst Percentage Max. Max.
Net profit (%): 167.17 Win/loss (%): 72.50 return ratio return return profitable consec. consec.
Annualized gain (%): 10.33 Avg. profit/loss (%): 0.85 periods profitable unprofitable
Exposure (%): 0.44 Avg. profit/loss ($): 20,895.72 Monthly 0.86% 1.05 8.44% -9.62% 43.33 5 8
Profit factor: 2.11 Avg. winner (%): 1.93
Payoff ratio: 0.97 Avg. loser (%): -1.98 Quarterly 2.60% 1.09 12.03% -7.43% 67.50 6 4
Recovery factor: 4.38 Avg. hold time (days): 5.74 Annually 10.82% 0.97 31.48% -6.67% 90.00 9 1
Drawdown Avg. hold time (winners, in days): 4.40
Max. DD (%): -12.70 Avg. hold time (losers, in days): 9.27
Longest flat days: 235 Max. consec. win/loss: 9/3 LEGEND: Avg. return — The average percentage for the period • Sharpe
ratio — Average return divided by standard deviation of returns (annualized)
LEGEND: Net profit — Profit at end of test period, less commission • • Best return — Best return for the period • Worst return — Worst return
Annualized gain — Compounded annual growth rate, or annual percentage for the period • Percentage profitable periods — The percentage of periods
return • Exposure — The area of the equity curve exposed to long or short that were profitable • Max. consec. profitable — The largest number of con-
positions, as opposed to cash • Profit factor — Gross profit divided by gross secutive profitable periods • Max. consec. unprofitable — The largest num-
loss • Payoff ratio — Average profit of winning trades divided by average loss ber of consecutive unprofitable periods
of losing trades • Recovery factor — Net profit divided by max. drawdown
• Max. DD (%) — Largest percentage decline in equity • Longest flat days Trading System Lab strategies are tested on a portfolio basis (unless
— Longest period, in days, the system is between two equity highs • No. otherwise noted) using Wealth-Lab Inc.’s testing platform.
trades — Number of trades generated by the system • Win/Loss (%) — The If you have a system you’d like to see tested, please send the trad-
percentage of trades that were profitable • Avg. profit/loss — The average ing and money-management rules to editorial@activetradermag.com.
profit/loss for all trades • Avg. winner — The average profit for winning
trades • Avg. loser — The average loss for losing trades • Avg. hold time — Disclaimer: The Trading System Lab is intended for educational purposes only to
The average holding period for all trades • Avg. hold time (winners) — The provide a perspective on different market concepts. It is not meant to recommend
average holding time for winning trades • Avg. hold time (losers) — The or promote any trading system or approach. Traders are advised to do their own
research and testing to determine the validity of a trading idea. Past performance
average holding time for losing trades • Max. consec. win/loss — The max-
does not guarantee future results; historical testing may not reflect a system’s
imum number of consecutive winning and losing trades
behavior in real-time trading.

www.activetradermag.com • January 2006 • ACTIVE TRADER

FUTURES
Trading System Lab continued from p. 63

$1,000,000 and we do not want to risk more FIGURE 4 ANNUAL PERFORMANCE

than 10 percent of our total equity ($100,000),
we would buy four contracts. Nine out of 10 years in the test period were profitable.
Had total equity been less than $250,000, we
would not have been able to take this position
because its dollar risk would exceed our sys-
tem’s 10-percent equity risk. This position-siz-
ing method keeps us out of risky trades that
have potential to ruin our account.

Starting equity: $1,000,000. Deduct $20 com-

mission per round-trip trade per contract.
Apply two ticks of slippage per market order.

Test data: Ratio-adjusted S&P futures continu-

ous contract (SP) data from Pinnacle Data
Corp.
FIGURE 5 EQUITY CURVE FOR THE LONG-ONLY SYSTEM
Test period: January 1995 to January 2005.
Similar performance during the bull market but a long flat period during
Test results: The system trades infrequently the bear market.
and exits quickly. There were 80 trades, with
an average holding time of only 5.74 days;
72.5 percent of trades were profitable. The
equity curve (Figure 2, p. 9) shows a steady
increase over the test period (the annualized
profit was 10.33 percent).
The drawdown curve (Figure 3, p. 10)
shows the system experienced only a few
drawdowns — most between 2 and 6 per-
cent and the maximum at 12.7 percent,
which occurred at the end of the test period.
Figure 4 shows 9 of 10 years were profitable.
These very positive results are even more
impressive in light of the fact the system has a
very low exposure level of 0.44 percent (the
small, light-green area of the portfolio equity
curve in Figure 2 represents the system’s expo-
sure). The average trade size was 12 contracts,
and the biggest trade was 31 contracts.
Overall, the test results suggest the system was
able to effectively detect oversold-overbought
situations.
As an additional confirmation we tested
the system without the rules for the short
trades (the original rules presented by Larry
Connors). Figure 5 shows the portfolio equi-
ty curve of the long-only system. There is a
long flat period from beginning of 2000 until beginning of the fact that the system was tested on only one instrument and
2004. This was a bear market period when the S&P traded it trades on average eight times a year, staying in the market
mostly below its 200-day SMA. Therefore, no trades occurred only 5.74 days per trade.
and consequently, the portfolio equity curve did not change. It might be that, although the system detects special situa-
The profit is also lower compared to the system including tions very well, there is a risk it’s not robust enough to with-
short trades. This confirms it was a good decision to include stand changing market conditions. Be careful when using it in
the short rules. the future. Instead, it might be wise to consider using it in
addition to a portfolio-based system.
Bottom line: The system test showed high profitability and
low risk over the simulation period. However, do not overlook —José Cruset of Wealth-Lab

www.activetradermag.com • January 2006 • ACTIVE TRADER

 TRADING Strategy

The VIX fix

A “synthetic” VIX calculation can be used in any market to reproduce
the performance of the well-known volatility index.

BY LARRY WILLIAMS

W
hen it comes to
describing what
markets do, Bernard
Baruch said it best:
“Markets fluctuate.”
That idea is embodied in the Chicago
Board Options Exchange (CBOE)
Volatility Index (VIX), which has become
a very popular measure of market risk
since it was introduced in 1993. The VIX, FIGURE 1: SYNTHESIZING THE VIX The behavior of the actual VIX (middle) and
the Williams synthetic VIX (bottom) are nearly identical.
which is derived from the implied volatil-
Source: TradeNavigator.com
ity of stock index options, is intended to
represent traders’ expectations of volatili-
ty over the next 30 days. (See “Key con- Unfortunately, the VIX is calculated First, let’s look at some market
cepts” on p. 88 for background informa- only for the S&P 500 index, Nasdaq examples.
tion on implied volatility and the VIX.) Composite index, and the Dow Jones
Essentially, the VIX reflects investor Industrial Average. What about other Parallel behavior
fear — high readings are associated with markets? Figure 1 shows the S&P 500 index with
high-volatility conditions (and market Luckily, it is easy to duplicate the VIX the actual VIX in the middle panel and
bottoms) while low readings are associat- for any market — Treasury bonds, gold, the Williams synthetic VIX (the “Williams
ed with low-volatility conditions (and silver, soybeans, even individual stocks VIX Fix,” or WVF) in the bottom panel.
market tops). — with a simple formula. The two indicators’ swings, levels, timing,

www.activetradermag.com • December 2007 • ACTIVE TRADER

and magnitude are almost identical.
Figure 2 shows how closely the VIX and
the WVF paralleled one another from
1994 to 2000.
If this synthetic VIX parallels the actu-
al VIX, then it would seem we have a
tool for measuring volatility and identify-
ing significant highs and lows in other
markets, as well as individual stocks.
Also, unlike the CBOE VIX, the WVF is FIGURE 2: S&P 500: 1994-2000 The synthetic VIX is based on a simple technical
not based on a time-consuming calcula- formula rather than option prices.
tion derived from option prices. Source: TradeNavigator.com
Figure 3 shows what the WVF looks
like in the 30-year T-bond futures (US)
back to 2003. The chart highlights

Investor sentiment
is more magnified
at market bottoms
than at tops.

volatility’s natural high-to-low cycle: Low

volatility is more often associated with
market peaks, while high volatility is
associated with market lows — the same
behavior we saw in the actual VIX rela-
tive to stock prices in Figures 1 and 2. FIGURE 3: T-BONDS When the WVF is particularly high, a market bottom is close
When the WVF is particularly high, a at hand and the daily ranges tend to be larger.
market bottom is close at hand and the Source: TradeNavigator.com

ACTIVE TRADER • December 2007 • www.activetradermag.com

Trading Strategy continued

daily ranges tend to be larger. Such times

are associated with a subsequent contrac-
tion in volatility..
A weekly chart of gold (GC) from
1993 to 1997 (Figure 4) also underscores
the relationship of the WVF to price.
Notice the index’s low readings typically
correspond to significant market peaks.
When volatility is “hanging out” at low
levels, the market is either in a rally lead-
ing to a top, or a top is about to occur.
This is true whether you are looking at a
market average or an individual stock or
FIGURE 4: GOLD The WVF’s low readings typically correspond to significant market commodity.
peaks. When volatility is “hanging out” at low levels, the market is either in a
In a way, the actual VIX or the synthet-
rally leading to a top, or a top is about to occur.
ic WVF turns price upside down — but
Source: TradeNavigator.com
with an important twist. Investor senti-
ment is more magnified at market bot-
toms than at tops. The VIX is much more
successful at highlighting the intense
investor emotion typically associated with
significant lows.
Figure 5 brings the gold chart and the
WVF indicator up to date. The past four
years show the same behavior evident in
the previous charts: Significantly high
readings are usually associated with mar-
ket bottoms. Even the May 2006 run-up
was not associated with a massive
increase in volatility as expressed by this
indicator. However, the market lows in
June and September were associated with
high readings. This is the same pattern
exhibited by the actual VIX in the stock
FIGURE 5: RECENT GOLD Significantly high readings are usually associated with market.
market bottoms. The chart of the British pound futures
Source: TradeNavigator.com (BP) in Figure 6 shows how the indicator
highlights significant market bottoms on

www.activetradermag.com • December 2007 • ACTIVE TRADER

the daily time frame, as well as expan-
sions of daily ranges. Remember, low
readings indicate the market is some-
where near a top during a rally.
There are two remaining questions:
Does the WVF work for individual
stocks? Also, are there ways of using the
index other than referencing its absolute
high and low readings?

Individual stocks
Figures 7-9 are charts of three of the
most actively traded U.S. stocks. Again
we see the same phenomenon — when
FIGURE 6: DAILY BRITISH POUND The WVF displays the same characteristics on the
volatility is high stocks tend to bottom;
daily time frame.
when volatility is low stocks tend to top
Source: TradeNavigator.com
out.
There are many ways to use this infor-
mation. For example, since you can buy
and sell the VIX itself, you might create a
trading strategy for that purpose. Also, in
terms of selling options, when the WVF
indicates volatility is extremely high, you
should expect a contraction in market
ranges and, of course, you are given
some insight into future direction.
Figure 7 is a chart of Microsoft
(MSFT) — perhaps the most widely fol-
lowed stock in the world. It shows the
WVF’s relevance on a weekly basis. Every
significant market low in the stock has
come at a time of increasing volatility as
measured by the WVF.
Figure 8 (General Electric) and Figure
9 (Starbucks) suggest this is a universal FIGURE 7: MICROSOFT Each significant low has come at a time of increasing
cycle of market activity — this is how volatility as measured by the WVF.
stocks actually move. That is not to say Source: TradeNavigator.com

ACTIVE TRADER • December 2007 • www.activetradermag.com

Trading Strategy continued

this is a be-all, end-all indicator, but it

does do a pretty good job of showing the
internal structure of the market and what
the next cycle of market activity is most
likely to be. It doesn’t matter what the
company does — it could make coffee or
manufacture coffee makers — the volatil-
ity cycle is the same.

Possible modifications
You might already have thought of ways
of using the WVF or making it better,
but here are a couple of suggestions.
First, because volatility does fluctuate and
may not have absolute levels of highs
and lows, you could plot Bollinger Bands
(see “Key concepts,” p. 88) to help iden- FIGURE 8: GENERAL ELECTRIC When the WVF indicates volatility is extremely high,
tify potential over extended zones with- you should expect a contraction in ranges.
out resorting to fixed values. Source: TradeNavigator.com
Notice in Figure 10 the VIX and the
WVF hit the Bollinger Bands at just
about the same time. This confirms what
we’ve seen earlier — these indicators are
essentially the same, or at least the
movement is almost identical.
What works in a stock market average
should work for an individual stock.
With that in mind, Figure 11 shows
Bollinger Bands on the daily WVF Index
for Starbucks. We see pretty much the
same thing as we have noted on individ-
ual stock indexes.
Another way to understand the WVF
a little better would be to construct a 14-
day stochastic (see “Key concepts,” p.
88) of the WVF, as shown in Figure 12.
Again, although these are daily charts,
the same concept occurs on weekly FIGURE 9: STARBUCKS The WVF highlights the internal structure of the market and
charts. what the next cycle of market activity is likely to be.
It is apparent that most of the market Source: TradeNavigator.com

www.activetradermag.com • December 2007 • ACTIVE TRADER

Trading Strategy continued

lows come when the stochastic of the

WVF posts a high reading (above 80 per-
cent) while most of the market lows
come when the index is below 20 per-
cent.

Synthetic VIX formula

There is no absolute trading strategy for
the WVF; it is more beneficial as a refer-
ence point for understanding what
volatility cycle a market is going
through, and also suggesting the most
obvious direction for the next move. The
formula for the index is very simple:
FIGURE 10: BOLLINGER BAND COMPARISON The VIX and the WVF hit the Bollinger
WVF = (Highest (Close,22) Bands at just about the same time, confirming the indicators’ similarity.
- Low)/(Highest(Close,22))*100 Source: TradeNavigator.com

This means you find the highest close

of the past 22 bars and then subtract the
low of the current bar from this price.
The result is divided by the highest close
of the last 22 bars. Finally, the result is
multiplied by 100 to normalize the indi-
cator readings.
There was no optimization involved in
selecting the indicator’s 22-day period.
The only reason this value was selected
is that the maximum number of trading
days in a month is 22. (Note: I have
found that values for all moving aver-
ages, oscillators, etc., return the best
results using a number between 20 and
22. I like 22 because that covers all
potential months.)
This allows us to look back on a time FIGURE 11: INDIVIDUAL STOCK The WVF’s behavior appears consistent on different
frame that relates to the stock market time frames and markets.
activity itself, as there are so many Source: TradeNavigator.com

www.activetradermag.com • December 2007 • ACTIVE TRADER

Trading Strategy continued

monthly influences in stock and com-

modity prices. You might want to change
the 22 setting on weekly charts to 26
weeks (six months). That was not done
in these examples, as the 22 and 26 set-
tings generally produce similar results,
but from a purist view point that would
seem wise.
Regardless, a simple formula that can
fill in for the complicated CBOE VIX is
well worth further research and experi- FIGURE 12: WVF WITH STOCHASTICS Most of the market lows come when the
mentation.
stochastic of the WVF posts a reading above 80 percent while most of the
market lows come when the index is below 20 percent.
Source: TradeNavigator.com

www.activetradermag.com • December 2007 • ACTIVE TRADER

 TRADING STRATEGIES

VIX options
Despite appearances, VIX options behave differently from other options.

BY MARC ALLAIRE

FIGURE 1 — VXO AND VIX, 1993-2007

The VIX has dropped below 10 just a few times since 1993, and it spiked sharply much

I
more often.
n February 2006 the
Chicago Board Options
Exchange (CBOE) listed
options on its S&P 500
volatility index (VIX). The VIX
tracks the implied volatility
(IV) of selected S&P 500
options and measures the mar-
ket’s volatility forecast over the
next 30 days.
VIX options represent a fair-
ly abstract concept — options
on an index that tracks the
implied volatility of S&P 500
index (SPX) options. Despite
their complexity, though, VIX
options have become popular
instruments among retail
traders. In October 2006, VIX
option average daily volume
was only 27,574, a figure that
jumped to 119,630 a year later. Source: CBOE
However, the VIX doesn’t
behave like a standard stock index or futures contract, and
February 1994 to November 2006.
its options are also somewhat counterintuitive.
But what about those frequent spikes? Table 2 lists all 40+
VIX readings and shows the index rose above that level 17
Historical VIX patterns times since 1993, seven of which were only intraday moves.
Figure 1 shows a weekly VIX chart since May 19, 1993 and
Table 2 also suggests the VIX’s highest close and its intraday
combines two symbols: The original volatility index (VXO),
high may have occurred on two different days.
which was revised in September 2003 to become the current
VIX. The original volatility index, which the CBOE still cal-
S&P 500 vs. VIX
culates, tracks the IV of eight at-the-money (ATM) options
The S&P 500 index and the VIX are mostly negatively
on the S&P 100 (OEX), while the current one measures the
correlated, which means when stocks climb, volatility tends
IV of various options on the S&P 500. For a detailed discus-
to fall and vice versa. Figure 2’s daily chart juxtaposes the
sion of the VIX and this change, see “The volatility index”
S&P 500 and the VIX from January 2006 to mid-March 2007.
(p. 20). The S&P 500 rose about 12 percent from early August
The VIX dropped below 10 just a few times, and it spiked through December, while the VIX declined from 15 to
sharply (albeit briefly) much more often. Let’s quantify around 11. Similarly, when the S&P fell 3 percent in May
these moves more precisely. Table 1 lists all the days in 2006, the VIX spiked from about 12 to nearly 20.
which the VIX fell below 10, either intraday or on a closing This phenomenon is easily explained by the concept of
basis. Single-digit VIX readings were quite rare, as the options as insurance. When the market rallies every day, the
volatility index stayed above 10 for more than 12 years from demand for options (e.g., insurance) dries up; and when

December 2007 • FUTURES & OPTIONS TRADER

 TABLE 1 — LOW-VOLATILITY DAYS
Over the past 15 years the VIX has dropped below 10 on just 25
buyers disappear, prices and IVs fall. If a market’s days. It remained above this level from Feb. 4, 1994 to Nov. 20,
decline seems dire, buying some insurance against 2006.
further declines appears prudent. Days VIX Intraday Closing
Dates below 10 basis basis
Interpreting the VIX Dec. 20-29, 1993 8 1 7
There are two ways to interpret the VIX: as a Jan. 27-Feb. 4, 1994 7 5 2
barometer that predicts where the S&P 500 could Nov. 20-22, 2006 3 1 2
be headed or as a thermometer that shows how Dec. 14-15, 2006 2 1 1
“hot” the S&P is. Traders who view the VIX as a Jan. 24-25, 2007 2 1 1
barometer see VIX extremes as possible market
Feb. 2, 14, & 16, 2007 3 3 0
turning points (see “Related reading”). This discus-
sion, however, is limited to interpreting the VIX as
a thermometer. futures’ theoretical price from the S&P 500 cash index, you
Because the VIX measures the temperature of S&P 500 should include carrying costs such as the risk-free interest
options, it is most useful to SPX options traders. According rate and the index’s dividend yield. Also the futures price
to Figure 2, the S&P 500 dropped 4 percent in the first half doesn’t stray too far from its theoretical value, because if it
of 2006, pushing the VIX up from 15 to the low 20s. At this did, traders could profit from arbitrage techniques such as
point, let’s assume you thought, “I’ve seen this before. The buying the S&P’s component stocks and selling the corre-
market has overreacted and will bounce back.” If you are sponding futures contract.
bullish, you could use the most basic bullish option strate- It’s a completely different situation with the VIX futures.
gy: buying calls on the S&P 500. Because no tradable underlying exists, VIX futures have no
What’s wrong with this scenario? Your explicit market theoretical value, and arbitrage isn’t possible. In other
forecast is bullish, but your implicit IV forecast is bearish,
assuming you believe the historical inverse relationship
between the S&P 500 and VIX will continue. If the market
does rise, call values will climb, but they will also be hurt by
a drop in implied volatility. The volatility index (VIX)
Buying calls may be profitable, but another strategy may
offer a better return on investment (ROI). The solution: The Volatility Index (VIX) measures the implied volatility
Enter a position that will benefit from both a rallying mar- of S&P 500 index options traded on the Chicago Board
ket and lower volatility. At the very least, make sure a bull- Option Exchange (CBOE). The VIX is designed to reflect
ish trade won’t be hurt by an IV drop. the market expectation of near-term (in this case, 30-
Strategies that fit these criteria include selling puts, bull day) volatility and is a commonly referenced gauge of the
call spreads (long call, short higher-strike call in same stock market’s “fear level.”
month), and bull put spreads (short put, long lower-strike The original VIX, launched in 1990, was derived from
put in same month). For these spreads, the short option’s eight near-term at-the-money S&P 100 (OEX) options
time premium should equal or exceed the long option’s (calls and puts) using the Black-Scholes options pricing
time premium. model.
The VIX underwent a major transformation in late
Look to the future(s) 2003. The current index is derived from both at-the-
An underlying security’s price is one of the variables need- money and out-of-the-money S&P 500 (SPX) calls and
ed to calculate an option’s price. (Other pricing variables puts to make the index better represent the full range of
include time until expiration, volatility, interest rates, and volatility. At the same time the CBOE applied the new
dividends.) If you trade options on stocks, indices, or calculation method to the CBOE NDX Volatility Index
futures, that variable is the current price or value. Stocks (VXN), which reflects the volatility of the Nasdaq 100
and indices can be bought or sold at their current prices, index.
even if buying or selling the 500 stocks in the S&P 500 index The exchange still publishes the original VIX calcula-
is impractical for most retail traders. tion, which can be found under the ticker symbol VXO.
But what about the VIX? It cannot be traded because it is Figure 1 shows both indices: the VXO from May 19, 1993
just a number, not a true underlying security. Therefore, it to Sept 19, 2003 and the new VIX over the next four
can be neither bought nor sold. VIX futures (VX), on the years.
other hand, can easily be traded. So the true underlying of For more information about the VIX and its calculation,
VIX options is the VIX futures, which makes a big differ- visit http://www.cboe.com/vix.
ence.
Consider the S&P 500 futures (SP). To calculate the

FUTURES & OPTIONS TRADER • December 2007

 TRADING STRATEGIES continued

TABLE 2 — HIGH-VOLATILITY DAYS

The VIX exceeded 40 more than a dozen times since 1993 and
stayed above this level for 105 days overall.
Days VIX Highest Intraday
Dates above 40 close high
Oct. 27-28, 1997 2 39.96 55.48 Where is the money?
Nov. 13, 1997 1 36.98 40.40 VIX options use VIX futures instead of the cash VIX
Aug. 27-Sept. 14, 1998 13 48.33 53.43 as the underlying. VIX options are based on the
Sept. 17-21, 1998 3 42.33 46.82 futures’ expected, or forward, VIX value at expira-
Sept. 30-Oct. 15, 1998 12 48.56 60.63 tion, which differs from current VIX values because
April 14, 2000 1 39.33 41.53 options in later-expiring months often have differ-
March 22, 2001 1 39.70 41.99
ent IVs.
When VIX futures trade significantly above or
April 3-4, 2001 2 39.33 40.77
below the cash level, the terms in-, at-, and out-of-
Sept. 17-25, 2001 7 49.04 57.31
the-money take on a slightly different meaning.
July 11-26, 2002 12 50.48 56.74 Assume, for example, that VIX is 16.00, but the
Aug. 1-15, 2002 10 49.31 49.83 futures, which expire in three months, trade at
Sept. 3-9, 2002 5 43.86 44.87 18.00. This suggests the market expects higher
Sept. 12-Oct. 25, 2002 32 49.48 50.48 volatility, because the futures price is driven by
Jan. 27, 2003 1 39.77 40.89 expectations, not arbitrage opportunities.
Feb. 10, 2003 1 37.70 40.48 If you buy a 16-strike VIX call, you may think it
Feb. 13, 2003 1 38.48 40.68 is ATM, but, in reality, that call is priced as if it is 2
March 12, 2003 1 38.99 41.16 points in-the-money. And if you buy a 16-strike
put, it will be priced as if it is 2 points out-of-the-
money, not ATM as it appears. In this scenario, the
words, no buy programs kick in if VIX futures seem over- call appears expensive and the put seems cheap.
priced and no sell programs are triggered by seemingly Although VIX put buyers may be duped by this discrep-
underpriced futures. The cash and futures VIX prices can ancy, they still have an inherent advantage, because VIX
diverge significantly from one another, and the gap can options settle to the VIX cash index, not VIX futures. The
remain for extended periods of time. VIX cash index represents the market’s volatility expecta-
tions within 30 days, so as VIX
options near expiration, the
FIGURE 2 — S&P 500 AND VIX actual VIX becomes more
The S&P 500 and VIX are negatively correlated, which means one tends to fall as the accurate as an underlying
other rises and vice versa. Notice how the VIX climbed to 20 as the S&P 500 fell in May security. As expiration
2006 but slipped below 10 as the S&P rallied in November. approaches, both VIX options
and the VIX index are based
on the S&P 500 options in the
same month.
Suppose you buy a 17-strike
VIX put when the cash VIX is
16 and the VIX futures trade at
18. When you enter the trade,
the put is priced one point
OTM, but it will expire one
point ITM if the VIX remains
unchanged.

Strategy implications
You should also remember the
VIX index behaves differently
from stocks and other indices.
For example, a stock can jump
Source: eSignal
from $100 to $400 and never

December 2007 • FUTURES & OPTIONS TRADER

 Related reading
Marc Allaire articles
“Putting put-call parity to work”
Futures and Options Trader, August 2007.
Theories don’t pay the bills. This analysis focuses on the
practical side of the put-call parity equation.

trade at $400 again (think Google, at least so far). But if “Rolling profitable covered calls”
VIX jumps from 10 to 40, the odds are fairly good that it Futures and Options Trader, April 2007.
will drop below 40 again. Table 2 shows the VIX has Taking profits on a winning covered call is tempting, but extend-
never stayed above 40 for more than 32 days. This his- ing the trade another month could generate additional profits.
This first installment of a two-part series examines the benefits
torical pattern has implications for strategies such as
and drawbacks of rolling a profitable covered call position as
selling naked calls.
expiration nears.
When you sell an uncovered call, you collect premium
in exchange for the obligation to sell stock to the call’s “Repairing a losing covered call”
holder at the strike price, which translates to theoretical- Futures and Options Trader, May 2007.
ly unlimited risk. Although you can debate whether this This sequel to the first article on covered calls shows how
upside risk is really unlimited, you can easily get wiped several repair strategies can help minimize a covered call’s
out if the market rallies and get caught on the wrong losses — and can occasionally even turn a loser into a winner.
side of a short naked call.
For VIX, the risk of selling naked calls is running out “Straddles vs. strangles, round two”
of money before the market proves you right. In other Options Trader, January 2007.
Neither strategy always outperforms the other. However, having
words, naked call sellers need to hold a great deal of
a clear price forecast makes it easier to select the best position.
capital so that if they get caught by one of the VIX’s
unexpected spikes, they will have enough staying “Selecting calls based on ROI”
power to outlast margin calls. Options Trader, October 2006.
Selling naked puts could be a portfolio-hedging strat- Traders seem drawn to complex options strategies, but some-
egy. A market correction could hurt you, but if you sell times simply buying calls is the best way to catch an up move.
VIX puts as a hedge, the premium you collect will help Learn how to weigh the possibilities by comparing various calls’
offset those losses. However, this strategy is risky if the return on investment.
market either goes nowhere or rallies consistently. In
these situations, the VIX will probably drop. Other articles
“Getting a grip on implied volatility”
Six degrees of separation
Options Trader, February 2006.
If the VIX index acts as a thermometer and shows how
Implied volatility is a crucial, but often misunderstood, concept.
“hot” SPX options are, how can you gauge the tempera- We explain what it means and how you can use it to improve a
ture of VIX options themselves? The answer is the trade’s chance of success.
implied volatility of VIX options. A note of caution here:
You might generate some bizarre results if you calculate “VIX-based system”
VIX options’ implied volatility with option-pricing soft- Active Trader, January 2006.
ware designed for stock or index options. This test uses a system described by Larry Connors in the
One of the assumptions behind these models is that TradingMarkets.com blog on Sept. 16. The system tries to find
an underlying’s returns are normally distributed, which oversold situations in the S&P by identifying VIX spikes.
means there is an equal possibility of a positive or nega-
“Forecasting the VIX”
tive return, and that most returns will be relatively small
Active Trader, June 2005.
with few very high or low returns. But the VIX is a
A novel approach to analyzing the VIX results in a volatility
mean-reverting index, so this assumption is inaccurate. forecasting technique and countertrend volatility trading
In other words, the VIX has a long-term tendency to method.
move back toward its mean, which is about 16 percent.
Expect unusual IV values when the VIX trades at “The volatility market connection”
extreme lows or highs. If VIX is above 40, then it is more Active Trader, March 2004.
likely to drop than climb — the odds of it going in either Is everything you know about volatility wrong? Find out what
direction are no longer even. An accurate implied history says about the volatility-market relationship — and what
volatility estimate would have to take into account this the VIX is saying about the stock market's 2004 prospects.
mean-reverting phenomenon.

You can purchase and download past articles at
http://www.activetradermag.com/purchase_articles.htm.

FUTURES & OPTIONS TRADER • December 2007

 TRADING STRATEGIES

Hedging with VIX options

The VIX’s unique characteristics give its options certain advantages
in constructing a stock-market hedge.

FIGURE 1: THE STOCK MARKET AND THE VIX

BY STEVE PAPALE

T
he stock market’s
increased volatility
has heightened
interest in the
CBOE Volatility Index (VIX),
which is often described as the
“fear index” because of its role in
reflecting investor sentiment. The
VIX represents the market’s The stock market (represented here by the S&P 500 tracking stock, SPY) and the VIX
expectation of 30-day volatility in tend to be inversely correlated. Unlike stock prices, however, VIX prices have a
the S&P 500 index (SPX), as mean-reversion characteristic.
Source: MetaStock
measured by the implied volatili-
ty of the SPX options.
As perceived risk in the stock demand for protective options goes market falls, the VIX rises, and vice versa.
market rises, investors tend to purchase down, lowering implied volatility and the This is shown in Figure 1, which com-
more options, particularly puts, for pro- VIX. pares the S&P 500 tracking stock (SPY)
tection against a market decline. The Because the VIX typically rises as to the VIX. Notice each time the VIX
greater demand for options causes investor fear increases, there is generally eclipsed 30, SPY tended to bottom. A
implied volatility to go up, which pushes an inverse correlation between price reading at or above this level represents a
up the VIX. Conversely, when investors movement in the VIX and price move- high level of nervousness and pessimism
perceive the market to be less risky, the ment in the S&P 500 — as the stock among traders. Such exceptionally high

www.activetradermag.com • July 2008 • ACTIVE TRADER

 The VIX is often cast
simply as a sentiment
indicator, but VIX
VIX readings may be used as a bullish extremely large price movements for an
contrary indicator. Conversely, a relatively extended period of time.
futures and options
low VIX reading (e.g., below 20) may But the VIX tends to fluctuate in a
indicate market complacency and func- range roughly between 10 and 35 — a
offer investors
tion as a bearish contrarian signal. characteristic known as mean reversion,
However, over time the VIX levels that or the tendency to return over time to a
expanded trading
signal market tops or bottoms can long-term average value. Interest rates
change. For, example, during the second and implied volatility tend to exhibit
opportunities.
half of 2006 and into early 2007 the VIX mean reversion, while exchange rates and
gradually declined until it bottomed out stock prices do not.
The VIX options can
around 10 and topped out around 13. This information is very useful to
The market sell-off that occurred on Feb. know when trying to decide the most
be used to speculate
27, 2007 signaled the start of a higher advantageous time to place a VIX trade.
trading level for the VIX that has contin- For example, if the VIX is extremely low
on expected moves
ued to today. and you want to hedge against a down
An investor who is long stocks can use move in the stock market, you may
in the stock market
the negative correlation of the VIX to pro- decide to put on a bullish position in the
tect against losses from a market decline. VIX instead of a more traditional hedge,
or as a hedge to offset
One hedging strategy would be to buy a such as a bearish position in an index.
VIX call or call spread. In a declining A VIX trade may be more advanta-
market risk.
market (when the portfolio is losing geous because if the market does fall,
money), the VIX should rally, causing the both VIX call spreads and SPY put
call or call spread to increase in value, at spreads should perform well. However, if
least partially offsetting any losses in the the market rallies, the SPY put spread will an April 143-141 SPY bear put spread for
stock market. lose a lot of money. The VIX, however, $0.50, or a total cost of $50. At the same
One interesting characteristic of the may not fall significantly since it is time you could have also purchased the
VIX is that its prices are not lognormally already at an extremely low level. April 12-14 VIX bull call spread for a net
distributed like stocks and stock index debit of $0.50, or the same total cost of
prices. (Lognormal distribution essentially Comparing hedges: $50. The April options had 57 days
means the price is just as likely to double VIX spread vs. SPY spread remaining until expiration at this time.
as it is to drop in half.) If the VIX went to Let’s compare how a VIX call spread and By the close on Tuesday, Feb. 27, 2007
zero, it would imply there is no expected a SPY put spread would perform during a the market had dropped substantially —
change in the daily value of the SPX. market sell-off. At the market close on SPY was trading at 139.85 — a 3.75-per-
Conversely, if the VIX were to reach an Friday, Feb. 23, 2007, SPY was at 145.31 cent decline. Table 1 (p. 25) shows
extremely high value and persist, that and the VIX was at 10.58. the performance of both trades over
would indicate the market expected At the time you could have purchased

ACTIVE TRADER • July 2008 • www.activetradermag.com

Trading Strategies continued

TABLE 1: VIX VS. SPY IN A DECLINING MARKET

Asset Opening trade description Closing trade description Gain ($) Gain (%)
SPY Buy 1 April 143 put @ $1.35 Sell 1 April 143 put @ $4.80
Sell 1 April 141 put @ $0.95 Buy 1 April 141 put @ $3.70 $70 175%
for a net debit of -$40 for a net credit of $110

VIX Buy 1 April 12 call @ $1.55 Sell 1 April 12 call @ $3.10

Sell 1 April 14 call @ $1.05 Buy 1 April 14 call @ $1.70 $90 180%
for a net debit of -$50 for a net credit of $140

The VIX and SPY option spreads posted nearly equal results when the stock market declined.

this period. than SPY declined, we realized nearly the the VIX to be a short-term event and that
The SPY put spread was sold for a net same return on both trades (175 percent it will return to more “normal” levels rela-
credit of $110, a gain of $70 or 175 per- vs. 180 percent). tively quickly.
cent. During this same time period, the Why would the positions have nearly Trying to keep conditions as close
VIX moved up to 18.22 (a 72.2-percent the same return when the VIX call spread as possible to the previous example, let’s
increase) and the VIX call spread was sold was much more in the money (on a per- compare how the two hedges would
for a net credit of $140, a gain of $90 or centage basis) than the SPY put spread? perform if the stock market rallied instead
180 percent. Notice that while the VIX Because the VIX is mean reverting. The of declined.
rallied much more (on a percentage basis) market expects the much-higher level in

TABLE 2: VIX VS. SPY IN A RALLYING MARKET

Asset Opening trade description Closing trade description Loss ($) Loss (%)
SPY Buy 1 May 136 put @ $2.30 Sell 1 May 136 put @ $0.70
Sell 1 May 134 put @ $1.75 Buy 1 May 134 put @ $0.65 $50 91%
for a net debit of -$55 for a net credit of $5

VIX Buy 1 May 18 call @ $1.00 Sell 1 May 18 call @ $0.60

Sell 1 May 20 call @ $0.60 Buy 1 May 20 call @ $0.45 $25 63%
for a net debit of -$40 for a net credit of $15

In an advancing market, the VIX’s mean-reverting characteristic contributed to the VIX spread’s much small losses relative
to the SPY spread.

www.activetradermag.com • July 2008 • ACTIVE TRADER

Trading Strategies continued

Related reading
Performance in
an advancing market “Tracking VIX swings”
Active Trader, January 2006.
Looking at Figure 1, an up move of near-
The VIX has been a widely discussed stock market barometer, but how
ly the same magnitude as the decline in reliably does it identify market turning points? This study turned up a few
the previous example occurred just a cou- surprises in analyzing how the S&P 500 tracking stock (SPY) responded to
ple weeks later between March 16, 2007 VIX highs and lows.
and March 21, 2007. Note: This article is also included in the discounted collection, “Market Pulse:
At the close on March 16, SPY was Stock market patterns and tendencies, Vol. 1.”
trading at 138.53 and you could have
“The VIX fix”
purchased a May 136-134 bear put
Active Trader, December 2007.
spread for a net debit of $0.55, or a total The VIX is calculated only for the S&P 500 index, Nasdaq Composite index,
cost of $55. The VIX was trading at 16.79 and the Dow Jones Industrial Average. What about other markets? A
and you could have purchased the May synthetic VIX calculation can be used in any market to reproduce the per-
18-20 bull call spread for a net debit of formance of the well-known volatility index.
$0.40, for a total cost of $40. The May
“VIX options”
options at this time had 64 days remain-
Futures & Options Trader, December 2007.
ing until expiration. Despite appearances, VIX options behave differently from other options.
By the close on March 21, SPY was
trading at 143.29, an increase of 3.44 “The volatility-market connection”
percent from when the trades were Active Trader, March 2004.
entered. Table 2 (p. 25) shows the per- Is everything you know about volatility wrong? Find out what history says
about the volatility-market relationship.
formance of both trades over this period.
The SPY put spread could be sold for a
You can purchase and download past articles at
net credit of only $5, resulting in a loss of www.activetradermag.com/purchase_articles.htm
$50, or 91 percent. By comparison, the
VIX fell to 12.19 (a 27.4-percent
decrease) and the VIX call spread could
be sold for a net credit of $15, a loss of opportunities. The VIX options can be
$25 (63 percent). used to speculate on expected moves in KC For more information about the
Although both hedges performed simi- the stock market, or as a hedge to offset following concepts, go to “Key concepts”
on p. 80.
larly when the market dropped in value, market risk.
the VIX hedge lost much less than the The difference in how the VIX and • Implied volatility
traditional hedge when the market rose. equity indices react to price changes in
• Vertical spreads (bull call spreads
the market make it worth taking the time and bear put spreads)
Expanded trading horizons to learn how to use the VIX as a trading
The VIX is often cast simply as a senti- and risk management tool.

ment indicator, but VIX futures and
options offer investors expanded trading

www.activetradermag.com • July 2008 • ACTIVE TRADER

 ADVANCED STRATEGIES

The VIX and market capitulations

It takes more than a VIX spike to identify a market bottom.

BY HOWARD L. SIMONS

H ang around markets long are those who use the Chicago Board of shareholders who hold these put options
enough and you will learn Options Exchange Volatility Index (VIX) now have less incentive to sell their
to associate key words and to define when we, much like the wino stocks. Both factors contribute to a dissi-
phrases with market struc- awakening in the gutter on a Sunday pation of selling pressure.
tures. If, for example, you hear someone morning, have hit rock bottom. The gen- The key phrase is “at some point.”
eral premise is quite valid: As most Figure 1, which shows the VIX’s history
blathering about the gold-silver ratio, you
investors are net long stocks and buy put (plotted inversely) going back to the late-
know we are in a bull market for pre-
options for insurance, they tend to bid 1994 breakout of the stock market,
cious metals. The opposite holds true for the price of put options higher as the shows numerous points where VIX spikes
terms such as “earnings visibility” and market declines. At some point, put corresponded to upside reversals in the
“capitulation.” Both come out of the option volatility surges to the point where S&P 500 (SPX). These include the onset
woodwork only when we are in a major the price of put option insurance incor- of the Asian crisis in October 1997, the
bear market for stocks. porates a worst-case loss and option buy- Russian/Long Term Capital Management
Associated with market capitulations ers cease and desist. In addition, nervous crisis of October 1998, the September
2001 terrorist attacks and, most
prominently, the September-
FIGURE 1: EXCESS VOLATILITY AND THE S&P 500 November 2008 collapse. Each of
these VIX spikes occurred at very
different absolute levels; the VIX’s
behavior does not look so much like
a long-term white-noise process
wherein spikes revert back to a static
mean, but rather a set of long-term
trends wherein spikes revert back to
different levels.

KC For more information about the

following concept, go to “Key concepts”
on p. 76.

• At the money
• Delta
• Floating-rate payor
Rather than consistently reverting back to a static mean value, the VIX has tended • Gamma
to form long-term trends in which spikes revert back to different levels. As a result, • Implied volatility
any indicator based on the absolute level of the VIX is the product of data mining, • Out of the money
over-parameterized back-testing, or both. • Variance swap

www.activetradermag.com • May 2009 • ACTIVE TRADER

 FIGURE 2: VIX SHOCK & REGRESS SINCE APRIL 14, 1997

The VIX ranges during the late-2008 collapse (within the magenta channel) were markedly higher than those during
the 1997-1998 bull market.

Moreover, the two long bullish periods The Volatility Index

in the SPX, the late 1990s and 2003-
2007, occurred during opposite trends in The Volatility Index (VIX) measures the implied volatility of S&P 500 index
the VIX. The 1990s rally occurred during options traded on the Chicago Board Option Exchange (CBOE). The VIX is
a period of steadily rising volatility while designed to reflect the market expectation of near-term (in this case, 30-day)
the 2003-2007 rally occurred within a volatility and is a commonly referenced gauge of the stock market’s “fear
period of steadily declining volatility. (As level.”
an aside into a much deeper topic, both The original VIX, launched in 1990, was derived from eight near-term at-
trends had good fundamental reasons. the-money S&P 100 (OEX) options (calls and puts) using the Black-Scholes
The 1990s rally was disbelieved by many options pricing model.
and occurred during an environment of The VIX underwent a major transformation in late 2003. The current index
uncertain global monetary policy and is derived from both at-the-money and out-of-the-money S&P 500 (SPX) calls
currency management; the 2003-2007 and puts to make the index better represent the full range of volatility. At the
rally occurred during an environment of same time, the CBOE applied the new calculation method to the CBOE NDX
very stable and predictable global mone- Volatility Index (VXN), which reflects the volatility of the Nasdaq 100 index.
tary policy and currency management.) The exchange still publishes the original VIX calculation, which can be
We should agree that any trading indi- found under the ticker symbol VXO. Figure 1 shows both indices: the VXO
cator based on the absolute level of the from May 19, 1993 to Sept. 19, 2003 and the new VIX over the next four
VIX has to be the product of data min- years.
ing, over-parameterized back-testing, or For more information about the VIX and its calculation, visit
both. Neither approach ever produces a www.cboe.com/vix.
robust system going forward.

ACTIVE TRADER • May 2009 • www.activetradermag.com

Advanced Strategies continued

FIGURE 3: S&P 500 VOLATILITY AFTER SEPTEMBER EXPIRATION

Moving beyond absolute

Let’s rearrange some of the VIX
and SPX data from mid-April
1997 through late January 2009.
The 1997 date was chosen
because, sadly, that was the cross-
ing point to the upside for SPX
prices revisited recently on the
downside. The daily high-low
range of the VIX will be mapped
against the SPX level.
Figure 2 (p. 28) shows the VIX
high-low ranges during the 2008
collapse occurred largely within
the confines of the magenta chan- A huge rise in SPX volatility at all strike prices is evident in the September-October
nel and are markedly higher than 2008 crisis period.
those during the 1997-1998 bull
market. The high-low range of the scale was with portfolio insurance during
bear market low to date, Nov. 20, 2008, the October 1987 crash. Every generation Traders who marched
is marked in red diamonds; the financial has to invent a new self-immolating
sector crash on Jan. 20, 2009 is marked trade. in to sell volatility
with green squares.
The chart confirms John Maynard Skew and you encountered not only
Keynes’ aphorism, “The market can stay Did the nature of volatility change during
irrational longer than you can stay sol- last fall’s credit crunch? It did not really
vent.” Not only did the VIX rise from change so much as it unveiled another
an absolute loss on
record levels, its ranges expanded as well. aspect of itself during the extreme market
Anyone who marched in to sell volatility conditions. instruments such as
encountered not only an absolute loss on We must remember a put-option buyer
instruments such as short VIX futures, should never pay more than the present short VIX futures,
long VIX puts, or short VIX call options, value of the strike; otherwise, the buyer is
they encountered lower liquidity and over-insuring the maximum loss. As most long VIX puts, or short
higher spreads because of the increased option buyers, call or put, tend to buy
volatility of volatility. out-of-the-money (OTM) strikes that VIX calls, they
Those losses might nonetheless be offer incomplete protection but cost less,
preferable to those involving short strad- the volatility of those strikes tends to get encountered lower
dles. There, the increased volatility was bid higher than that of at-the-money
compounded by the delta and gamma (ATM) options. This produces the charac-
losses created by the SPX’s price collapse. teristic “smile” of option volatility where
liquidity levels and
Even worse would be a floating-rate the deep in- or out-of-the-money volatili-
payor position in SPX variance swaps. ties are lower than the ATM volatility. As higher spreads
Variance tends to rise as the square of an aside, the purchase of lower-cost,
volatility, and as volatility rises on the lower-gamma OTM options is contrary to because of the
way down, those who are short variance the option strike selection algorithms dis-
are forced to sell ever-greater quantities of cussed in The Dynamic Option Selection increased volatility
stock at ever-lower prices to hedge their System (Wiley, 1999).
positions. The last time such “sell in the As noted, the asymmetric nature of of volatility.
hole” trading was tried on an industrial stock ownership places put options in

www.activetradermag.com • May 2009 • ACTIVE TRADER

 FIGURE 4: S&P 500 VOLATILITY AFTER SEPTEMBER EXPIRATION

relatively high demand, even at

higher volatility; this skews the
smile toward strikes lower than the
ATM. (That systematic skewing
creates a trading opportunity for
put-write strategies relative to buy-
write strategies, as discussed in
“Putting the put-write right,” Active
Trader, August 2008.)
How did the SPX skew change
during the September-October
2008 crisis? Let’s map the daily
SPX option volatility between the
September and October 2008
option expirations against “money-
ness,” or a level at a percentage of As volatility increased, the higher-priced strikes’ volatility increased as a percentage of
the ATM strike (Figure 3). the ATM volatility, while the opposite occurred for the lower-priced strikes.
As the huge rise in SPX volatili-
ty at all strikes is the dominant
feature of this chart, we should FIGURE 5: EXCESS VOLATILITY AND THE S&P 500
rearrange the data and display the
readings as a percentage of the
ATM volatility for each day
(Figure 4).
Here a pattern emerges. As
volatility in general rose, the high-
er-priced strikes’ volatility
increased as a percentage of the
ATM volatility, and the opposite
occurred for the lower-priced
strikes. The pattern for the lower-
priced strikes suggests put-option
buyers became more willing to
buy the higher-cost, higher-
gamma near-ATM strikes as
volatility rose. Whether they
should have engaged in that
behavior in the first place is a
moot point.
As yet another aside, the SPX excess volatility readings below 1.00 are infrequent, but they have occurred at
mechanics of VIX construction market bottoms such as October 1998, April 2000, October 2002, and March 2007.
places a greater weight on the
ATM strikes. Even if nothing occurred in simple and grim: The put-option buyer is Lehman Brothers, Fannie Mae, Freddie
overall volatility levels, the VIX would not buying a floor price underneath a Mac, and Washington Mutual, just for
have risen by virtue of higher ATM stock or index so much as protection the starters.
volatility relative to OTM volatility during likelihood a stock is going to go to or The volatilities for individual financial
this particular time period. near zero quickly. That happened to a stocks were indeed pricing in the risk of
What can explain this change in strike number of financial stocks during ruin during this period. Volatility on
preference? The answer in this case is September-October 2008; think of

ACTIVE TRADER • May 2009 • www.activetradermag.com

 Advanced Strategies continued

Morgan Stanley reached 495 percent environment replete with such stocks were characterized by high absolute
when its capital infusion by Mitsubishi look like? Let’s compare the ATM implied volatility.
UFJ came into question during October volatility of the SPX — not the same We can conclude, therefore, that
2008. An October $12.50 put expiring in thing as the strike-weighted VIX — to the when absolute implied volatility is high
a week closed at $4.80 while the stock SPX’s high-low-close volatility. This meas- but is less than absolute high-low-close
itself closed at $9.68. If that is not evi- ure incorporates the effects of intraday volatility, the market is walking away
dence of a binary bet on whether Morgan price range as well as interday price from options as insurance. It is this
Stanley was going to stay in business, change (Figure 5, p. 30). relative low volatility, not the high
what was it? Excess SPX volatility readings less than absolute volatility, which defines a
1.00 are uncommon, but they have capitulation bottom in a non-parameter-
Relative volatility occurred during market bottoms such as ized fashion.

If such a bet on an individual stock is a October 1998, April 2000, October 2002,
bet on capitulation, what does an index and March 2007. Several of these periods

Related reading: Other Howard Simons articles

“ The industrial pulse of freight rates”” “ Nothing could be finer than being a data miner””
Active Trader, April 2009. Active Trader, September 2008.
Good trading ideas are often found away from the beaten Analysis of large moves in three markets has interesting
path — for example, the interaction between shipping rates, implications for system traders.
the yuan-dollar rate, commodity prices, interest rates, and
inflation. “ Putting the put-write right””
Active Trader, August 2008.
“ Bonds today, stocks tomorrow”” Looking for a different edge in the stock market? See what
Active Trader, March 2009. comparing the Buy-Write (BXM) and Put-Write (PUT) indices reveals
Dig into the stock-bond relationship to find out what it will about the path of least resistance in the stock market.
take for a new bull market in stocks to emerge.
“ Crush spreads in a biofuel age””
“ Gasoline, productivity, and inflation”” Active Trader, July 2008.
Active Trader, February 2009. Take a look at how the biofuel industry is changing the soybean
Higher prices can do two things: change demand and induce market crush spread.
new supply. Both of these occurred after mid-2008.
“ Energy stock movers and shakers””
“ Volatility and swap spreads”” Active Trader, June 2008.
Active Trader, January 2009. The surprising results of analyzing the connection between oil and
The relationship between swap spreads, the yield curve, and natural gas prices and the performance of energy stocks.
the term structure of fixed-income volatility offers valuable
insights into the direction of corporate bonds and stocks. “ TIPS, treasuries, and insurance””
Active Trader, May 2008.
“ REITs as macro indicators”” Do TIPS really have an advantage over regular T-notes and
Active Trader, December 2008. T-bonds?
Take a look at the sometimes surprising performance of REITs
over the past several years. “ Gold: Sound and fury, signifying nothing””
Active Trader, April 2008.
“ Deconstructing the commodity surge”” Gold has burst to new highs as the U.S. stock market and dollar
Active Trader, November 2008. have tanked, but don’t believe the easy explanations about the
Analysis of the factors driving the commodity boom offers yellow metal’s role as an inflation barometer or hedge.
insight regarding the appropriate way to approach futures
markets. “ Had enough of the dollar and stuff? ”
Active Trader, March 2008.
“ The flight-to-quality trade”” Analysis shows the relationship between the dollar and
Active Trader, October 2008. commodity prices isn’t what most people think.
The rule has seemingly always been: Don’t ever be short bonds
when stocks are getting clobbered. Is it still valid? “ Howard Simons: Advanced Currency Concepts, Vol. 1””
A discounted collection that includes many of the articles listed
here.
You can purchase and download past articles at http://store.activetradermag.com

www.activetradermag.com • May 2009 • ACTIVE TRADER

 TRADING
TRADING Strategies
Strategies

Fair value and the VIX futures

The relationship between the VIX futures and the underlying index can shed
light on the market’s expectations for future volatility and market direction.

BY RUSSELL RHOADS

O
ne of the unique features of the futures contracts on the Fair value
CBOE Volatility Index (VIX) is how they trade relative to One of the best illustrations of a Go to p. 70-71 for more
the VIX. The futures are cash-settled based on an opening futures contract with a fair value information about:
VIX calculation on expiration day. Until that time, the relationship is the S&P 500 index
relationship between the VIX and VIX futures may appear to be futures (SP). Before the stock mar- • At-the-money
nonexistent. This lack of an obvious quantifiable relationship ket opens each day, traders might • In-the-money
between the two stems from an interesting aspect of the VIX: turn on the TV to their favorite • Out-of-the-money
There is no method to replicate the price action of the VIX business channel, looking for the • Variance and standard
through buying a basket of securities. quote that shows where the S&P deviation
The VIX is calculated using the midpoint of the bid-ask spread 500 futures contracts are trad- • VIX
of a wide range of at-the-money and out-of-the money S&P 500 ing. An anchor might refer to the • Volatility
index options. These prices are fed into a formula that creates S&P 500 futures and say some-
a synthetic at-the-money S&P 500 index option contract that thing like, “Based on fair value we
expires 30 days from the moment the calculation was made. The should see the stock market open-
implied volatility from this synthetic option contract is what is ing the day up five points higher.” Of course, the anchor says this
quoted as the VIX. (If you have noticed the VIX tends to chop without explaining the fair value relationship between the S&P
around during the day rather than moving in small ticks, you are futures and the underlying market.
not imagining things: Because of the complexity of its calculation, A basket of securities can be purchased to replicate the perfor-
the VIX is updated every 15 seconds during the trading day.) mance of the S&P 500 index. This could be a position in all 500
The VIX calculation is a measure of 30-day implied volatility as stocks that comprise the index, or possibly a position in fewer
indicated by S&P 500 index options. However, it is not possible to stocks that will closely track the performance of the index. Since
isolate this volatility level through trading option contracts or any the index can be replicated, the ability to go long the index and
other instruments. Because the VIX cannot be replicated in a port- short S&P 500 futures contracts also exists. If a large buy order
folio, there is no tradable arbitrage opportunity; the relationship hits the S&P 500 futures and the contract price rallied to rise to
that keeps most futures contracts in line with an underlying market a certain premium relative to the S&P 500 index, a trader could
or instrument does not exist between the VIX and VIX futures con- theoretically take a long position in the index and short the
tracts. This relationship is commonly referred to as fair value.

www.activetradermag.com • • ACTIVE TRADER

 Trading Strategies

FIGURE 1: S&P 500 INDEX AND FUTURES futures. Similarly, if the S&P futures are trading at enough
of a discount to the index, the futures can be bought and
the index portfolio sold short. This relationship causes the
S&P 500 futures to stay within a certain price range around
the S&P 500 index.
Figure 1 illustrates this relationship. It is a one-minute
chart of the S&P 500 index and the September S&P 500
futures from Aug. 23, 2011. The blue line, which is this
higher of the two on most of the chart, represents the
S&P 500 index; the red line represents the futures. Notice
throughout the day the spread between the two is fairly
consistent — averaging around 2.50 points over the course
of the day. The spread was at 2.38 points at the beginning of
the day and 2.77 points at the close.
The absence of this type of relationship between the VIX
index and VIX futures results in the VIX futures sometimes
trading at a significant premium or discount to the VIX. At
This one-minute chart of the S&P 500 index (blue) and the expiration the VIX futures and index will converge, so when
September S&P 500 futures (red) shows the spread between the VIX futures are at a discount this may indicate traders
believe the index should be lower between the current date
the two is fairly consistent, on this day averaging around
and expiration. Conversely, if the futures are at a significant
2.50 points.
premium, this may suggest traders believe the direction of the
VIX will be higher between the current date and expiration.
FIGURE 2: VIX AND VIX FUTURES Figure 2 is a one-minute chart of the VIX and the Sep-
tember VIX futures (VXU11) from Aug. 23, 2011. On this
chart the higher blue line represents the index and the low-
er red line represents the futures contract. However, on this
chart the spread between the futures and index throughout
the day progressively narrows.
On Aug. 23 the VIX dropped from the previous day’s
close of 42.44 to 36.27, a loss of 6.17 points, while the
September VIX futures fell from 36.45 to 34.30, for a loss
of 2.15. Over the course of this one day the spread between
the two went from 5.99 to only 1.77 points (the index
being at a premium to the futures). This is an exaggerated
example; the VIX had fallen approximately 15 percent from
the previous day’s close, which is a relatively large move
for the index. However, this is a good example of how the
futures and index can have vastly different price movements
in the same day.
The spread between the VIX and VIX futures behaves
differently than the one between the S&P 500 index and S&P Mean reversion
The implied volatility of the market is a nontrending
500 futures shown in Figure 1.

www.activetradermag.com • • ACTIVE TRADER

 Trading Strategies

FIGURE 3: VIX VS. FUTURES, AUG. 22, 2011 measure that is “mean reverting” over time. When the VIX
is at lower-than-average levels, the expectation will be for a
rise in the index, while when the VIX is at higher levels the
expectation is for a lower VIX over time. During times of
higher volatility the spread between the index and futures
can widen tremendously because tumultuous market condi-
tions are expected to calm over time.
These expectations can be depicted by a graph like the
one in Figure 3, which shows the level of the VIX and active
VIX futures contracts (represented on the horizontal axis) at
the close of Aug. 22, 2011. This was a time of higher-than-
normal volatility for the stock market; the VIX closed above
40 for a few days. Over that period, all the VIX futures
contracts were trading at prices well below the index price.
This could be considered an indication traders believed the
market’s volatility would subside over time.
Figure 4 depicts the closing VIX and VIX futures prices
At this time of higher-than-normal volatility, all the VIX futures on Aug. 23, 2010. On this date, about a year prior to the
contracts were trading well below the index price — an first chart, the VIX closed at 25.66 and all actively traded
indication traders believed the market’s volatility would futures contracts were trading at a premium to the VIX in-
subside over time. dex. This may be considered an indication traders believed
higher volatility was on the horizon.
More often than not, the VIX futures trade at a premium
FIGURE 4: VIX VS. FUTURES, AUG. 23, 2010 to the underlying index. With more time left until expira-
tion, there is more of a possibility of some sort of unusual
volatility event that can result in a spike in the VIX and
higher VIX futures prices. This relationship between the fu-
tures and index has dual implications for trading. There are
potential trading strategies that can be based on the price
relationship between the futures and the index, or even the
various futures contracts. Also, because the VIX futures are
priced relative to the index based on market expectations,
there is the possibility of developing a stock market indica-
tor.

Trading considerations
If you are considering trading VIX futures contracts, it is
of the utmost importance to understand the relationship
between the futures and the index. The price action from
Aug. 23, 2011, is a prime example. The VIX dropped
In this case the successive futures contracts were trading at almost 6 points, but the front-month futures declined only
a premium to the VIX, suggesting traders believed higher about 2 points. Being short a VIX futures contract might be
volatility was on the horizon. a little disappointing, considering the much larger move in
the index.

www.activetradermag.com • • ACTIVE TRADER

 Calculating VIX futures fair value
There is a complex fair value relationship that exists
between the markets and VIX futures pricing. The value
of a VIX futures contract can be derived by pricing the
forward 30-day variance that underlies the settlement
price of VIX futures. The forward price of the 30-day
variance is determined by the pricing of S&P 500 index
option contracts that expire 30 days after the futures
expiration date.
If you are interested in digging deeper into this
relationship, visit cfe.cboe.com.

Also, always keep in mind that at expiration the VIX futures outlook for the VIX can be considered an outlook for the direc-
contract price will converge with the index. Referring again to tion of stocks. Therefore it might be possible to apply the relative
Figure 4, the VIX closed at 25.66 and the September 2010 VIX price of VIX futures to the VIX as an indicator to trade the direc-
futures closed at 29.00. If the outlook is that the VIX will be tion of the stock market. At a minimum the futures prices relative
stagnant until September expiration, a short position can be to the index can be considered an indication of where traders
taken in the September VIX futures contract. If this outlook is believe volatility is headed over the near term. ◆
correct and September 2010 VIX settlement is 25.66 (unchanged
from the Aug. 23 close), the short futures position would realize
a 3.34-point profit.
FIGURE 5: S&P 500 AND VIX
Market indicator
VIX futures trade at a discount or premium to the VIX
based on trader expectations. An expectation of higher
volatility can result in the futures trading at a premium,
while expectations of lower volatility can result in the
futures trading at a discount.
The volatility of S&P 500 index options tends to fall in
rising markets and rise in falling markets. Because of this
relationship the VIX and S&P 500 index tend to move
in opposite directions. Figure 5 highlights the inverse
relationship between the S&P 500 and the VIX over a
seven-week period in July and August 2011. The blue line
represents the daily closing prices of the S&P 500 while
the red line shows the daily changes for the VIX.
Another way to consider this relationship is, if the index
and VIX futures move in opposite directions, then an The VIX (red) tends to move in the opposite direction of the
opinion about the direction of the S&P 500 would also S&P 500 index (blue).
be a projection of the price changes for the VIX. Also, an

ACTIVE TRADER • December 2011 • www.activetradermag.com

 TRADING
TRADING Strategies
Strategies

Trading with the VIX Put/Call Ratio

There are several put/call ratios, and they all have different characteristics.
Find out how the VIX Put/Call Ratio can signal trade opportunities in the stock market.

BY RUSSELL RHOADS

A
put/call ratio measures the number of put options traded index option volume, and VIX index option volume — all of
vs. the number of call options traded in a day. The ratio is which are tracked on the exchange’s website (www.cboe.com)
most commonly viewed as a gauge of demand for bearish and have downloadable data in spreadsheet form.
protection against a downturn in equities. On days a put/ The common use of the different option contracts that form the
call ratio is above 1.00, traders and investors are buying more basis of these put/call ratios varies a bit. Equity options tend to
put options than call options — a possible indication market have more balanced put and call volume, while SPX options con-
participants are more focused on market weakness. On days sistently have more trading in puts than calls; VIX option volume
a put/call ratio is below 1.00, it means call option volume has depends on the specific market events occurring at a given time.
exceeded put option volume and the market is buying calls in As a result, the put/call ratios based on these options exhibit
anticipation of higher stock prices in the near term. unique characteristics.
The Chicago Board Options Exchange publishes daily data for
several put/call ratios, shown in Table 1: total exchange volume, Comparing put/call ratios
index option volume, equity option volume, S&P 500 (SPX) Figure 1 shows the Equity Put/Call Ratio for the third quarter
of 2011 along with the S&P 500
Index. The put option volume
TABLE 1: PUT/CALL RATIOS
outweighs the call option volume
Ratio Based on
in only a couple of instances;
CBOE Total Exchange Put/Call Ratio Volume of all puts and calls the spikes above 1.00 in August
CBOE Index Put/Call Ratio Volume of stock index puts and calls occurred on very bearish days in
CBOE Equity Put/Call Ratio Volume of individual equity puts and calls the stock market and mark the
CBOE Volatility Index (VIX) Put/Call Ratio Volume of VIX puts and calls two roughly equal troughs near
the bottom of the summer sell-off.
CBOE S&P 500 (SPX) Put/Call Ratio Volume of S&P 500 puts and calls
Most of the time, there is more
The Chicago Board Options Exchange (CBOE) calculates and publishes several put/
volume in calls than in puts in
call ratios on a daily basis.

www.activetradermag.com • January 2012 • ACTIVE TRADER

 FIGURE 1: EQUITY PUT/CALL RATIO

equity options. But when there is heightened

concern about protecting against a downside
move in individual stocks, there is usually an
increase in put volume relative to call volume
and the Equity Put/Call Ratio rises.
Figure 2 shows the SPX Put/Call Ratio for
the same time period. In this case the put/
call ratio line never dips below 1.00, which
means during the third quarter there was not
a single day when S&P 500 call volume ex-
ceeded put volume. Although this character-
istic is in stark contrast to the Equity Put/Call
Ratio in Figure 1, it is fairly typical in the
SPX Put/Call Ratio because institutions often
buy SPX put options to hedge their equity
portfolios. However, much like the Equity The Equity Put/Call Ratio (bottom) pushed above 1.00 (indicating higher
Put/Call Ratio, this ratio moves to higher put volume than call volume) only twice during the third quarter.
levels when there are increased concerns a Data source: www.cboe.com

bear market might be on the horizon.

Figure 3 shows the VIX Put/Call Ratio. FIGURE 2: SPX PUT/CALL RATIO
Here, the line is mostly below 1.00, but a
very large spike formedd toward the end
of the quarter, when put volume exceeded
call volume by more than three to one. It is
the nature of the VIX to spike during times
of market uncertainty and then revert to its
mean during more stable equity market con-
ditions. The trading activity in VIX options
reflects this recurring pattern, where demand
for puts spikes relative to calls when the
overall stock market sells off.
Figure 4 compares the S&P 500 and VIX
during the third quarter. Notice there is basi-
cally an inverse relationship between daily
price changes in the S&P 500 index and VIX.
In fact, the VIX closes down on approximate-
ly 75 percent of the days the S&P 500 closes
higher than the previous day; the VIX closes
higher on around 75 percent of the days the Because institutions often buy SPX put options to hedge their equity
S&P 500 index closes lower. Between Jan. 1, portfolios, the SPX Put/Call Ratio is typically above 1.00.
Data source: www.cboe.com

ACTIVE TRADER • January 2012 • www.activetradermag.com

 Trading Strategies

2000 and Sept. 30, 2011, the S&P 500 closed

FIGURE 3: VIX PUT/CALL RATIO
down by 2.5 percent or more exactly 100
times. The VIX rallied on 99 of those days,
and its move was greater on a percentage
basis than the drop in the S&P 500 94 times:
The average S&P 500 drop on those 100 days
was 3.74 percent, while the average gain in
the VIX index was 14.7 percent.
VIX options commenced trading in 2006,
and the daily volume in these contracts often
approaches 1 million. Unlike any other put/
call ratio, the VIX Put/Call Ratio spikes when
there is excessive bullish sentiment. More VIX
put volume than call volume indicates market
participants are expecting a lower VIX, which
implies a rebound in the stock market.
The point behind taking a look at what
the VIX does on truly bearish days in the
stock market is to highlight what VIX option
The VIX Put/Call Ratio is mostly below 1.00, but it formed a big spike
traders do when concerned about the future
toward the end of the third quarter.
Data source: www.cboe.com direction of the market: They purchase call
options. Therefore, increased call volume
FIGURE 4: S&P 500 AND VIX occurs when traders believe there is a reason
to be concerned about the market turning
bearish. This sort of bearish move would also
need to occur in the form of a dramatic drop
in the stock market for the VIX to react with a
spike to the upside.
Let’s look at how this information can be
incorporated in a trading approach.

VIX Put/Call Ratio strategies

Applying the VIX Put/Call Ratio as a short-
term buy signal for the stock market pro-
duces some interesting results. We’ll look at
two simple methods, tested using data from
Jan. 1, 2007 to Sept. 30, 2011, that go long
the S&P 500 based on the closing VIX Put/
Call Ratio value.
The first approach involves taking a long
position in the S&P 500 for one day when
VIX put option volume exceeds VIX call op-
There is a general inverse relationship between daily price changes in the
S&P 500 index and VIX. tion volume — that is, the VIX Put/Call Ratio
Source: TradeStation closes above 1.00. When the put volume ex-
ceeds the call volume it indicates traders may

www.activetradermag.com • January 2012 • ACTIVE TRADER

be expecting the VIX to trade lower, which FIGURE 5: VIX PUT/CALL RATIO SIGNAL 1
implies expectation the stock market is going
to move higher over the near term.
There have been 1,197 trading days be-
tween Jan. 1, 2007 and Sept. 30, 2011, and
the VIX Put/Call Ratio has closed above 1.00
on 166 of them. If a trader bought the S&P
500 on the close of days the ratio was above
1.00 and held until the next day’s close, they
would have profited on just more than 56
percent of trading days. Overall, the S&P
500 closed higher on about 54 percent of all
days in the period.
However, more impressive than the slight
improvement in winning percentage is the
net gain from the system relative to simply
being long the S&P 500 index. Over the test
period the S&P 500 lost 286.88 points while
this simple signal gained 235.94 points. Fig-
ure 5 compares the S&P 500 (blue) to a long Although the S&P 500 (blue) lost 286.88 points during the analysis
position based on the VIX Put/Call Ratio period, the first VIX Put/Call Ratio signal (red) gained 235.94 points.
long signal (red). Data source: www.cboe.com
The second approach is a little more com-
plex. If within the previous three days the FIGURE 6: VIX PUT/CALL RATIO SIGNAL 2
VIX Put/Call Ratio has closed below 1.00 one
day and then closed above 1.00 the follow-
ing day, go long the S&P 500 for a one-day
hold. There are periods where this type of
activity has produced multiple (overlapping)
signals — that is, times when a signal is trig-
gered when a previous long signal is still in
effect. When this occurs, the second signal is
not acted upon; only the long position from
the first signal continues to be maintained.
This approach improves a bit on the first
method’s performance, returning 382.90
S&P 500 points during the analysis period.
Figure 6 shows the performance of the S&P
500 vs. this more complex trade screen.
Although neither of these approaches
would work as a pure, stand-alone system,
their positive results suggest the VIX Put/
The second VIX Put/Call Ratio approach returned 382.90 S&P 500 points
Call Ratio is a useful indicator for identifying during the analysis period.
trade opportunities. ◆ Data source: www.cboe.com

ACTIVE TRADER • January 2012 • www.activetradermag.com

 TRADING
TRADING Strategies
Strategies

The VIX-neutral
option trade
When trading VIX
options, it’s the VIX futures
price that counts.

BY RUSSELL RHOADS, CFA

O
ptions on the CBOE Volatility Index (VIX) have special options are priced throughout their lifetimes using the futures
properties because of their relationship to the VIX itself but settle in terms of the index.
and the VIX futures (VX). While the most appropriate This unique relationship gives traders the opportunity to
underlying price of a VIX option is provided by the VIX structure a position that can benefit from a neutral outlook on the
futures contract that shares the same expiration date, VIX options VIX, capitalize on the passing of time, and limit risk to a long-
are settled at expiration using the VIX. In other words, VIX options position.
Before detailing this long VIX options
FIGURE 1: TYPICAL VIX CURVE strategy, let’s further explore the relationship
between the VIX and VIX futures.

VIX futures and the VIX

Although the strategy is executed using VIX
option contracts, the premise is based on
understanding how VIX futures are valued
relative to the VIX.
Typically, VIX futures contracts trade at a
premium to the VIX — that is, they trade at
a higher price. This premium varies based
on the market’s expectation of future implied
volatility, as well as how much time remains
The more distant the contract, the larger the premium at which the VIX until the futures expire. As a specific contract
futures usually trade to the VIX.
month’s expiration approaches, the futures

www.activetradermag.com • August • ACTIVE TRADER

 FIGURE 2: TRADING THE VIX CURVE

The trade idea, which uses a neutral forecast for the VIX, capitalizes on
the VIX futures converging (lower) to the VIX value by expiration.

price will begin to converge with the underlying (cash) VIX price. to the VIX (in which case the shape of the curve will invert), but
Typically, the more time until expiration, the wider the VIX- the curve usually returns to its “normal” shape over time. The
futures spread will be, and the slower the day-to-day gravitation VIX curve is in this normal condition, shown in Figure 1, ap-
of the futures toward the index. proximately 80 percent of trading days.
Figure 1 shows the VIX curve based on closing prices on This price spread, with the VIX futures trading at a premium to
Jan. 20, 2012. The first data point represents the VIX, while the index, provides the basis for our trade strategy, which consists
the remainder are the closing prices of successive VIX futures of purchasing a put option on the VIX.
contracts. Note the more distant the contract month, the greater
the premium at which the futures trade to the VIX. This is a fairly VIX option pricing
typical shape for the VIX curve. Options on the VIX are cash settled based on a specific calcula-
However, the shape and steepness of the curve can change. tion of the VIX expiration date (for more details, go to
There are instances when the VIX futures will trade at a discount www.cboe.com/products/indexopts/vixoptions_spec.aspx). Like
the majority of index options, VIX options use “European-style”
settlement, which means they can be exercised only at expira-
TABLE 1: DECEMBER VIX OPTIONS
tion. (“American-style” options can be exercised at any time up to
Put option Bid Ask
expiration.) This European settlement results in some interesting
Dec 28 5.10 5.20
pricing characteristics for VIX options.
Dec 29 5.70 5.80 Although they are cash-settled against the VIX, the best way
Dec 30 6.30 6.40 of valuing VIX options during their lifetimes is by using the VIX
These December VIX put option prices reflect futures contract with the same expiration date (e.g., using the
the VIX trading at 20 and the December VIX
futures trading at 30.

ACTIVE TRADER • August 2012 • www.activetradermag.com

 Trading Strategies

September futures for the September options). For example, let’s appear mispriced if the index (rather than the futures contract)
say the VIX is at 20 and a VIX futures contract is trading at 30. was being referenced as the underlying security.
A put option for an underlying security trading at 30 is going Because VIX options are specifically referred to as “index op-
to have a different price than a put with an underlying security tions,” this pricing characteristic can cause some confusion. For
trading at 20. As a result, a VIX put option in this situation could
example, the December VIX put option prices in Table 1 (p. 41)
reflect a situation that would arise if the VIX
FIGURE 3: LONG PUT PROFIT PROFILE was at 20 and the December VIX futures were
trading at 30. (There are many more strikes
than the ones in this table, but these three
are a good way of leading up to the trading
strategy.)
Consider the VIX Dec. 30 put, which could
be purchased at the ask price of 6.40. A
novice VIX trader who saw the VIX at 20 and
the Dec. 30 put trading at 6.40 might believe
the option is mispriced. The normal expecta-
tion is that a put option with a strike price of
30 on a stock trading at 20 would be worth
As the price of the VIX futures declines, the trade’s profit increases. at least 10 (its “intrinsic value” — the amount
the option is in the money). A 30-strike put
FIGURE 4: TRADE LIFETIME that could be purchased for $6.40 would
present a significant arbitrage opportunity:
The put could be purchased for $6.40 and
100 shares of the stock could be purchased
at $20. The right to sell 100 shares at $30
would then be exercised. The result would be
a stock purchase at $20 and a sale at $30 for
a $10 profit — $3.60 more than the $6.40
cost of the 30-strike put. (Unlike VIX options,
which can be exercised only at expiration,
options on stocks are American style and can
be exercised at any time until expiration.)
Now let’s see how all these characteristics
lend themselves to a simple VIX-options
trade.
As April expiration approached, the spread between the futures and the
VIX narrowed rapidly.

www.activetradermag.com • August • ACTIVE TRADER

 Related Reading
Russell Rhoads articles
Long VIX put trade
Figure 2 (p. 41) revisits the VIX curve from Figure 1 and illus- Trading with the VIX put/call ratio
trates an interesting trade using a VIX put. The basic idea is the Active Trader, January 2012
VIX will be fairly neutral over the next three months and settle There are several put/call ratios, and they all have
different characteristics. Find out how the VIX Put/
near 18.25 (the Jan. 20th VIX price shown in the chart). If this
Call Ratio can signal trade opportunities in the stock
scenario played out, the result would be an April VIX settlement market.
at 18.25. The chart highlights the April futures price and the
underlying VIX value. Fair value and the VIX futures
Active Trader, December 2011
Although the strategy is based on a forecast for a relatively stag-
Analysis of how the relationship between the VIX
nant VIX over a three-month period, the trade consists of buying futures and the underlying index can shed light on the
an April put because this forecast also means the April VIX market’s expectation for future volatility and market
futures contract, which closed at 25.30 on Jan. 20, would con- direction.
verge to the expected April settlement price of 18.25. Checking
Quantifying the black swan
prices, the most attractive choice is the VIX April 30 put, which Active Trader, October 2011
is trading at 7.00. With an April VIX settlement of 18.25, the VIX Analysis of the CBOE SKEW index and how it can
April 30 put would be worth 11.75 at expiration. Through cash signal trading opportunities.
settlement, long positions in the VIX April 30 put would receive
Other articles
a payment of $1,175 (11.75*$100). The option cost $700, so the
result in this case is a $475 profit. Hedging with VIX options
Figure 3 shows the payoff for the trade, based on the neutral Active Trader, July 2008
VIX outlook. This profit diagram is typical for a long put at expi- The VIX’s unique characteristics give its options certain
advantages in constructing a stock-market hedge.
ration. As the price of the VIX futures declines (moves to the left),
the trade’s profit increases. Note the April VIX futures and VIX VIX options
prices are also displayed. The location of the two emphasizes the Futures & Options Trader, December 2007
unique nature of VIX options. The pricing is based on one price Despite appearances, VIX options behave differently
from other options.
while settlement is based on the other.
The actual outcome of this trade was fairly close to the projec- Tracking VIX swings
tion. Before checking out the final outcome, let’s look at how the Active Trader, January 2006
April VIX futures and VIX traded over the life of this trade. The VIX has been a widely discussed stock market
barometer, but how reliably does it identify market
Figure 4 compares the two from Jan. 20, 2012 through April ex-
turning points? This study turned up a few surprises
piration on April 18, 2012. The April VIX settlement was 19.06, in analyzing how the S&P 500 tracking stock (SPY)
so the option was worth 10.94 at expiration. The relationship responded to VIX highs and lows.
between the April VIX futures contract and the VIX shown here is
The VIX and market capitulations
fairly typical. As April expiration approaches, the spread between
Active Trader, May 2009
the two narrows. One of the more interesting sections on this It takes more than a VIX spike to identify a market
bottom.

ACTIVE TRADER • August 2012 • www.activetradermag.com

 Trading Strategies

chart occurs before and after March 19. During this period, the of 1.60 if the VIX settled at 17.40
VIX was in a trading range, but the April VIX futures fell sharply, in August. However, in late July
converging to the index as April became the front-month futures and early August, the VIX started
contract. to rise dramatically as the stock
market sold off, and the result was
Assessing the risks an August VIX settlement of 32.73;
Finally, a word of caution about what can go wrong with this the VIX Aug. 25 put expired Go to p. 68 for more
trade. Figure 5 shows an example of how the VIX behaves when worthless. The result of using this information about:
it experiences a rapid move to the upside, as occurred in August strategy would have been a total • Exercise
2011 when the stock market came under pressure because of re- loss of $600 (the cost of the put). • In the money
cession fears in the U.S. and renewed worries about the European The interesting aspect of this • Intrinsic value
debt crisis. trade is that we are able to benefit • Out of the money
• Strike price
On May 20, 2011, the August VIX futures were trading at from the passage of time while • Volatility
21.50 and the VIX was at 17.40. Based on a similar outlook to only incurring the risk of a long • VIX
the previous trade, we might have bought a VIX put. The VIX option contract. The alternative
Aug. 25 put was trading at 6.00, which would result in a profit to buying a put would be sell-
ing a call that would benefit from the price
FIGURE 5: MOVING AGAINST EXPECTATIONS
convergence of the April VIX futures to the
index along with the time decay of a short
call option. However, the risk incurred for a
short call like this is theoretically unlimited.
With the price spikes that occur in volatility,
the risk taken on with a short call should not
be taken lightly.
The independent price movements of VIX
futures and options relative to the underly-
ing VIX can be a bit confusing at first glance.
However, the opportunity to benefit from a
neutral outlook on the VIX through a long
put option position can make getting past the
initial confusion worth the effort. ◆

Stock-market declines, such as the one that occurred in July-August 2011,

can send the VIX quickly and dramatically higher.

ACTIVE TRADER • August 2012 • www.activetradermag.com

 TRADING
TRADING Strategies
Strategies

Go to p. 64 for more
information about:

• Fair value
• Sharpe ratio
• Variance & standard
deviation
• Variance swap
• VIX
• Volatility

The VIX basis trade

The unique nature of VIX futures offers a simple but significant trade opportunity.

BY EUAN SINCLAIR

T
FIGURE 1: S&P 500 AND VIX rading opportunities take various
shapes — temporary price
dislocations, long-term trends or
biases, and seasonal factors, to name
a few. In the case of the VIX futures (VX),
the nature of the underlying itself presents a
unique trade opportunity.
The Chicago Board Options Exchange
(CBOE) Volatility Index (VIX) — sometimes
called the “fear index” because it tends to rise
during periods of market tension or panic —
uses the prices of the out-of-the-money S&P
500 options to create a composite volatility
number that represents the option market’s
best guess of the S&P’s volatility over the next
30 days (Figure 1).
The VIX is an interesting trading tool for
several reasons. It can be used as a confirming
signal in price-based strategies, in a simi-
Derived from options implied volatility, the VIX represents the market’s
lar way to volume, and it is one of the few
“best guess” at S&P 500 volatility over the next 30 days.
products that are reliably and significantly

www.activetradermag.com • y • ACTIVE TRADER

 Related Reading

The VIX Futures Basis:

Evidence and Trading Strategies
by David Simon and Jim Campasano (2012).
negatively correlated to the stock market. Also, volatility has been Available at www.ssrn.com (SSRN-id 234797).
widely studied. The academic research paper repository
Books by Euan Sinclair:
www.ssrn.com has been adding approximately 500 new papers
a year with the word “volatility” in the title. Finally, volatility is Volatility Trading (+CD ROM, Wiley Trading, 2008)
much more predictable than price. The updated second edition of this book, which offers a
quantitative model for measuring volatility in order to gain
Futures valuation and the VIX an edge in option trading, is slated for publication in April
VIX futures were first listed in March 2004 (VIX options followed 2013.
in 2006). At expiration, the VIX futures are settled against the
cash VIX price. Option Trading: Pricing and Volatility Strategies and
Techniques (Wiley Trading, 2010)
The standard way to value futures contracts is through “cash-
and-carry” arbitrage: This assumes we borrow money to buy the Articles by Euan Sinclair:
underlying instrument, store it, and then deliver it to the buyer of
the futures contract. Because this position has the same cash flow The real effect of stops
as the future, it must also have the same value. Active Trader, November 2008
Although this model works well when the underlying in- There are real costs to using stop-loss orders, although
strument is a financial product, such as a bond or a stock, it is most traders don’t understand how high the price can
sometimes be. This article analyzes the benefits and
somewhat less relevant for many commodities. In some cases it
drawbacks of using stops.
might not be possible to buy the underlying commodity because,
for example, the corn might still be in the ground or the cows Equity options and earnings
may not yet be born. Further, storing commodities can be risky: Active Trader, June 2012
Grains can rot and cattle can get sick. There are two basic ways to trade successfully: Develop a
And in other cases the cash-and-carry model fails completely. complete model of the game you’re interested in, or look
For example, electricity futures can’t be priced this way because for special situations where you might have an edge. This
electricity is basically impossible to store. The model also fails article explores how to trade equity options before and
during quarterly earnings announcements.
for the VIX because it is impossible to trade the cash VIX index.
Basically, this is because the VIX is defined as the square root of a
variance swap, and the square root cannot be “statically repli-
model for the future evolution of implied volatility. VIX futures
cated.” This means it’s not possible to do a single trade that “locks
fair-value calculations are model-dependent.
in” this value. For example, we can statically hedge a bond future
This also means the relationship between the VIX cash and
by trading a bond once. However, if we were to hedge an option
VIX futures (the “basis”) is fundamentally different than for many
we need to “dynamically replicate” it by continually trading the
other products. And this leads to a trading opportunity.
underlying because the option has exposures that change as the
underlying moves or time passes.
Because the VIX index cannot be statically replicated, the usual
VIX basis as a predictor of the futures
In their paper “The VIX Futures Basis: Evidence and Trading
futures pricing relationship based on cash-and-carry arbitrage
Strategies” (see “Related Reading”), David P. Simon and Jim
does not hold. Instead, VIX futures need to be priced using a

ACTIVE TRADER • January 2013 • www.activetradermag.com

 Trading Strategies

FIGURE 2: VIX FUTURES FORWARD CURVE

The mean reversion of implied volatility leads to one of

the most common characteristics of the futures term struc-
ture. Generally, when the VIX is high the futures curve will
slope downward, and when the VIX is low the curve will
slope upwards (Figure 2). However, this phenomenon is
far from universal. For example, when the VIX was below
20, the futures curve was downward sloping 22% of the
When the VIX is high the futures curve will generally slope time and when the VIX was between 40 and 50 the curve
downward, while it typically slopes upward when the VIX is was upward sloping 46% of the time.
low. However, these characteristics do not always hold. Such seemingly anomalous behavior can lead to some of
the best trading opportunities.
Campasano presented evidence that the basis can be used to pre-
dict VIX futures prices: If the futures are trading above the cash Testing a VIX-basis strategy
VIX, the futures will tend to fall, and if the futures are trading The raw VIX-basis trade is:
below the cash VIX, they will tend to rise.
This is not what is predicted by the rational expectations hy- 1. Sell short one VIX futures contract if the front-month
pothesis of economics, which hold the futures prices are unbi- contract price is above the cash VIX value and the expected
ased predictors of the future value of the cash VIX. If this were daily convergence between the cash and futures is greater
true, the basis would have no predictive power. than 0.1 VIX points. (For example, if the basis was 2 points
The Simons and Campasano study analyzed the VIX from with 10 trading days remaining, we would expect the aver-
2006 to 2011, a period during which the cash VIX averaged 23.7 age daily rate of convergence to be 0.2 points.)
and the two nearest futures contracts averaged 23.8 and 24.4, 2. Buy one VIX futures contract if the front-month contract
respectively. The VIX high during this period was 81.1 and its price is below the cash VIX value and the expected daily
low was 9.9. convergence is greater than 0.1 VIX points.
Because volatility reverts to the mean, the range for the futures 3. Hold the trade for five days.
prices was smaller: The front month (the contract nearest to
expiration at any given time) had a range of 10.3 to 69.0. The Table 1 summarizes the results of this strategy. Most of the
front-month basis averaged 0.66 (meaning the futures price was profits come from realizing the anticipated convergence — that
higher than the cash by this amount), with a low decile of -1.40 is, the futures coalesce to the cash VIX price at the expected rate.
and a high decile of 2.01. The average trade was positive for both longs and shorts, and
60.48% of trades were profitable overall.
TABLE 1: VIX-BASIS TRADE SUMMARY Of course, there’s much more to consider than average profits
when assessing a trading strategy. The first thing to do is to look
Shorts Longs
at the equity curve as a function of time (Figure 3). This chart
Mean profit $656 $1,040
tells us a couple of interesting things. First, the strategy was very
Winners/losers 53/29 22/20 profitable during the equity market turmoil of 2008. Second, the

www.activetradermag.com • January 2013 • ACTIVE TRADER

 TABLE 2: HEDGED VIX-BASIS TRADE SUMMARY
trade has been consistently profitable since, albeit less dramati- Shorts Longs
cally so. Mean profit $539 $908
The strategy seemed to struggle during the early part of the Winners/losers 55/27 27/15
test period. Obviously this would be cause for greater concern if
it had occurred at the end of the period, but it is worth knowing hedged trade are shown in Table 2.
the strategy can go through long periods of mediocre perfor- As we might expect, hedging slightly decreased the profitability
mance. of the average trades, but smoothed the results by increasing the
winning percentage, particularly for long trades, which increased
Adjusting the strategy from 52.38% to 64.29% (the overall winning percentage in-
Despite its net-positive performance, it’s clearly possible to creased to 66.13%).
improve this strategy. Both the 0.1-point per day convergence To indicate how robust this idea is, another test was performed
rate and the five-day holding period could be changed. Possibly that simply assumed a single hedge ratio — specifically, 500
they could be made adaptive or tied to the VIX level. Instead of a shares of the S&P 500 ETF (SPY) for every VIX future — was
time-based holding period, it might be better to use profit targets
and stops to exit trades.
These ideas and other modifications to mean- FIGURE 3: VIX BASIS TRADE EQUITY CURVE
reverting strategies have been discussed in
numerous books and articles. The bottom line
is the specifics of the trading strategy aren’t as
important as understanding the phenomenon
itself. This makes it relatively easy to find ways
to trade it. With this in mind, there are a couple
of other features of the trade to appreciate.
As previously mentioned, there is a strong
negative correlation between volatility changes
and S&P 500 returns. This means while we are
trying to collect the convergence premium, we
will be exposed to substantial risk if the VIX
moves, and the largest of these moves will be
driven by the movement in the stock index.
Simon and Campasano also discussed a
version of the trade that hedged this risk by
purchasing S&P futures when long VIX futures
and selling S&P futures when short VIX futures. After struggling early in the test period, the raw basis-trade strategy
was very profitable during the 2008 equity market collapse and
They estimated the hedge ratio by regressing VIX
consistently profitable afterward.
returns on the index returns. The results of this

ACTIVE TRADER • January 2013 • www.activetradermag.com

 Trading Strategies

FIGURE 4: HEDGED VIX BASIS TRADE EQUITY CURVE

The hedged strategy was superior to the unhedged version on a risk-

adjusted basis. Hedging slightly decreased average trade profitability,
but it also smoothed the results by increasing the winning percentage.

good enough to use for every trade. Figure 4 shows the equity A distinctive tool
curve for this crudely hedged trade. Unlike most products, the VIX basis is a good predictor of the fu-
We can estimate a Sharpe ratio by dividing the daily dollar ture’s prices. The analysis suggests a very simple trading strategy
P&Ls by their standard deviation, and then annualizing the result can profit from the convergence of the VIX futures toward the
by multiplying by the square root of the number of trading days cash VIX. This effect is both statistically and financially significant
in a year. This isn’t a true Sharpe ratio because it ignores the risk- and seems to hold across a wide range of volatility regimes. It can
free return and uses dollar profits rather than returns. However, be traded directly or combined with other futures and options
it’s a very useful “dirty” estimator because we don’t have to worry strategies. It might also be worthwhile to explore the idea’s poten-
about account size or funding requirements, which can vary from tial as a predictor for the direction of the S&P, taking advantage
trader to trader. of the negative correlation between price and volatility.
In terms of this estimated Sharpe ratio, the hedged trade And now that we know the VIX cash-futures basis contains
clearly dominates, with a ratio of 2.2 compared to the unhedged valuable information, we could extend this idea by looking at the
trade’s ratio of 1.55. This is important because if a trade’s returns entire term structure of the futures. Is the effect more pronounced
are smooth enough, we can employ leverage to adjust the returns when the curve is steeper? Ultimately, the most important thing
higher. Another important point is that any hedge size from 400 to take away from this analysis is that the effect is real and sig-
to 1,100 shares resulted in a Sharpe ratio above 2.0, so the trade nificant. Once we know that, we can utilize many approaches to
is fairly robust. The “optimal” hedge size was 760 shares, which monetize it.◆
produced a ratio of 2.42. In practice, the hedge size would be
adjusted periodically.

www.activetradermag.com • January 2013 • ACTIVE TRADER