Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Note: The following article is an excerpt from the Third Quarter 2010 Letter to Investors from Artemis Capital
Management LLC published on September 30, 2010.

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 2
520 Broadway, Suite 350
Santa Monica, CA 90401
(310) 496-4526 phone
(310) 496-4527 fax
www.artemiscm.com
c.cole@artemiscm.com

Unified Risk Theory - Correlation, Vol, M3, and Pineapples

The summer of investor discontent was
VIX Index Surface - Third Quarter 2010
made glorious by the sun of quantitative easing
as the equity markets finished their best
September since 1939, largely due to
expectation for further monetary stimulus.
Earlier this month the Federal Reserve signaled
its intention to renew its Treasury bond
purchase program (known as Quantitative
Easing 2 or QE2) sending the markets up, the
dollar down, and our trading partners into a
frenzy. In this brave new normal our monetary policy has somehow managed to make any asset yielding above 1% look
sexy while sparking a global race to depreciate every globally traded currency. While the noble intention is stabilization
and job growth the Federal Reserve is inadvertently throwing the global monetary world into complete disarray. The
correlations (implied and realized) of asset classes are at multi-decade highs negating the benefits of diversification and
mystifying even the most seasoned investors. In addition volatility surfaces steepened significantly over the quarter with the
high volatility skew emblematic of a growing distrust of the prevailing equilibrium. There is a definitive connection
between current monetary policy, unusually high correlations, and the historically steep volatility surface. Even as the
domestic equity markets rise, the volatility and correlation markets are flashing signals of increased systemic risk seething
under an innocuous surface. The following letter uses detailed quantitative analysis to make this case, but as you will see,
the simple metaphor of selling pineapples in a Hawaiian farmer's market may suffice to understand why risk is increasing
exponentially in today's global economy.
34
32
30
28
26
24
22
20

1-Jul-2010
13-Jul-2010
22-Jul-2010

2-Aug-2010

11-Aug-2010

20-Aug-2010

31-Aug-2010

10-Sep-2010

21-Sep-2010

Rise of the Correlations: The paradox of the economic recovery is a phenomenon whereby different asset classes are
moving in sync at highest levels in decades, making diversification futile for professional and retail investors alike.
Correlation is a statistical measurement used to determine how closely the prices of different securities move in tandem and
is directly related to volatility. Correlations are an important tool to test our trust in the market simulacrum because they
provide a way to measure randomness. High correlation in markets can be described as a representing a lack of randomness
in the price discovery mechanism. When markets become less random, paradoxically, they become riskier but also
attractive for arbitrage opportunities.

www.artemiscm.com

42.5

Risk-On/Risk-Off - Average 16 Month Correlations - 1972 to Present

S&P 500 Index / 10yr UST Yield / CRB Commodity Spot Index / Swiss Franc to USD

Average 16 Month Correlation

32.5

22.5

12.5

2.5

-7.5

-17.5

2009
2008
2007
2005
2004
2003
2001
2000
1999
1997
1996
1995
1993
1992
1991
1989
1988
1987
1985
1984
1983
1981
1980
1979
1977
1976
1975
1973
1972

Correlations typically rise sharply with volatility during

financial crashes (like in 2008) as investors panic and sell all
assets resulting in less information being embedded in prices.
Oddly during the recent recovery professional money managers
are perplexed as to why asset classes are dancing to a simplistic
high-correlation tune called: Risk-On / Risk-Off. Risk-On, is the
beat whereby economic expectations are perceived as good and
investors flock to equities, commodities, emerging markets, and
sell the dollar along with US government bonds (resulting in
higher yields). Risk-Off, is the alternate chorus of bad economic
expectations leading to the exact opposite behavior. As you can
see from the top chart, the 16 month average correlations for
prices of domestic equity, US bond yields, commodity prices,

(310) 496-4526

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 3

and the dollar (compared to the franc) are at multi-decade highs.

60

Estimated Monthly Average Correlation of Dow Jones Industrial Average (6 month MA) vs.
Weekly Returns
160%

40
110%

30
60%

Weekly % Return

Correlation (-100 to 100)

50

20

10%
10

-40%

2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1976
1975
1974
1973
1972
1971
1970
1969
1968
1967
1966
1965
1964
DJIA Weekly Return (-)
DJIA Weekly Return (+)
Dow Jones Industrial Average - Estimated 21 day average correlation (6m MA smoothed)

Ranked 21 day Historical Correlations of 50 of the Largest Capitalization Stocks in the SPX
(2005 to Present)

100
80

21 day Historical Correlation

Astonishingly individual stocks are also rising and falling

together at the highest levels in decades. In this sense, it doesn't
matter whether you own Microsoft or Google, Chevron or
Exxon. This phenomenon is frustrating traditional value
investors, as the market is not differentiating between good
companies or bad companies, rendering stock and sector
diversification meaningless. Now more than ever professional
stock pickers are having trouble beating the market. Between
1995 and 2007 50% of mutual funds focusing on high growth,
large cap companies beat the Russell 1000 Growth Index. In
2010, only about 24% of those funds beat that index(1). For a
detailed look at how correlations are changing over time consider
the second chart down demonstrating that average correlations
for Dow Jones Industrial Index Average component stocks are at
their highest levels since 1964(2). Consider the third graphic down
showing rankings of the highest and lowest values of realized
correlations (21 days in the past) between pairs of the 50 largest
capitalization stocks in the S&P 500 index. At the peak in
November of 2008 (represented by the blue line) the 50 largest
capitalization stocks in the S&P 500 index had an average
correlation of 0.76. The amazing thing is that we are in the
middle of the recovery and correlations are at 0.62! (see the red
line). Correlations are almost as high now as they were at the

60
40
20
0
-20

Current (Correlation at 0.63)

11/13/2008 (Highest Avg. Correlation at 0.76 )

-40

11/3/2006 (Lowest Avg. Correlation at 0.10)

-60
1

101

201

301

401

501

601

701

801

901

1001

1101

1201

Ranking (Lowest to Highest)

height of the 2008 crash!

Implied S&P 500 Correlation Index (12 month constant adjustement)
80

S&P 500 Index Implied Correlation (12 month)

The market's prediction of future correlations, derived from

option prices, is also at the highest levels ever recorded when
measured by the CBOE's S&P 500 Implied Correlation Index.
These index values are market-based estimates of the average
correlation of the stocks that comprise the S&P 500 index and
have been trending upward since their introduction. Currently the
one-year constant maturity of this index shows highest implied
correlation levels since they started in 2007.

75
70
65
60
55
50
45
40
35
30

Sep-10
Aug-10
Jul-10
Jun-10
May-10
Apr-10
Mar-10
Feb-10
Jan-10
Dec-09
Nov-09
Oct-09
Sep-09
Aug-09
Jul-09
Jun-09
May-09
Apr-09
Mar-09
Feb-09
Jan-09
Dec-08
Nov-08
Oct-08
Sep-08
Aug-08
Jul-08
Jun-08
May-08
Apr-08
Mar-08
Feb-08
Jan-08
Dec-07
Nov-07
Oct-07
Sep-07
Aug-07
Jul-07
Jun-07
May-07
Apr-07
Mar-07
Feb-07

www.artemiscm.com

Index Volatility Sensitivity to Component Volatility at Different

Correlation Levels
100%
90%

Index Volatility Sensitivity

Correlation and the Volatility Surface: If volatility are the waves

in the ocean correlation is the undercurrent. The study of
correlations is important to anyone who trades volatility because
there is a clear mathematical relationship. It is not by coincidence
the volatility surface steepened in the third quarter concurrent with
rising implied correlations. The implied volatility of an index,
such as the S&P 500, is more sensitive when the average
correlations between the components of the index are greater. For
example, according to our formulas if average correlation of index
components rises from 0.35 to 0.70 the volatility of that index
should be 45% more sensitive. Put in other words, the higher the
level of correlation the greater the volatility of volatility (VOV)
and vice versa. The mathematical relationship between index
volatility sensitivity and component correlation is shown by the
chart to the right.

80%
70%
60%
50%

Higher Correlations =
Higher Index Volatility
Sensitivity

40%
30%
20%
10%
0%
0.00

0.10

0.20

0.30

0.40

0.50
Correlation

(310) 496-4526

0.60

0.70

0.80

0.90

1.00

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples
In the third quarter the expectation of future volatility (seen by the slope of
the line) rose substantially in conjunction with higher correlations despite a
rising stock market. The term structure of the VIX futures (see chart) is
now currently at the high end of its historic range. This is partly because
bearish investors are hedging their portfolios by buying volatility futures,
but also because market makers must adjust to fact that volatility is
literally more volatile due to high correlations! The average correlation of
S&P 500 index components is mathematically embedded in the slope of
the volatility surface, but very few ever people make this connection. As
correlations rise volatility become more volatile!

Normalized Slope of Volatility Surface (1=spot)

1.60

Page 4

Normalized Slope of the Volatility Surface

September 2010 vs. Historical Slopes

1.50

1.40

Steeper Slope =
Expectation of Higher
Future Volatility

1.30

1.20

1.10

1.00

0.90
Spot

Month 1

Month 2

September-10

Month 3

May-10

Month 4

Month 5

1 year average

Month 6

Month 7

Average (2004 to Present)

The subtlety of the relationship between volatility and correlation should

not be underestimated. To give an example, in the third quarter I heard
Volatility Surface July to September 2010
many pundits claim that long-dated volatility was significantly overvalued.
35
This was largely due to the observation that the back-months of the VIX
30
term structure rose above 30 even as the VIX dropped to the lows 20s. I
25
heard one "expert" on CNBC make the point (erroneously) that five month
20
7/1
7/12
7/20
7/28
VIX futures at 30 were implying the S&P 500 index would move +/-2%
8/5
8/13
on average per day through the end of the year. This simplistic
8/23
8/31
interpretation ignores the concept that when you sell volatility you are, in
9/9
essence, selling insurance against an unknowable adverse event (a market
9/17
crash). The premium of the insurance contract must be probabilistically
9/27
VIXM6 VIXM7
VIXM4 VIXM5
adjusted by the shifting likelihood of that event occurring. The trader who
VIXM2 VIXM3
VIX Index VIXM1
is short volatility must build a premium into their expectation of future
volatility to account for the fact vol rises faster than it drops. As component correlations rise the probability of violent
exponential jumps in index volatility are greater! Hence the premium over today's volatility level (current VIX) should be
much higher to appropriately account for the added risk. The analyst who advises shorting long-term VIX futures simply
because they are above 30, while ignoring correlation a risk source, is missing a big part of the picture. At high correlation
levels shorting volatility becomes much more dangerous. Consider the chart below which tracks the relationship between
the slope of the intermediate-term VIX futures surface (months 4 and 7 measured by proprietary index) and implied
correlations. As indicated, the slope of the VIX term structure (on an absolute value basis) can be thought of as a risk
premium measuring expectations of future volatility of volatility. The risk premium increases when implied correlations
stay elevated over extended periods of time.
70%

Implied Correlation of S&P 500 Index vs. Slope of VIX Term Structure

60%

80
Slope (Absolute Value) of Month 4 to Month 7 VIX
Term Structure

50%

S&P 500 Index Implied Correlation (12 month)

70

40%
60
30%
50
20%
40

10%

30

0%

Aug-10

Sep-10

Jul-10

May-10

(310) 496-4526

Jun-10

Mar-10

Apr-10

Jan-10

Feb-10

Nov-09

Dec-09

Oct-09

Aug-09

Sep-09

Jul-09

May-09

Jun-09

Mar-09

Apr-09

Jan-09

Feb-09

Nov-08

Dec-08

Oct-08

Aug-08

Sep-08

Jul-08

May-08

Jun-08

Mar-08

Apr-08

Jan-08

Feb-08

Nov-07

Dec-07

Oct-07

Aug-07

Sep-07

Jul-07

May-07

Jun-07

Mar-07

Apr-07

Feb-07

www.artemiscm.com

Slope of Month 4- Month 7 VIX Term Structure

S&P 500 Index Implied Correlation (12 month)

90

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 5

Outside of higher correlations I'd like to highlight three other reasons why the volatility surface increased in the third
quarter: (1) First, a liquidity shortage on the long-end of the OTC volatility surface emerged as sophisticated players
covered short positions following substantial losses on volatility derivatives in May. Among the firms affected were some
of the best and most sophisticated including Goldman Sachs and Warren Buffet's Berkshire Hathaway. (3) Berkshire
Hathaway had underwritten $37 billion in volatility and equity-linked derivative contracts alone since 2004 based on SEC
filings. Although Berkshire has retained those positions Buffet renegotiated their terms and shorted their durations, hence
reducing risk exposure. I find it very amusing that Warren Buffet is so intricately involved in the volatility and derivatives
markets despite his well-crafted image as a plain-vanilla value investor. As for Goldman Sachs, during a July 20th
conference call their CFO said short volatility positions were primarily to blame for a poor second quarter in the equity
derivatives business that saw profits drop 62% from a year earlier. These are just two examples of important institutions
that lost some of their risk appetite for short volatility exposure in the third quarter; (2) Secondly, there has been a recent
proliferation of new "tail risk" or "black swan" hedging strategies and funds. These funds attempt to profit from extreme
events and have provoked increased purchases of out-of-the money options and directionally long volatility derivatives.
Most recently, Pacific Investment Management Company, Citigroup, and Deustche Bank have launched their own "Black
Swan" funds following in the footsteps of Nassim Taleb's Universa. (3) Lastly, a psychological bias called the "snake bite
effect" may continue to affect investors who have emotional scars from the turbulent fourth quarter of 2008. In theory, bad
memories may induce investors to be overly cautious about hedging their portfolios using volatility during the seasonally
weak months of September and October. The late-summer "snake bite" bias was very prevalent last year as well.
80

50 Largest Capitalization S&P 500 stokcs 21 day Rolling Correlation vs. VIX Index
(2005 to Present)

120
60
100

80
40
60

30

40

20

20

10
S&P 500 Correlation (21 day rolling)
60 day Correlation Moving Average

Jan-05
Feb-05
Mar-05
Apr-05
May-05
Jun-05
Jul-05
Aug-05
Sep-05
Oct-05
Nov-05
Dec-05
Jan-06
Feb-06
Mar-06
Apr-06
May-06
Jun-06
Jul-06
Aug-06
Sep-06
Oct-06
Nov-06
Dec-06
Jan-07
Feb-07
Mar-07
Apr-07
May-07
Jun-07
Jul-07
Aug-07
Sep-07
Oct-07
Nov-07
Dec-07
Jan-08
Feb-08
Mar-08
Apr-08
May-08
Jun-08
Jul-08
Aug-08
Sep-08
Oct-08
Nov-08
Dec-08
Jan-09
Feb-09
Mar-09
Apr-09
May-09
Jun-09
Jul-09
Aug-09
Sep-09
Oct-09
Nov-09
Dec-09
Jan-10
Feb-10
Mar-10
Apr-10
May-10
Jun-10
Jul-10
Aug-10

While the financial media has given attention to the correlation phenomenon they are not asking the right questions.
Most pressing is why correlations are at levels associated with an economic crash when we are in the middle of a recovery?
If this recovery was truly healthy correlations would be dropping and the volatility surface flattening, not the opposite!!
Are high correlations and steep vol curves an omen that profound systemic risk is building underneath the surface of this
market?
There are a number of theories for the higher correlations including globalization, new types of investment funds, and
the fact retail investors have abandoned the market. A key structural change is the proliferation of high frequency trading
strategies and exchange traded funds. The media likes to blame these funds for the higher correlations, but I think they are
more of a red herring masking the real problem at hand. I readily admit these funds very likely have had a strong impact on
rising correlations over the past few years starting long before May's Flash Crash. Despite this fact, indexing and high
frequency strategies may not entirely explain the full range of higher than normal cross-asset and cross-currency
correlations. It is always easy for politicians and regulators to find blame in whatever new technology is affecting markets
while ignoring the elephant in the room. Truth be told, there are much larger structural and macroeconomic forces at work.
I prefer the theory that heightened global correlations are the unintended consequence of aggressive monetary

(310) 496-4526

VIX Index

50

Unified Risk Theory - Correlation, Vol, M3, and Pineapples

www.artemiscm.com

140

70

S&P 500 - 21 day Rolling Correlation Index

All of these factors likely contributed to the initial rise in the

long-term volatility surface from May to early August. I see
correlation as the primary reason why the surface remained
steep even as the VIX index dropped into September. (see
chart to the right). At current correlation levels (both realized
and implied) the long-end of the volatility surface is not wildly
overvalued as some seem to claim. As long as realized
correlations remain above 0.60 on the S&P 500 I do not
believe shorting long dated volatility is a bargain. This is not
to say that cataclysmic market crash is imminent, but it does
mean the probability of the VIX breaching 30 for an extended
period of time over the next 6 months outweighs the chance it
doesn't.

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 6

expansion in developed economies amplified by global structural trade imbalances. The graph below elaborates on this
theory by comparing average cross-asset correlations to the volatility of YOY changes in M3 money supply (actual and
estimated from shadowstats.com) (4). The volatility of the money supply measures the change, both up and down, of the
amount of currency in circulation. In theory, up-and-down shocks to the money supply from both financial panics and/or
government intervention (e.g. money printing & QE) lead to higher volatility of money (green bars) and potentially
elevated periods of cross asset-correlations. In simple terms, if the Federal Reserve floods the global system with money,
after a lag period during which the stimulus is absorbed, correlations should eventually rise. For those who prefer common
sense to statistics let me explain this macro-theory using a micro-metaphor.
50

70

Volatility of Changes in YOY M3 Money Supply (qtrly/annualized) vs. Avg.Correlations of USEquity, Bond Yields, Swiss Franc/USD, and Commodities (16 months)
1972 to Present

40

60

50

Higher volatility of
money supply =
increased crossasset correlations?

20

40

10

30

20

-10

10

-20

0
2010

2009

2008

2007

2006

2005

2003

2004

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1989

1990

1988

1987

1986

1985

1984

1983

1982

1981

1980

1979

1978

1977

1975

1976

1974

1973

1972

Volatility of M3 (YOY changes - qtrly - annualized)

Volatility of M3 (%)

Average Correlation (%)

30

16 Month Average Correlation (SPX,Commodities, CHF/USD, 10yr UST Yield)

This past summer my wife and I visited the beautiful Hawaiian island of Kauai and rented a cottage near the Napali
coast. Twice a week the locals hold a farmers market next to the elementary school which we visited to buy fruits and
vegetables for grilling with freshly caught Ono for dinner. This one local family sells delicious golden pineapples but you
have to get to them early because they sell out fast. In the farmer's market, just like the stock market, the higher quality
goods eventually command a premium over less desirable produce. Those pineapples were the Hanalei Bay equivalent of
Apple stock selling at a high P/E ratio. Over time the demand for high quality produce like the pineapples should be
uncorrelated with demand for lower quality fruit.
Hypothetically let's imagine that Ben Bernanke shows up in the parking lot of the Hanalei Bay farmers market. Over a
bullhorn he agrees to give away $1,000 to everyone in attendance on the condition they spend the money immediately.
What will happen? The quality of the produce instantly becomes of secondary concern to simply realizing the full value of
the hand out. Whether or not you are buying star fruit, pineapples, or bananas it is largely irrelevant because if you don't
buy something you will lose out on the free subsidy. The demand for all produce will rise near equally as the correlations of
fruit increase and the Hanalei Bay farmer's market ceases to function as a true marketplace.
So why are the rules of financial physics radically different for the global economy compared to the example of the
farmer's market? Is it wrong to substitute star fruit for yen or gold for pineapples? We know that correlations of assets rise
sharply when the money supply contracts violently and cash is dear, but why can't cross-asset correlations increase when
the same violent adjustment happens in reverse? They can and they are. The reason we don't feel it day-to-day is that global
economy evolves in continuous time while the farmer's market only lasts three hours. Hence market imbalances may take
years (or decades) to be fully recognized and assimilated in the former while the adjustment is near instantaneous in the
latter.
I don't understand why CNBC pretends like it is some great mystery as to why stocks, gold, bonds, and commodities all
increased in tandem in September. The Federal Reserve is clearly signaling, through QE2, the intention to limit downside
price risk by making any asset yielding above 1% desirable while artificially propping up household net worth! Taken in
context of the farmer's market, if Ben Bernanke shows up the simple strategy is to buy as many of the good pineapples as
possible, but when those run out you will be forced to spend the remainder of your cash on less desirable fruit. If you don't

www.artemiscm.com

(310) 496-4526

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 7

use it (cash) you lose it because there is a negative real return in today's economy. When the good pineapples are increasing
at the same rate as the bad mangos, the situation is unsustainable. Price discovery is not only broken, it's crazy. Why it is
happening is obvious, but what happens going forward is the problem!
Elevated levels of correlations are a bad omen for the long-term health of the global
economy. We will all face vast long-term unintended consequences from the monetary
policy of the developed world ranging from increased systemic risk, asset bubbles, trade
wars, and/or violent currency fluctuations. Higher correlations increase the potential for
volatility and amplify global risk in a self-reinforcing cycle. I am bracing for more shocks
down the road, however I think they will manifest first through violent sell offs in the global
bond markets, and then ripple into domestic equities. Exact timing is impossible to tell and
spot volatility may look normal until some as yet unknown tipping point is reached. This
higher risk is what the steep volatility surface is foreshadowing. What to do in the meantime?
Bring your wheel barrel to the market, have fun while it lasts, but carefully hedge your
portfolio over a pineapple salad.

Sincerely,
Artemis Capital Investors, L.P.
Christopher R. Cole, CFA
Managing Partner and Portfolio Manager
Artemis Capital Management, L.L.C.

www.artemiscm.com

(310) 496-4526

Artemis Capital Management, LLC | Unified Risk Theory - Correlation, Vol, M3, and Pineapples

Page 8

THIS IS NOT AN OFFERING OR THE SOLICITATION OF AN OFFER TO PURCHASE AN INTEREST IN ARTEMIS CAPITAL
INVESTORS, L.P. (THE FUND). ANY SUCH OFFER OR SOLICITATION WILL ONLY BE MADE TO QUALIFIED
INVESTORS BY MEANS OF A CONFIDENTIAL PRIVATE PLACEMENT MEMORANDUM (THE MEMORANDUM) AND
ONLY IN THOSE JURISDICTIONS WHERE PERMITTED BY LAW. AN INVESTMENT SHOULD ONLY BE MADE AFTER
CAREFUL REVIEW OF THE FUNDS MEMORANDUM. THE INFORMATION HEREIN IS QUALIFIED IN ITS ENTIRETY BY
THE INFORMATION IN THE MEMORANDUM.
AN INVESTMENT IN THE FUND IS SPECULATIVE AND INVOLVES A HIGH DEGREE OF RISK. OPPORTUNITIES FOR
WITHDRAWAL, REDEMPTION AND TRANSFERABILITY OF INTERESTS ARE RESTRICTED, SO INVESTORS MAY NOT
HAVE ACCESS TO CAPITAL WHEN IT IS NEEDED. THERE IS NO SECONDARY MARKET FOR THE INTERESTS AND
NONE IS EXPECTED TO DEVELOP. NO ASSURANCE CAN BE GIVEN THAT THE INVESTMENT OBJECTIVE WILL BE
ACHIEVED OR THAT AN INVESTOR WILL RECEIVE A RETURN OF ALL OR ANY PORTION OF HIS OR HER INVESTMENT
IN THE FUND. INVESTMENT RESULTS MAY VARY SUBSTANTIALLY OVER ANY GIVEN TIME PERIOD.
CERTAIN DATA CONTAINED HEREIN IS BASED ON INFORMATION OBTAINED FROM SOURCES BELIEVED TO BE
ACCURATE, BUT WE CANNOT GUARANTEE THE ACCURACY OF SUCH INFORMATION.
The General Partner has hired Unkar Systems, Inc. as NAV Calculation Agent and the reported rates of return are produced by
Unkar for Artemis Capital Fund. Actual investor performance may differ depending on the timing of cash flows and fee structure.
Past performance not indicative of future returns.
(1) Source: Morningstar
(2) Estimated average correlation of Dow Jones Industrial Index stocks is derived from proprietary calculations using data from
yahoo finance. The dataset is subject to survivorship bias and represents a best attempt at estimating the average correlation
levels.
(3) Source: Bloomberg, "Volatility Trade Buffet Embraced Backfires for Wall Street Hedge Experts" July 29, 2010
(4) Source: www.shadowstats.com / Note: the Fed ceased publishing M-3, its broadest money supply measure, in March 2006.
John Williams Shadow Government Statistics M-3 Continuation estimates current M-3 based on ongoing Fed reporting of M-3s
largest components (M-2, institutional money funds and partial large time deposits) and modeling of the balance. Please see John
Williams Shadow Government Statistics subscription service for more details.

www.artemiscm.com

(310) 496-4526