Issue 11 - The Technical Analyst

Transcription

july/aug 2005
The publication for trading and investment professionals
www.technicalanalyst.co.uk
DrKW’s
TECHNICAL
STRATEGISTS
Sending out clear trading signals
Outlook for bunds
Volatility arbitrage funds
Commodity currencies
DeMark indicator signals
major reversal
Fimat’s Rami Habib
discusses the strategies
Which are the
genuine articles?
WELCOME
Algorithmic execution strategies (AES) are increasingly being adopted by investment
managers and hedge funds as a means of reducing both transaction and commission
costs. However, until now there has been little data available on the full extent of AES
usage. In this issue we present the results of a Banc of America Securities sponsored
survey - the first of its kind to report on AES application across buy-side equity firms
in the US.
We also look at how Dresdner Kleinwort Wasserstein has developed a web-based
proprietary market sentiment tool to provide their clients with daily market calls and
timing recommendations across a range of asset classes. PIA, (Price Information
Advantage), is DrKW's attempt to leverage their in-house technical analysis expertise
in the competitive prime brokerage market.
Matthew Clements, Editor
CONTENTS 1 > FEATURES
Bund futures
Warning of major reversal
JUL/AUG
>12
Gordon Kolling, technical analyst at
Commerzbank in Frankfurt, alerts us to a
significant DeMark signal on bund futures.
Intermarket analysis
Marc Zaffran of Societe Generale in London
explains their 'Market Behaviour and
Historical Patterns' - a tool for forecasting
short-term FX rates.
Sending clear signals
Max Knudsen of Dresdner Kleinwort Wasserstein
in London discusses his team's client
focused research product that looks to provide
clear, user friendly trading recommendations.
© 2005 Clements Biss Economic Publications Limited. All rights reserved. Neither this publication nor any part
of it may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without the prior permission of Clements Biss Economic
Publications Limited. While the publisher believes that all information contained in this publication was correct
at the time of going to press, they cannot accept liability for any errors or omissions that may appear or loss
suffered directly or indirectly by any reader as a result of any advertisement, editorial, photographs or other
material published in The Technical Analyst. No statement in this publication is to be considered as a
recommendation or solicitation to buy or sell securities or to provide investment, tax or legal advice. Readers
should be aware that this publication is not intended to replace the need to obtain professional advice in
relation to any topic discussed.
July/August 2005
>15
> 39
>>
THE TECHNICAL ANALYST
1
8&#XXXDRHDPN
&."*-DPOUBDU!DRHDPN
-0/%0/
1"3*4
3&"-5*.&
5&$)/*$"-"/"-:4*4
2605&4
/&84
015*.*;&
.0/&:."/"(&.&/5
#"$,5&45*/(
*/%*$"5034
$6450.
53"%&4:45&.4
'03.6-"4
/08*/4065)"'3*$""/%5)&.*%%-&&"45
Gary Stafford explains how Gann angles
should be properly applied.
“Volatility arbitrage fund managers probably have around $3 billion under management globally.” - Rami Habib, Fimat.
24
33
CONTENTS 2 > REGULARS
Editor: Matthew Clements
Managing Editor: Jim Biss
Advertising & subscriptions:
Louiza Charalambous
Marketing: Vanessa Green
Events: Adam Coole
Design & Production: Paul Simpson
INDUSTRY NEWS
Special feature: Algorithmic trading survey
04
06
MARKET VIEWS
Gold: Long-term trend will hold
FTSE 100: First Fibonacci target at 5450
Bunds: DeMark indicators warn of major reversal
08
10
12
TECHNIQUES
DeMark Retracements and Trendlines
Intermarket analysis: Forecasting short-term FX rates
Commodity cycles: Technicals and fundamentals coincide
The proper use of Gann Angles
Setting stop-losses using price volatility
15
18
21
24
27
The Technical Analyst is published by
Clements Biss Economic Publications Ltd
Unit 201, Panther House,
38 Mount Pleasant,
London WC1X 0AN
Tel: +44 (0)20 7833 1441
Web: www.technicalanalyst.co.uk
Email: editor@technicalanalyst.co.uk
SUBSCRIPTIONS
Subscription rates (6 issues)
UK: £150 per annum
Rest of world: £175 per annum
For information, please contact:
subscriptions@technicalanalyst.co.uk
ADVERTISING
For information, please contact:
advertising@technicalanalyst.co.uk
PRODUCTION
Art, design and typesetting by
all-Perception Ltd.
Printed by The Friary Press
ISSN(1742-8718)
INTERVIEW
Rami Habib,
Quantitative analyst, Fimat International
33
SUBJECT MATTERS
Identifying commodity currencies
36
SOFTWARE
Dresdner Kleinwort Wasserstein's PIA
39
BOOK REVIEW
Mechanical Trading Systems
by Richard L. Weissman
43
COMMITMENTS OF TRADERS REPORT
LONG-TERM TECHNICALS
EVENTS
44
46
48
July/August 2005
3
Industry News
EQUIS INTRODUCES METASTOCK
FX PRO
Equis International has introduced
MetaStock Pro FX, a real-time version of MetaStock specifically created for FX trading. The package
includes a FX multiple time frame
system, allowing traders to get a "top
down" view of a FX currency pair,
and studies including the FX Keltner
System. It looks for patterns that
occur on a very infrequent basis and
informs the trader of subsequent
price moves. Keltner System 2
serves the same purpose but for patterns that occur on a more frequent
basis. Also included is the FX
Pattern System which combines
Bollinger Bands, Standard Error
Bands and a Stochastic Oscillator.
Interbank FX has developed a proprietary execution engine incorporated into the MetaStock Pro FX soft-
ware allowing traders to deal directly
LSE to launch
first oil ETF
CBOT EXTENDS ELECTRONIC
TRADING HOURS
The LSE will launch the first oilbacked EFT at the end of July. Oil
Securities, the company that set up
the first gold-backed ETF in
Australia will probably list two oil
securities on July 28 in London.
Graham Tuckell of Oil Securities
said one security would track the
price of the front-month Brent contract traded on the IPE while the
other would track the West Texas
Intermediate contract traded on the
NYMEX. He said he was confident
that the investment in the oil ETFs
would surpass the $3.5bn so far
invested in the gold ETFs.
The Chicago Board of Trade
(CBOT) has announced the expansion of trading hours on its electronic platform, a response to global
demand for additional trading
opportunities at the exchange.
CBOT products will be available for
an additional hour each day.
The new operating hours for the
CBOT electronic platform will be
6:00 p.m - 4:00 p.m. (the following
day), with the pre-opening session
now beginning at 5:30 p.m.
Previously, the hours of operation
were from 7:00 p.m. - 4:00 p.m.,
4
July/August 2005
from the charting software.
with the pre-opening beginning at
6:30 p.m. The change will be completed in conjunction with the
upgrade to the CBOT electronic
platform on October 9, 2005
Industry News
ON THE
MOVE...
Alan Johnson has joined Macquarie
Bank in London as consultant Asian
Technical Analyst. Johnson will provide analysis on Asian equity markets
including Japan for Macquarie clients
such as long-term pension funds and
hedge funds. Previously he was a
consultant analyst with HSBC and in
1987 founded Market Directional
Analysis, a research company specialising in interest rate and Asian
Pacific equity technical analysis.
Alan
Johnson
EXCHANGE NEWS:
IPE energy futures see volume record
after switch to electronic trading
The Intercontinental Exchange (ICE) say a new volume record
in the benchmark IPE Brent crude oil futures contract was set
on July 12. IPE Brent volume reached 185,144 contracts,
exceeding the earlier record of 181,182 contracts set on
September 14, 2004. ICE say the figures reflect the increase in
electronically traded contracts following the transition in April
from IPE's open-outcry floor to fully electronic trading.
The CBOT announced that exchange-wide average daily volume
climbed to a new record in the second quarter of 2005. ADV
rose to 2,889,059 contracts, up 14.8% from the same period last
year. Total volume in the quarter also reached a record
184,899,800 contracts, up 16.6% from the same period in 2004.
Euronext.liffe traded a record month in its flagship contract, 3month Euribor, in June at over 18 million contracts representing an average daily volume of over 834,000. All short term
interest rate products traded a total of 36 million contracts during June, up 34% on June 2004.
July/August 2005
5
Special Feature
ALGORITHMIC TRADING
The first buy-side survey shows firms have yet to wholeheartedly
embrace algorithmic execution strategies
A
lthough algorithmic
trading is a fast developing area within the
financial community, until now
there has been no detailed data
on the extent of usage of algorithmic execution strategies
(AES). Financial technology
research house, Financial
Insights, and Banc of America
Securities have recently conducted the first extensive survey
of algorithmic trading across
the trading and investment
community. The results show
that equity buy-side firms have
yet to fully embrace the benefits
offered by AES.
The use of algorithmic execution
strategies has grown in recent years as a
result of several factors. These include
the decimalisation of security prices
producing fragmented liquidity across
multiple prices which has made achieving best price from execution more
complex. In addition, recent FSA recommendations for best execution practices have put pressure on fund managers to provide a process that ensures
they are achieving best execution for
their clients. The market has also seen
falling costs of electronic trading tools
and increased pressure to reduce transaction costs.
The survey shows that 91% of buyside firms now use order management
systems (OMS) in their trading. Use of
electronic communication networks
(ECNs) stands at around 80% of firms
with crossing networks usage at 78.3%
6
and direct market access (DMA) at
73%. The prevalence of these various
electronic trading practices show that
buy-side traders are now actively working their orders in the market in order
to achieve lower commission rates and
reduced information leakage. Above all,
the increased used of electronic trading
tools means that transaction costs are
now seen by fund managers as a significant source of negative portfolio performance.
Financial Insights surveyed 60 head
equity traders from the top 477 investment management firms, pension
funds and hedge funds in the US
regarding specifically their use of algorithmic execution strategies. The survey
shows that on the buy-side 66% of registered investment advisors (RIAs) and
92% of hedge funds use algorithmic
trading to some extent (see Chart 1).
Nevertheless, these figures belie the
real extent to which AES are being
deployed as 70% of those surveyed say
they are doing less than 10% of their
trading using algorithms. However,
usage increases sharply for larger firms
with $70 billion and above under management - 30% of large firms and
33.4% of hedge funds of this size
reported algorithmic trading usage
above 10% of total order flow.
The main reason given by buy-side
firms for not adopting algorithmic
trading was that it does not support
their fiduciary responsibility for best
execution. This says Rob Flatley, cohead of sales for electronic trading
services at Banc of America Securities
in New York, illustrates that the application and methodology of algorithmic
trading is not yet fully understood.
Benefits of algorithmic trading
The survey shows that the buy-side
perception is that the benefit of algorithmic trading is in increased productivity rather than improved trading performance - head traders say they are
happy to send their easiest orders to
algorithms to get them out of the way
RIA
Hedge funds
Chart 1. 66.7% of firms say they use algorithmic trading to some extent
July/August 2005
Special Feature
algorithmic trading is executed via the
most basic algorithms available" he
says. "Moreover, at only 5% of total
order flow, AES use is presently less
than that suggested by recent financial
media coverage. As such, the scope for
growth and development of AES use is
enormous". For example, the survey
shows that only around 40% of firms
are customizing their algorithmic trading models to suit their own trading
even though system vendors usually
provide customisation at no extra cost.
80
75
70
65
60
55
50
Lowers
commission rate
Provides
increased
anonymity
Convenience
Higher trader
productivity
Provides the
Reduces explicit
ability to scale
trading costs
without increasing
headcount
Reduces implicit
trading costs
Yields more
consistent
performance than
traders
Chart 2. Benefits of using algorithmic trading (% of firms)
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Don't know
>50%
26-50%
11-25%
5-10%
1-4%
Chart 3. Algorithmic trading as % of total trading (% of firms)
so they can focus on those that are
more difficult. According to Flatley, a
greater focus on transaction costs
analysis (TCA) will lead to the realisation that a more consistent trading performance is also a significant benefit
from adopting algorithmic trading.
Hedge funds have led the way in the
use of algorithmic trading systems
largely because they have had more to
gain from increased productivity and
there is a greater need amongst the
smaller companies to reduce staff
costs. Generally, low commission costs
and anonymity are seen as the main
benefits of algorithmic trading on the
buy-side (see Chart 3). However, these
same respondents don't see reduced
implicit trading costs (market impact
plus opportunity cost) or the consistent
out performance of algorithms over
traders as key. Banc of America
Securities stress that these are two of
the most important benefits of algorithmic trading. At present, only 57%
of firms surveyed said they expect their
use of algorithmic trading to grow over
the next 2 years.
Even with those firms that regularly
use AES, Flatley says the full benefits
offered by algorithmic trading have yet
to be fully exploited. "The majority of
July/August 2005
The sell side
The development of algorithmic trading was, in part, due to a response by
the sell-side to pressure on commission
rates
from
fund
managers.
Consequently, algorithmic trading has
become integral with around 30-35%
of trading volume at sell-side bulge
bracket firms now executed using algorithms. This has led to reduced costs
with the ability to scale volume - for the
same costs, more orders can be traded
and desk head counts can be reduced.
Rob Flatley concludes, "Algorithmic
trading will extend into futures, foreign
exchange and Treasuries in the not too
distant future. It is feasible that algorithmic trading volume could double
each year over the next two years. Banc
of America Securities expects the level
of algorithmic trading order flow adoption to increase to 20% by 2007 with
the use of more sophisticated strategies. The key for traders is determining
the best algorithm and its appropriate
settings. This should lead to a more
even usage across equity market capitalisations. The challenge for the sell-side
is to invest in quantitative resources and
supporting technologies to improve
strategy performance for less liquid
stocks."
Further information can be
obtained from Sean Mogle at
smogle@webershandwick.com
7
Market Views
GOLD:
LONG-TERM TREND WILL HOLD by Ron William
P
erseverance is the word that best describes the trend
in gold as we press through the hot summer months.
Originating from the cyclical lows of $252, this
metal's spectacular performance relative to major currencies,
equities and bonds, has heightened its portfolio appeal within the investment community. That said, after registering a
sixteen-year high at $456 in late December, we are currently
seeing a period of tempered consolidation. Ultimately, this
year's tremendous rally in the US dollar, compounded by
acute euro weakness (as EU economic growth and unity
issues loom), continues to hamper gold on the upside.
Nevertheless, the easing of the inverse correlation between
gold and the USD in recent months provides welcome traction for speculative gold bugs and gives gold the opportunity to push on higher.
Price targets
The seeds of change were planted in 1999, a time when
gold ended its twenty year secular decline and started rising.
In that year the US stock market was booming and the dollar had appeared to be the universal store of wealth for the
21st century, while gold had almost become a relic.
Following its price slump at the cyclical lows of $252, a
double bottom formed. This base would later dominate the
positive pattern of higher peaks and troughs.
Furthermore, gold's bull-phase became
strongly guided by the rising long-term 70week moving average (Figure 1). By late
December 2004 it had retraced almost one
half of what had been lost and - although
trend followers made further linear projections upwards - momentum then began to
wane.
In practice, market swings remain complex
as bulls and bears battle sideways within a
converging-line pattern. Confirmation above
$446/$456 promises to slice through our triangle pattern objective at $483, yielding
accelerated swings towards the December
1987 peak at $503, equivalent to the impulsive 5th wave of an Elliott Wave structure.
Figure 1. Weekly gold chart guided by the long-term 70 week moving average.
Settlement below the $410 level would put
this scenario on hold and would provide
downside targets at the 25% or 38.2%
Fibonacci retracement levels (of the primary
trend), before regaining composure.
Assessing the real value of gold against other
currencies adds further weight to the longterm bullish gold theme. Figure 2 shows the
price of gold in terms of euros, yen and sterling all developing positive patterns of higher
peaks and troughs above their rising longterm 200-day moving averages, a widely
watched moving average. Interestingly, closer
examination of the EUR/Gold chart shows
there has been a sustained monthly breakout
from a seventeen year-old resistance area.
Figure 2. Daily gold priced in euros, yen and sterling. Note rising trends above their 200- Admittedly, this could be a short-lived emoday MA.
tional reaction to EU unity and growth chal8
July/August 2005
Market Views
lenges. However, while EUR/Gold machinates above €352
the risks to the upside remain. Technical measurements
favour €405 as a minimum upside target (€352 + €53).
Gold/USD correlation
Figure 3 shows that the inverse relationship between gold
and the USD index (trade weighted) follows a cyclical pattern, and looking at the 1-month annual rolling correlation
tells us that it is beginning to weaken once again. While the
past couple of years have seen a fall in the USD index
matched almost exactly by an advance in gold prices, in
recent months this trading relationship has eased from -0.9
to -0.7, a reflection of the fact that while EUR/USD
descended to $1.20 from just below $1.25, gold managed to
turn upwards towards the $443 level.
The loosening relationship between these two powerful
asset-classes provides welcome traction for speculative gold
bugs. Moreover, there is growing confidence that with
prices being increasingly driven by fund managers carving
out a bigger place for gold within their porfolios, the drag
from USD price action can be overcome. If this is the case,
then it provides another reason for seeing gold advancing
back to the $456, and possibly onwards to the $500/503
barrier.
Figure 3. Rolling annual correlation coefficient showing the inverse
relationship between gold and USD index following a cyclical pattern.
Note weakeness in recent months.
Figure 4. Illustrates how net-long speculative positions have been one
of the primary drivers for gold. Note sizeable long positions remain.
July/August 2005
Watch the non-commercials
Whether you are following gold's positive pattern of higher
peaks and troughs or are simply interested in playing the
break, it is always worth identifying trends in market positioning. To do this, we take a look at the Commitment of
Traders Report (COT). Founded by the Commodity
Futures Trading Commission (CFTC), it requires
traders/investors to report their daily positions. The CFTC
classifies traders into three groups: commercial traders
(hedgers), non-commercial traders (large speculators), and
small traders (small speculators). Most TA textbooks recommend watching the commercials, as larger hedgers are
traditionally assumed to be the smart money. However, this
has not been the case with gold over the last five years.
Indeed, it has been heavily sold short by the commercials
ever since the new bull phase started. With similar negative
positioning against the Euro, commercial sentiment has
favoured the safe haven of the Swiss Franc.
Instead, the renaissance of gold was helped by diminished sales on the part of central banks, reduced downside
hedging activity by producers and a lack of returns on equities post the bursting of the tech bubble, thus marking the
beginning of the speculative long position on the current
gold up-trend. Figure 4 shows that net-long speculative
positions have successfully tracked the resurgence of gold.
Following a record acceleration of speculative activity into
2005, positions have since been scaled back. However, current sizable speculative long positions coupled with a likely
increase in physical demand (driven by seasonal summer
trends), should alert us to renewed upside impetus. (Note
gold demand cycles are historically stronger during the summer period when wholesalers typically build up inventory
for the end of year retail Christmas pick up).
Conclusion
The major upward trend in gold remains dominant, with
any short to medium-term weakness indicative of consolidation. These corrections should present longer-term
investors with good buying opportunities. In price terms,
confirmation above $456 promises to slice through our pattern objective at $483, yielding accelerated swings towards
the December 1987 peak at $503. Indeed, with oil driving
the USD, gold should benefit from any medium-term weakening of the USD. This effect is supplemented by the
increase in gold demand from the Middle East, as the
strength in oil prices stokes demand through wealth creation. Moreover, persistent underperformance of other
asset classes, coupled with the generalized increase in overall
commodity demand from emerging markets, looks set to
entice gold back above the peak levels of $850 in the
longer-term.
Ron William is technical strategist at IDEAglobal in
London.
9
Market Views
FTSE 100:
FIRST FIBONACCI TARGET AT 5450
by Gökhan Erem
I
n the second quarter of 2005 stock markets showed us
their positive side again and for many market participants it was an unpleasant surprise. Price action in the
equity markets now reminds me of the second half of the
nineties when we saw markets extending without any obvious reason. More and more people were, and now again are,
increasingly worried over fundamental issues like the oil
price, the EUR/USD exchange rate and what Mr
Greenspan is going to do and say at the next FOMC meeting. Yet equity markets continue to rise beyond fundamentally-based reasoning. And today, market participants are
displaying the same stubborn unwillingness to change their
mind set and their position in the market.
A more technical description of recent developments
would be that short interest in the markets has risen too
much to be able to yield serious profits on downward
focussing positions. In this environment the reaction to any
decline is often short term profit taking or loss limiting buy
orders that immediately provide demand and thus take
prices back to the highs and sometimes even beyond those
recent extremes.
Long-term target (Figure 1)
The major trend has been up since the lows set in 2003, but
let's focus on the second half of this trend where the trend
was finally confirmed.
In 2005, we saw the first serious test of the major trend.
Prior to this test, the trendline had been drawn under two
points making it a trendline under construction at most.
The decline came and brought prices to the 61.8% retracement level of the most recent rise prior to the decline. That
support held and the trend was fact for the first time.
Figure 1.
10
July/August 2005
Market Views
Figure 2.
We can now use Fibonacci levels to look for the next target within this bullish trend: The rise prior to the decline
started at 4283 and reached a maximum of 5077. After that,
the FTSE corrected to a Fibonacci level at 4773. If we
make a 100% projection of the rise from 4283 to 5077
(about 800 points) and add it to the new support at 4773,
the target becomes 5573.
This target remains valid as long as the stop-loss just
below the trendline at 4900 remains untouched. Below that
level the trend is broken but because that would mean letting a lot of profits evaporate before changing direction, I'd
place a stop just below the last top in the trend, in this case
5050.
Short-term target (Figure 2)
The decline on the 7th of July following the London
bombings started at the top end of a short term uptrend.
The rise was already developing characteristics of an exponentially rising trend, suggesting the move was becoming
overextended and unsustainable. At least it needed a correctional decline in the short term. So the decline came and
took prices through the short term uptrend until the 50%
retracement level was reached.
As on the longer term outlook where the natural reaction
lows were formed around Fibonacci levels, the 50% retraceJuly/August 2005
ment was met and provided support. In Figure 2 this is
illustrated by the use of a Speed/Resistance fan where the
middle diagonal is the 50% retracement line. After measuring the price action from the bottom to the end of the
trend, a projection for future prices can be made. The top
at that moment was formed at 5225 and the starting point
was just below 4773, which leaves a move of at least 450
points. When we project these points upwards from the
5022 level, where prices stopped falling on the 7th of July, a
target of 5450 is found.
Once again, support is found at Fibonacci levels and a
second target can be calculated using Fibonacci projections.
This conjuction of two targets increases the probability of
the target being met.
From a safety point of view, I choose the lowest and thus
the easiest target to reach. This does not mean that the rise
in prices ends there but implies a high probability of that
target being met before correcting severely.
To summarize: As long as the uptrend in the FTSE 100,
currently at 4900, doesn't give way, the first target for this
uptrend is the 5450 level.
Gökhan Erem is senior technical analyst with the
Hedge Fund Team at Rabo Securities.
11
Market Views
BUNDS:
DEMARK INDICATORS WARN OF MAJOR REVERSAL
by Gordon Kolling
S
ince the Fed started raising interest rates on June 30th
2004, the logical conclusion must have been that longterm interest rates are due to go higher. To the surprise of almost all market participants they have not. The
Fed Chairman has talked about a "conundrum " and many
research houses have thrown in the towel and stopped predicting higher long-term rates ahead. Nevertheless, away
from the fundamentals, DeMark indicators have proved to
be highly reliable in forecasting the direction of longer-term
rates.
The second quarter of 2005 produced a very rare signal the first quarterly TD Combo sell signal since the Bund
futures started trading in 1990. The signal implies significant long-term market exhaustion and presents an especially
good trading opportunity.
Tom DeMark has frequently stressed that the best trades
usually appear when different time periods give similar signals and when similar products confirm. Therefore the key
components in judging the quality of the quarterly sell signal are the other time frames as well as the situation in
other related interest rate futures markets.
Monthly Countdown
Figure 1 shows that the monthly Bund chart gave a "13"
TDAC exhaustion indication in May. This is a rare signal in
the 10-year German government debt market. There are
two prior instances when it timed a market turn extremely
well. The first time was in December 1993 when it appeared
in the 10-year yield chart (Figure 2) and the second time
was in December 1998 when it appeared in the futures market (Figure 3). The chart in Figure 1 highlights the "power
of the 13". In both cases, the market sold off by more than
200 basis points in yield terms within one year. In terms of
the fixed income market this is a crash. A monthly TD
Combo exhaustion signal which appeared in June also confirms the massive corrective potential of the current setup.
Weekly sell signal
A look at the weekly chart in Figure 4 displays an interesting situation. A TD Combo "13" sell signal appeared on
May 13. Simultaneously a new Setup had started counting
and had already reached a "6". In such situations the recommended strategy is to wait for a completed Setup or a flip in
the opposite direction, i.e. a closing below the close four
weeks ago.
That strategy helped to avoid getting into the trade too
12
July/August 2005
Figure 1.
Figure 2.
Figure 3.
Market Views
early. Three weeks later - on June 3rd - the 13(9) signal was
there. Looking at the past, such signals are reliable weekly
turning points. Such signals appeared at the bottoms in
1994 and 2000 and at the top in 2003.
Daily situation
On June 12 we received a daily 9-13-9 sell signal that
proved very accurate in timing the expected move lower.
The market sold off from 123.62 to exactly 121.28. At that
point four signals occurred that warned about the quality of
the down move. 121.28 is exactly the TDST level and a
break below would have been disqualified on that day.
Simultaneously T-Note and Bund futures both had hourly
"13" buy signals and most importantly the market had sold
off from the top for three consecutive days - often a reliable sign that the move is prone to reversal. Consequently
the market traded back up and even made a new high.
When the market spiked up after the London bombings we
received a new sell signal on an intra day and weekly basis.
The new highs have produced evidence that the market is
ready for a swift reaction lower.
Downside potential
In the past, reactions to monthly TDAC signals led the
→
DeMark glossary
The TD Combo
The TD Combo consists of two distinct stages that are designed to anticipate trend reversals: Setup and Countdown.
Setup: The initial setup phase consists of a series of at least nine consecutive closes less than the close four trading bars
earlier for a buy setup and at least nine consecutive closes greater than the close four trading bars earlier for a sell setup.
The Setup establishes the environment for the market and determines whether a trader should be looking to buy or sell
the market.
Countdown: Once a Setup is complete, the TD Combo refers back to day 1 of the Setup and begins a Countdown on
that day. The key elements of a sell Countdown are threefold:
1. A series of 13 successive closes less than or equal to the low two price bars earlier
2. Each countdown day's high is greater than the previous trading day's high
3. Each successive Countdown day's close is greater than the previous Countday day's close and the previous trading day's
close
The TD Aggressive Combo (Version 2)
TDAC has lesser requirements for Countdown Bars 11, 12, and 13. They only need to close successively higher for a Sell
Countdown and successively lower for a Buy Countdown. The aggressive version of the TD Combo is designed to forewarn of an impeding exhaustion of the trend and often signals a counter-trend correction. Usually the market continues
after such a correction and completes a Countdown 13 in the non-aggressive TD Combo version.
The TD Termination Count
The Count allows an even looser definition for Countdown Bar 13, only requiring that its opening or closing price be higher than the closing price of Countdown Bar 12.
"Recycling"
Recycling is a concern that could arise in a strongly trending market. Generally speaking, if a subsequent Setup occurs prior
to completion of countdown, then a new countdown process must begin.
TD Setup Trend (TDST)
These are important support and resistance levels. The TDST is defined as the highest price of bars 1 to 9 in a Buy Setup
or the lowest price of bars 1 to 9 in a Sell Setup. Furthermore, there are certain rules that are used to determine whether
a break through these support or resistance levels should be followed through or if they should be treated cautiously, i.e.
false breaks. If the DeMark criteria are met (not discussed here), then the break is qualified. If the criteria are not met,
then the break is disqualified (false). See page 15.
The TD Combo and the Bund market
Over the last few years, combining the TDAC and the TD Combo has proved to be a very accurate method for timing
corrections to the Bund futures market. In Figure 2 you can see two TD Combo "13" signals - one in the second quarter
of 2003 and one in Q2 2005. Both signals were preceded by a marked downside correction. The size of the correction
would have stopped out many long positions. This is where the TDAC is of outstanding value. In Figure 1, I have marked
out the signals created by the TDAC. It shows how well they predicted the corrections. According to DeMark, corrections
following a TDAC signal usually last for three to 5 periods. I have examined the results for Bunds and found that this rule
can be fully applied to Bunds. Interestingly the market tends to correct 3-5% after such a signal. I suggest using short-term
signals (hourly charts etc) and qualification criteria to time the correction exactly.
July/August 2005
13
Market Views
Conclusion
When looking at the combination of quarterly and monthly
13 signals in the Bund futures and related markets, the possibility of a major turn in the market must be taken seriously.
DeMark indicators are an extremely powerful tool in timing trading decisions. The signals currently in place warn all
bond bulls to be very cautious. With quarterly, monthly and
weekly sell signals in place the chances are that we are going
to see a significant move lower. The situation compares in
many ways to the two big Bund sell offs in 1994 and 1999. I
am confident that Tom DeMark's indicators will guide us
through this process with impressive reliability and will help
us to time the down move with the same accuracy as was
shown on the way up.
Figure 4.
market back towards the monthly TDST level. This is located at 114.19. There is a good probability that we shall trade
towards that level in the next six to twelve months. To
judge the end of the down move, we need to watch price
behaviour at retracement and TDST levels. The first daily
TDST is at 121.28 and the first weekly TDST is at 118.59.
Qualified breaks of such levels are good indications that the
move is not exhausted.
14
Gordon Kolling is technical analyst at Commerzbank
Group Treasury and Research.
Sources: CQG; "New Market Timing Techniques" by Tom
DeMark, John Wiley & Sons, 1997.
July/August 2005
Techniques
DEMARK RETRACEMENTS AND TRENDLINES
by Trevor Neil
DeMark takes the guesswork out of judging the most important
support and resistance levels.
T
om DeMark, the wizard's wizard*, is probably best known
for his TD Sequential and TD
Combo patterns but two others that
have impressed me are TD Lines and
TD Absolute Retracements. TD Lines
because they really made me think
about how I was drawing my trendlines
and offer a new perspective on what
constitutes a proper line break. TD
Absolute Retracements are a new spin
on Fibonacci retracements which have
spectacularly called some major market
turns.
TD Lines
We know that some lines are more
important than others: the more points
touched, the longer the line, the less
acute the angle, the more important a
line is typically considered. The break
of a line that has been tested many
times and remained unbroken for a
long time and is not climbing sharply is
normally considered an important line
break and a big move is expected. But
Tom Demark says we have missed the
point. He says we have got it wrong.
There are many problems with line
drawing. Two technical analysts may
well draw differing lines. Some consider a small line break allowable. Others
strictly join rising lows to draw an
uptrend and falling highs to draw a
downtrend. There are no 'rules' for
drawing a line. Further, technicians do
not agree on what constitutes a line
break. Some say a one-tick movement
through the lines is a break, others a
close through the line, others a percentage movement through the line and so
on. TD Lines have strict rules for their
construction, importance, the tar- →
Trevor Neil and Tom DeMark
July/August 2005
15
Techniques
from the line itself, traders are then left
high and dry by the break and will be in
a rush to get out of positions that are
now well away from the market. This
will cause a big move after the break.
Figure 1. TD Lines on EUR/USD. The solid lines are Qualified and when they are broken a target is set. Note how on the break of 11/04/05 at 1.2940 an automatic target of 1.31 was set and
reached a few days later. Note also that the line breaks on June 1st and 14th were false. The line
itself went dashed indicating it was Un-Qualified. This happened on the day of the break and
told us it was false. A short term counter trendline break trade could have been taken at these
points. TD Lines can be used on intra-day charts for more trades or weekly charts for a longer
term perspective. All lines are drawn by the software package.
get and when a line has had a good
break and when the break is likely to be
false. This means that a computer can
draw them, set targets when they are
broken and indicate if the break is a
good one that will result in a trend
change or a 'false break' which may
even be tradable against the direction
of the break itself.
Line construction
Instead of drawing lines from left to
right, down across highs or up joining
lows, Tom DeMark suggests looking
from the right for a TD Point to the
next TD Point to the left then extend
the line to the right. TD Points are the
key. A one-point TD Point would be a
bar which has a low below the low of
the prior bar and the subsequent bar.
Imagine a support point that is so powerful that is can stop a fall in the market
when it has gone on for four bars and
caused a rally which extends for four
bars. This is a powerful support point.
Of course if a fall of say six bars is broken and a rally of six bars follows, this
is an even more important support
point. TD Uptrend Lines are drawn by
finding the next TD Point to the left
16
with the same number of falling bars
followed by rising bars. TD
Downtrends are drawn by joining two
TD Points constructed using the same
number of higher highs followed by
the same number of lower highs. His
point is that the shape of the pivot
point that matters in deciding how
important an uptrend or downtrend
line is. The more highs and lows surrounding the pivot point the more significant the resistance point. This
means a line can be drawn using just
two pivot points because the pivot
points themselves are so significant.
Setting targets
Once a line is broken a target is set. The
target is calculated by taking the maximum excursion in the direction of the
trend and measuring that amount from
the break point. The idea is the amount
of overexcitement during the life of
the line will be reflected by proportionate disappointment after the break. So
if the line has been hugged by the market during its life there will be a mild
reaction after the break. But if the market got so excited during the uptrend
that traders pushed the market far away
July/August 20052004
Qualified and Unqualified line
breaks
Conventional technical analysis has
some difficulty in determining when a
line is actually broken and the market is
changing direction. There are too many
'false breaks', an innocent name for a
losing trade. Tom DeMark gives rules
from which it can be determined
whether the break is for real and is
gong to result in a profitable trade. By
first qualifying a TD Line, the trader
reduces the risk that intra-day entry
(going with the TD Line Breakout) will
fail, and, at the same time, increases the
chance of success should the user fade
the disqualified trade (going against the
TD Line Breakout). To be a Qualified
TD Line, price must satisfy one of the
following conditions:
1. The close of the bar prior to
upside/downside
penetration
is
below/above the close two bars ago,
OR
2. The open of the breakout bar is
above/below the TD Line and also
opens above/below the previous bar's
close, OR
3. The difference between the previous
bar's close and its true low/high added
to/subtracted from the previous bar's
close is less than/greater than the TD
Line. A true high is that bar's high or
the close of the previous bar, whichever is greater; a true low is that bar's low
or the close of the previous bar,
whichever is less.
To be a Disqualified TD Line
(dashed line), NONE of the above
conditions exist, and the reverse logic
applies.
TD Lines give traders: an exact a reliable way to draw lines using significant
support and resistance lines; automatic
targets following the break; and an
automated determination of the validity of the break itself.
Techniques
Figure 2. TD Absolute Retracement off of the 24/03/00 S&P500 cash high day's close. The
TD Absolute Retracment levels downside are 944.75 and 764. The 944.74 was the exact 9/21/01
low and the 22/10/02 low fell 2 points short of the objective. See X's on chart.
Absolute Retracements
Just as there are subjective problems in
drawing trend lines, there are also the
same problems when deciding how to
measure Fibonacci retracements. The
problem is choosing the measuring
points. Where to start looking for the
38.2%, 50% and 61.8% levels? We
know these levels can be very effective
but which high or low is the right one
to measure from? Tom demark offers
us an objective way to make these
measurements which does not require
us to guess which price level we are
reacting from.
Tom DeMark suggests taking one of
the two knowns: the all-time high or alltime low and measuring down to zero
or up to the high and dividing these
into 31.2%, 50% and 61.8% levels.
TD Absolute Retracement has an
impressive record of finding important
support and resistance levels that hold
significance for many years. All this
without the trader having to guess the
starting point for the Fibonacci measurement.
Tom DeMark with his TD Lines and
TD Absolute Retracements has given
us two mechanical and objective methods of describing trends, determining
when they are broken, setting targets,
and a way to find important support
and resistance levels without having to
make subjective judgments over the
starting point for measurement.
Trevor Neil is a hedge fund manager at T-Capital in Cape Town. He
was head of technical analysis at
Bloomberg for 4 years before moving to South Africa to form a new
technically traded fund. He offers
institutional training through is
company BETA Group.
www.betagroup.co.uk.
* Jack Schwager
Figure 3. TD Absolute Retracement upside from the close of the day after the October 2002
low. A rally to the 150% TD Absolute Retracment objective at around 1252 would be equivalent
to the rally subsequent to the April 1930 DJIA rally following the 1929 crash and prior to the
huge fall to the 1932 low. That would be comparable to the S&P subsequently declining by half
but very possibly by 61.8%.
July/August 2005
Reference: TD Lines and TD Retracements, please
refer to Thomas R. DeMark, The New Science of
Technical Analysis, New York: John Wiley & Sons,
1994.
17
Techniques
INTERMARKET ANALYSIS:
FORECASTING SHORT-TERM FX RATES
by Marc Zaffran
C
orporates and institutions often seek ways
to predict currency
rates in order to decide when
to hedge or when to take a
position in the market. The
methods they use are typically based on an assessment of
fundamental or technical
analysis.
The forward and futures markets can
also provide a good indicator of rates
for the purposes of corporate planning. However,
their use should not be
confused with mistaken
notions about the accuracy
of their predictions. After
all, market rates are determined by the best
informed participants and
by those participants who
are able to move first, not
by market consensus.
To provide a better
insight into future market
behaviour, the Societe
Generale Corporate and
Investment Banking FX &
Derivatives Sales Team in
London perform analyses
called "Market Behaviour
& Historical Pattern"
(MBHP).
Looking for parallels
Corporate and institutional clients are
constantly asking the sales desk if
recent events in the market are similar
18
to events that have happened before
and, if they have happened before,
what was the impact on the FX markets
that followed.
To provide an answer to these questions, our MBHP analyses essentially
ask two questions of our proprietary
database: Can we find historical parallels for recent significant events in the
markets? If so, what happened next?
The analyses are executed weekly and
are based on events in stocks, bonds,
commodities, energy, and foreign
exchange markets, and sometimes even
economic and political data such as US
payroll figures and general elections.
With this database, which holds data
going back a minimum of 35 years, we
can ask questions such as "when prices
went up in x market and down in y mar-
July/August 2005
ket over a z time period, what happened to ab currency pair in the following week?" The database is sufficiently
deep and sophisticated in its formulation to allow us to ask almost any permutation of question, although limiting
ourselves to a maximum of three to
four parameters tends to yield the most
reliable results.
We generally look for, and publish,
results that suggest a 70% or greater
bias in behaviour. Of the 40 weekly
analyses we have published in the last
year, 29 have been successful.
Correlation
The questions we pose
our database are based
on an understanding or
even a hunch about the
relationships between
markets. Thus performing historical analysis
leads us to highlight
inter and intra FX market correlations.
Correlation does not
imply causation in any
way. In other words, just
because two events are
correlated does not
mean that one causes
another, or has anything
to do with the other.
Correlations deal only
with observed instances
of events. However, a
strong correlation does often warrant
further investigation to determine causation. It is important to understand
how different currency pairs move in
relation to each other.
Techniques
“IT IS IMPORTANT
TO UNDERSTAND
HOW DIFFERENT
CURRENCY PAIRS
MOVE IN RELATION
TO EACH OTHER.”
Two illustrations of MBHP
1. In the one week from the 22nd
October 2004 to the 29th October
2004:
•
•
- MARC ZAFFRAN
Having this knowledge allows you to
effectively diversify and manage your
portfolios. Furthermore, we can tell
from our dynamic correlation charts
how a correlation between two underlings evolves through time, providing
important information about the
volatility of the correlation and of
whether the current correlation coefficient is reflective of a recent trend or
whether it is fundamentally high.
Historical analysis
Essentially, the goal is to identify consistent and reliable patterns (or inconsistent ones) - based on history. Many
currencies don't exhibit any particular
patterns of movement and this is information that is useful in itself, but for
those that do our analysis may provide
some powerful trading opportunities.
MBHP allows us to investigate the
past in a flexible and efficient way,
drawing patterns in the market that are
likely to occur again today. Even
though the analysis is not usually statistically significant - the number of
occurrences we are working with are
often as small as seven - MBHP has a
good track record of highlighting situations which confirm the trend or indicate its imminent reversal.
Marc Zaffran, FX & Derivatives
Sales team, Societe Generale
Corporate & Investment Banking,
London.
EUR/USD gained more than 1%
The Dow Jones Industrial Average
gained more than 2% (269 points)
NYMEX Light Crude Oil future
lost more than 5%
We asked our historical database how
EUR/USD, CABLE and USD/CHF
had performed WHEN:
•
•
•
EUR/USD had been up over the
week &
NYMEX Oil Futures had lost
more than 5% over the same week
&
The DJIA30 had gained more than
2% over the last three trading days?
We found 17 prior occurrences of this
scenario since 1986 and history
showed that after these events:
•
•
•
•
EUR/USD rallied in 82% of the
cases with an average of 1.90%
over the next 10 trading days.
The bearish signal for the dollar
was confirmed on other currency
pairs:
GBP/USD rallied in 70% of the
cases by an average of 1.40% over
the next 10 trading days.
USD/CHF decreased in 82% of
the cases with an average of 1.70%
over the next 10 trading days.
between EUR/USD and aluminium.
Worried about its short term
EUR/USD and aluminium exposure,
our client wanted us to determine
whether or not EUR/USD would
continue its two week downward
trend. Over the same period we saw a
correction in aluminium's six months
bull-trend with contracts trading on
the London Metals Exchange below
their 200-day moving average. Had
we seen such a simultaneous move in
EUR/USD and aluminium before?
We asked our historical database:
How had USD performed against the
Euro and the Yen WHEN:
•
•
•
We found 8 prior occurrences (from
1993 to today) and history showed
that:
•
•
The actual result:
EUR/USD gained more than
1.40% from 29/10/04 to
12/11/04 on a closing bid basis.
2. In early May one of our clients with
a significant exposure to aluminium
asked us to perform an analysis that
would explore the co-movement
July/August 2005
EUR/USD lost in 7 cases out of
8 by an average of 1.00% over
the next 10 trading days and
USD/JPY rallied in 7 cases out of
8 by an average of 1.10% over
the next 10 trading days.
The actual result:
•
•
EUR/USD had been going down
over the previous 10 days &
LME Aluminium prices were up
but below the 200-day moving
average
The US dollar had risen in each
of the prior four trading days
•
While many market participants
were expecting EUR/USD to
head back up, this study allowed
us to tell our client that based on
the
historical
AluminiumEUR/USD relationship, the
EUR/USD downtrend would go
on.
EUR/USD lost more than 2.00%
from 05/05/05 to19/05/05 on a
closing bid basis.
19
Asia and Japan Hedge Fund Directory
2005
Qsjodjqbm!Tqpotps;!
Tfdpoebsz!Tqpotps;!
Techniques
COMMODITY CYCLES:
TECHNICALS AND FUNDAMENTALS COINCIDE
by Dave Hightower
T
he commodity markets
have received a lot of
attention from talk
that long term commodity
cycles are in the process of
turning higher. There has
been so much bullish talk
about commodities that some
contrary players think the
March 2005 high in the CRB
Index marks the end of the
upward cycle. There are a
number of forces coming
together however that should
extend commodity prices further upwards.
Empirical studies suggest that a soft
asset to hard asset rotation typically
lasts about 12-15 years and with the low
in hard assets possibly coming in on the
1998-1999 low in the CRB Index at the
earliest it is possible that generally ris-
ing commodity prices could be in place
until 2010 (see Figure 1). There is also
compelling evidence to suggest that
both longer term fundamental and
technical issues are at work and that
commodity prices won't become
exhausted on the upside very soon. In
fact, it could be a mistake to discount
the combined impact of the "new
money" coming into the commodity
markets and the rampant globalization
of markets in general. Commodity
markets have never seen the type of
money that has begun to flow their way
until recently, and they certainly have
never seen the type of demand patterns
that have been developing as a result of
the rapid economic expansion in China.
In addition to long term technical
cycles (like the Kondratieff wave) turning upward, a number of fundamental
cycles are also positive for commodities. Rising consumption in India and
China alone might present the commodity markets with the most concentrated expansion in demand that has
been witnessed in modern times. In
fact, considering the opening up of
economies that were previously closed
to capitalism and their growth in population since the end of World War II,
we suspect that commodity demand
increases could easily outstrip the
boom of the 1940's and 1950's.
Furthermore, after decades of just-intime inventory buying and a desire to
optimize profitability, many producers
lack the infrastructure to run production at record levels for an extended
period of time. Copper is a good example and Figure 2 shows how stocks are
now heavily depleted in this metal.
Likewise, the oil sector is another prime
example of a commodity market that
has been unable to satisfy world
demand. Even with significant capital
flowing toward energy investments, it is
clear that meeting and exceeding the
world's needs is at best a difficult task.
In some commodity markets, technology has certainly stepped up to help
production keep pace with the →
Figure 2.
Figure 1.
July/August 2005
21
Techniques
Figure 3..
growth in world demand. However, in
markets that are subject to the effects
of nature and, in a sense, weather
cycles, it is possible that meeting growing demand with ever increasing record
production may prove to be too difficult to achieve, creating periodic volatility events. For instance, the grain markets from 1997 to the middle of 2003
saw extremely favourable global weath-
er conditions, but those favourable patterns seemed to come to an end in
August 2004 and have continued to
worsen into the spring and summer of
2005. Therefore, even though genetic
engineering and advances in crop technology have resulted in unprecedented
growth in crop yields, they are still susceptible to the vagaries of weather.
While the world supply of soybeans,
Table 1.
22
July/August 2005
for example, has been able to sustain a
near linear uptrend, world demand has
sometimes grown at an exponential
pace. For example, it is estimated that a
strikingly large percentage of the
income growth in China is spent on
food and food products. China's soybean usage was near 10 million tonnes
in 1981, hit 20 million tonnes by 1998
and is projected to be 43 million tonnes
in 2005.
On top of traditional cycles for commodity markets, commodity cycles can
sometimes be weighted with a natural
cycle which is seasonal in nature. A 91
day cycle has been a useful tool for
technical analysis in the grains and livestock because production seasons (i.e.
spring, summer, fall, winter) appear to
have a significant influence on prices.
For example, grain traders watch for
signs of tops or bottoms around March
21st, June 21st, September 21st and
December 21st. The November 2005
soybean contract appears to have put in
a significant cycle top near June 21,
2004 (628), September 22, 2004 (565
1/2), December 23, 2004 (569 3/4),
March 21, 2005 (606) and the current
contract highs on June 22, 2005 at 754.
In addition to the historical confluence of technical cycles and rampant
globalization, the commodity markets
are seeing an unprecedented influx of
"new money" flowing their way in the
form of commodity fund activity (see
Figure 3). This can be seen in the open
interest figures - in 2004 all time highs
were achieved for nearly every commodity (Table 1). In fact, the money
flowing into commodities since the
beginning of 2003 has the potential to
swamp all but the most substantial
commodity markets.
On April 19, 2005 almost every physical commodity market on the board
rallied in a concentrated fashion and
reports from the trading floor suggested that a single commodity fund was
responsible for sparking the buying
wave. Coincidentally, on that same day
the commodity markets saw an unrelated story concerning the launch of the
Techniques
first exchange-traded fund that was
based on a commodities index.
The importance of this new form of
instrument and the expanding presence
of fund activity in commodities could
have a profound influence on how
prices are established on all major commodity exchanges. While anyone can
thoroughly document the influx of
fund money into commodities, the creation of exchange-listed commodity
specific funds could facilitate an even
more significant flow of money into
the markets, and that in turn could
change a number of historical assumptions. In other words, the potential to
overwhelm various commodity markets
is increasing. In fact, in the coming
environment even if fundament developments prove to be only moderately
bullish or bearish, it is possible that historical assumptions of both low and
high values for specific commodities
will be redefined constantly and that
commodity price volatility will become
explosive.
Kondratieff Wave
The Kondratieff Wave represents an attempt to explain patterns of regular,
structural changes in the world economy. It is named after Nikolai
Kondratieff, a Russian economist working in the 1920s, who saw the capitalist economies as evolving and self-correcting, contradicting the Marxist
notion that capitalism was doomed to collapse. Kondratieff looked at prices
and output in Britain and the US going back to the 1790s. Based on his studies, he determined that economic activity could be measured in "waves" lasting 50-54 years.
The cycle begins with what is called the "upwave" during which prices start
to rise slowly as economic expansion begins. During the course of the
expansion inflation but gradually picks up steam so that after about 25-30
years it is running at very strong rate. This sets the economy up for a severe
recession that is longer and deeper than anything that the economy had
experienced during the upwave. Eventually prices stabilize, the economy
recovers and a second wave of expansion begins, but the growth is nothing
like what was seen during the long upwave. This persists for about ten years,
but growth is anemic and the economy never reaches its previous dynamic
state. This secondary plateau ends with a sudden shock, such as a financial
panic or a stock market crash, and the economy moves into its secondary
correction phase, which is far deeper and which leads a depression and
deflation.
There are different theories as to what causes this wave. One theory is that
some key innovation drives a period of economic realignment and expansion, such as the industrial revolution in the 1800s, electronic and motor
vehicles in the early 1900's and the information revolution in the latter part
of the 20th Century. This realignment either brings about or is necessitated
by the wave action. Another theory is that a new wave begins every three
generations or after that the group that experienced the previous wave dies
off.
Determining at what phase of the cycle the economy is currently becomes
a matter of interpretation. Some students of Kondratieff Wave theory have
put the beginning of the mid-20th Century a wave at 1940 (the end of the
Great Depression and start of the expansion at the beginning of World War
II), while others believe the wave began around 1949 with the post-war
expansion. Using these dates as a point of reference puts the start of the
next upwave at 1990-94 or 1999-2003, respectively. On the other hand, there
has been further theorizing the Kondratieff Wave may have increased in
length to 60 years or more as life spans have increased. Using a 60 year wave
length would move the start of the next upmove to either 2000 or 2009.
Dave Hightower is president of The
Hightower Report, specialists in
futures research and forecasting
(www.futures-research.com).
Clearly, the timing of the Kondratieff wave is not an exact science.
However, what is consistent is that in order for the theory to hold, the market will have to face some sort of crisis prior to starting the next upmove.
Was that crisis the bursting of dot.com bubble or is there something else in
store?
July/August 2005
23
Techniques
THE PROPER USE OF GANN ANGLES
by Gary Stafford
M
ost people probably have software
that allows them to
look at Gann Fans or Angles.
However it might surprise
you to find that the software
is probably doing it completely wrong. What we read
regarding the principles from
the Gann literature available
is not necessarily what is
being applied.
One of the worst examples I have
come across from a major software
package is where you put a trendline
on a chart and the program then plots
the other angles based on the original
trendline. This is a common and fundamental mistake and bears no resemblance to what Gann actually said.
Gann was looking for geometric relationships between price and time. The
trendline/Gann fan completely ignores
this principle.
Even if the software adheres to what
Gann literally said you will still being
doing it wrong in most cases.
As an example, we are going to look
at the FTSE 100 just before the all time
high in 2000. Figure 1 shows a Gann
fan, but based on what Gann literally
said. As you can see from the chart, I
have angles spreading out from the
October low and have drawn the daily
1x1, 2x1, 4x1, 8x1, 16x1. However, as
you can see, these angles do not work.
Now we are going to do this correctly. Every stock, commodity, currency
etc works in its own unique way, the
skill is in finding how it works.
For instance the Dow Jones works
slightly differently to the FTSE 100,
the mathematical principles do not
24
Figure 1.
Figure 2.
July/August 2005
Techniques
change. The FTSE does not work on
daily angles. Gann in his writings called
it "The Master Time Factor". Looking
at Figure 2, the correct angle series is
put on using the correct time factor.
The angles Gann talked about are as
follows;
1x1, 2x1, 4x1, 8x1, 16x1, 32x1, 64x1,
128x1, 256x1 etc…
In Figure 2 we have put the same
angle progression on but with the correct Master Time Factor. As you can
see the 512x1 angle did not support the
price, the next angle below 512x1 is the
256x1 which gives support in February
1999. Remember that this angle is not a
trendline, it is a mathematically calculated angle which is drawn from the
October low, almost 5 months in
advance. The next angle drawn from
the same point is half of the 256x1,
which is the 128x1. This gives us the
low in August 1999, the next angle for
the future is the 64x1 which works but
is not shown on the chart.
There are other angles
Quoting directly from Gann's writings,
he states "After 50 years of research,
tests, and practical applications I have
perfected and proven the most important angles to use." These numbers are
shown above, i.e. 1x1, 1x2, 1x4 etc. The
important thing here is that they are not
the only angles that work.
Figure 3 shows the angles halfway
between the major angles.
The first two steepest angles on
Figure 2 are the 512x1 and the 256x1.
The halfway angle between these two is
the 384x1 angle. Half of this then
becomes a 192x1, then 96x1, then 48x1
and so on.
Looking at Figure 3 you can see that
the 384x1 gives us the low in December
1999, the 192x1 give us the low in June
2000 and the 96x1 gives us the exact
low in October 2000. Now compare
this chart with the angles in Figure 1
and you can see how easy it is to get
Gann angles completely wrong.
→
Figure 3.
“THE UNDERLYING PRINCIPAL BEHIND
[GANN]…IS THE FACT THAT SPECULATIVE
MARKETS ARE MATHEMATICAL IN NATURE.”
- GARY STAFFORD, GANN MANAGEMENT
Figure 4.
July/August 2005
25
Techniques
“AFTER 50 YEARS OF
RESEARCH…
I HAVE PERFECTED
AND PROVEN
THE MOST
IMPORTANT ANGLES
TO USE.”
- W.D. GANN
Figure 5.
Another example: GBP/USD
Figure 4 is a weekly chart showing the
1985 major low, when the dollar almost
reached parity. It was from this low we
have drawn a 1x12 angle which gives us
the major low in 1986 and 1987. This
angle is what we call a half angle, the
one between Gann's major angles of
1x8 and 1x16. As this angle has worked
i.e. supported the price, the next angle
we would watch with great interest is
twice the 1x12, which is a 1x24. As you
can see, this angle gives you all the lows
in 1989. The next angle we would
watch with interest is the 1x48, which
you can see in Figure 5.
In Figure 5 we have changed to a
monthly chart. As can be seen the 1x48
did not support the price, hence this
did not work. From the logic so far you
would expect us to use the next angle
twice 1x48 which would be 1x96.
However I found out many years ago
that if an angle does not work, it can
change what will work in the future.
To recap, the major angles are:
1x1, 2x1, 4x1, 8x1, 16x1, 32x1, 64x1,
128x1, 256x1 etc…
We will call this series 1.
26
The angles halfway between series 1
are:
3x1, 6x1, 12x1, 24x1, 48x1, 96x1,
192x1 etc…
This is called series 2.
We can see that GBP/USD has been
working on series 2. However the 1x48
did not work. This means that we will
change to series 1 angles for the future.
So the next angle below the 1x48 is the
1x64, which is a series 1 angle. From
the chart you can see that it gave us the
exact low made in May 1996 - this angle
was put on the chart eleven years ago
from the major low in 1985.
The 1x64 has worked so we stick with
series 1. The next angle is therefore the
1x128 which is twice 1x64, and you
should be able to see the bottom was
the exact low in June 2001, 16 years
from the low in 1985.
Mathematical markets
The underlying principal behind these
rules - rules which I have developed
over the last 14 years - is the fact that
speculative markets are mathematical in
nature. Recognising this fact should
give traders a fighting chance in profiting from the markets.
July/August 2005
Gary Stafford is technical director at
Gann Management Limited in the
UK. (www.gann.co.uk)
Please note there are a series of videos
to accompany this article. These can be
accessed online at www.gann.tv (click
on Free Gann Video button)
Techniques
SETTING STOP-LOSSES
USING PRICE VOLATILITY
by Cynthia A. Kase
B
y accounting for
volatility, the variance
in volatility, and
volatility
skew,
Kase
DevStops provide a statistically sound basis upon which
to place stop-loss and stopand-reverse orders.
The idea of using volatility-based
stops first occurred to me after I read
Welles Wilder's New Concepts in
Technical Trading, published in 1978.
In this book, Wilder sets out a stop and
reverse system centred around a measure called "True Range". It was my
reading of that book along with the
burst in volatility that took place in the
oil market in 1990 that led me to investigate the use of True Range in setting
the size of stops, and ultimately to the
development of the Kase DevStop system in 1991.
The DevStop
A number of trading techniques use
average True Range (ATR) values with
fixed multipliers to determine where to
place stop and reverse orders - for
example, stops might be set at 3 x ATR
(see Box for definition of TR and
ATR). Yet while True Range is proportional to volatility and using such an
approach is an improvement over a
fixed-value stop, the use of the average
(ATR) is insufficient to truly capture
market behaviour.
Think of two populations, Population
A and Population B. Population A is
comprised of chorus girls and has an
average height of 5'10" - the shortest is
5'9" and the tallest 5'10½". Population
True Range
True Range (TR) is the full price range of a period and is defined as the largest
value of the following three calculations:
1. TR = Highest price of current period("H") - Lowest price of current
period("L")
2. TR = Highest price of current period("H") - Closing price of
previous period("C")
3. TR = Closing price of previous period("C") - Lowest price of current
period("L")
Average True Range (ATR) is the simple average of True Range over the past n
periods or an exponential moving average.
Source: The Encyclopedia of Technical Market Indicators, Colby, 2003
B is comprised of professional basketball players and their elementary school
aged children, and has an average
height of 5'10" - the shortest is 3'2" and
the tallest 6'11". Obviously, even
though both populations have the same
average height, a higher value is needed
to be equivalent to, say, the 95th percentile.
The same is true for a market. A stop
that is going to perform optimally must
consider the variability of range, not
just the average. So I developed the
DevStop, where "Dev" stands for standard deviation around the mean.
The value of the DevStop is calculated
as follows:
DevStop Amount = average(2 x TR, n)
+ stddev(2 x TR, n)
For the simple reason that True Range
values based on single bars are too
small, notice the DevStop employs the
True Range of two bars (2 x TR),
July/August 2005
which we call the True Range Double
(TRD).
We can use this calculation to determine the points at which the stops
should be set. For a normal bell curve,
for example, a stop placed at the one
standard deviation level has an 85%
chance of being hit and a stop placed at
the 1.65 standard deviation level has a
95% chance of being hit.
Volatility skew
The problem with using a normal bell
curve is that it doesn't accurately reflect
how range is really distributed. When
we look at range as a proxy for volatility (which is a logarithmic term) the
range is more or less log normally distributed as opposed to normally distributed. Thus it has a right hand skew.
Figure 1 shows the distribution of the
daily TRD values for the Live Cattle
contract from August 1969 to August
2004. It can clearly be seen that there is
a severe right hand skew. Indeed a full
two-thirds of the data are below →
27
Techniques
Figure 1. Distribution of TRD
the average TRD of 1.4 cents.
Therefore a correction of the normal
standard deviation levels must be made
to account for the skew.
Back in the early 1990's when we first
developed the DevStop we performed
research on 100 bar subsets of a total
of 10,000 bars of 10 minute British
Pound data to estimate the skew corrections that would be needed to equate
to 1.0, 2.0 and 3.0 standard deviations.
We found that the one standard deviation level, on average, did not require a
correction, and that the two and three
standard deviation levels needed to be
corrected by about 10 percent and 20
percent respectively, that is, to 2.2 and
3.6 standard deviations over the mean.
Figure 2 shows an example of the
DevStops on the daily chart for the
October 2004 Lean Hogs contract,
where each line has been calculated as
follows:
Warning Line =
average(TRD, n) + 0stddev(TRD, n)
Dev 1 =
average(TRD, n)+1stddev(TRD, n)
Dev2 =
average(TRD, n) + 2.2stddev(TRD, n)
Dev3 =
average(TRD, n) + 3.6stddev(TRD, n)
28
The computer does not know, of
course, if one is long or short, so a
moving average crossover system is
embedded into the front end of the
code. If the fast moving average, normally defaulted to five (5) is above the
slow, defaulted to 21, then the computer assumes you are long and trails the
stop from the highest high since the
moving averages crossed. If the fast is
below the slow, the DevStops trail from
the lowest low. The stops are flipped
from below the data to above the date
when the averages flip from long to
short and vice versa. But the front end
could be programmed for any entry
system.
Adjusting for skew
Recently I decided that it was time to
update the original study. Rather than
using 10-minute bars on one contract,
we decided to test daily data over a
range of physical commodity and
financial futures from contract inception to September 2004, amounting to
more than 50,000 data points. In addition, we also evaluated the data in 250
bar sets which give a more representative population than the 100 bar sets,
stepped backward in 100 bar increments. Table 1 lists the study data.
The first question I addressed as I
looked at the results is how the data
July/August 2005
compares with our prior adjustments,
i.e.: no correction at the one standard
deviation level, a 10 percent correction
at the two standard deviation level and
a 20 percent correction at the three
standard deviation level, resulting in
one, 2.2 and 3.6 standard deviations
over the mean for the DevStops. Table
2 shows the percentile rankings and
number of standard deviations for a
normal bell curve versus the average
and median values for the actual data.
I found that the warning line, which
is set at the average of the data, or zero
(0.0) standard deviations over the
mean, actually captures about 60 percent of the observations. This can be
seen by looking across the row marked
"60" in bold under the percentile column. Where a normal bell curve would
require a setting of 0.3 standard deviations over the mean to be at this point,
the actual data only requires a setting of
zero.
At the 85th percentile the one standard deviation setting for normal bell
curve corresponds to about 0.95 for
the actual data, which means that a set-
Table 1. Data used to September 13, 2004
Table 2. Normal Bell Curve vs. Actual Data
Techniques
“A STOP THAT IS
GOING TO PERFORM
OPTIMALLY MUST
CONSIDER THE
VARIABILITY OF
RANGE.”
CYNTHIA KASE
Figure 2. DevStops on October 2004 Lean Hogs Daily Chart
ting of 1.0 for DevStops is roughly
equivalent to a very slightly higher percentile reading at about the 86th percentile. At the 90th percentile, marked
in italics, the curves cross from the
actual data being at a standard deviation
level slightly less than that generated by
the bell curve to slightly more. So using
1.28 for a DevStop setting would on
average generate a stop at the 90th percentile, the same as a bell curve.
At two (2.0) standard deviations, a
correction of 20 percent would be
needed to correspond with the actual
data, the same as has been used all
these years for the three (3.0) standard
deviation stop. Rather than correspon-
ding to 2.0 standard deviations over the
mean at the 97.5 percentile, a setting of
2.2, on average, corresponds with
about a percent lower, that is, with the
96.5 percentile. So the use of 2.2 rather
than 2.4 has been accurate within one
percent of its original intent. As noted,
the use of 3.6 - a 20 percent correction
to 3.0 - is almost exactly correct.
Finally, I looked at the degree of variation among the futures contracts studied (Table 3). While a fairly wide range
of variation between each data set was
found (each data set being 250 bars),
there is little variation among futures
contract. The average standard deviation for the results by data set is 0.13
Table 3. Average Standard Deviation Results by Commodity vs. Percentile of Data
July/August 2005
(results not presented here), but the
average variation by commodity is only
0.03, less than 25% that of the variation
by data set. This means that the range
of variation by data set is about the
same no matter what is traded.
In summary, the methodologies that
were used a dozen years ago to develop
the Kase DevStop are still valid. The
one standard deviation level used for
Dev1 and 2.2 standard deviation level
used for Dev2 are within a percentile of
the settings that were the original
intent, specifically the 86th vs. 85th percentile for Dev1 and 96.5 versus 97.5
for Dev2. For Dev3, the original setting
at 3.6 is right on the mark. There is
some variability around the mean values found, but there is insignificant
variation by commodity. That it,
regardless of the commodity studied,
the results come out about the same, so
no adjustment must be made to the
DevStops based on the futures contract
traded.
Cynthia Kase is president of Kase
and Company, Inc.
(www.kaseco.com). Kase DevStop
is available on CQG, TradeStation,
eSignal, Aspen Graphics and DTN
ProphetX.
29
presents
Technical Analysis in the Commodity and FX Markets
A series of one day seminars designed to equip traders, analysts,
hedge funds and fund managers with the most effective and up-todate charting techniques to trade the commodity and FX markets.
With contributions from leading market analysts, the seminars will
present a range of studies and market views to enhance both short
and long term trading returns.
Cape Town, South Africa
Johannesburg, South Africa
Tuesday August 23rd, 2005
Venue: ArabellaSheraton Grand Hotel
Thursday August 25th, 2005
Venue: Intercontinental Sandton Sun & Towers
Topics covered include:
▪ Intermarket analysis
▪ Maximising the value of technical indicators
▪ Using Fibonacci to trade grains
▪ Assessing the rand and gold markets
▪ Using DeMark to generate buy & sell signals
▪ Commodity markets using candlestick charts
▪ Using Fibonacci to trade grains
▪ Assessing the rand and gold markets
▪ Using DeMark to generate buy & sell signals
David Sneddon,
Pieter Van Wyk,
Kevin Edgeley,
CSFB
Commodity Investment Goldman Sachs
Services
Judy Padayacee,
ABSA Bank
Paddy Osborn,
TraderMade
Register online at www.ta-conferences.com
Jeremy Goldwyn,
Sucden
Sponsors
Registration fee = £399 (Dubai £599)
£100 discount when you register early
How to Register
1. Go to www.ta-conferences.com
2. Call +44 (0)20 7833 1441
3. Email events@technicalanalyst.co.uk
Dubai, UAE
to: events@technicalanalyst.co.uk
London,
UK
Tuesday September 27th, 2005
Venue: Intercontinental Dubai Hotel
Tuesday October 25th, 2005
Venue: Trinity House EC3
▪ Trading base metals
▪ Using volatility bands as an overlay on oscillators
▪ Assessing the outlook for precious metals & crude
▪ Applying behavioural finance to the oil market
▪ Using Fibonacci
▪ Commodity currencies
▪ Trading base metals
▪ Intermarket analysis.
Thomas Anthonj,
ABN Amro
John Noyce,
Citigroup
Shaun Downey,
CQG
Robin Griffiths,
Rathbones
Trevor Neil,
T-Capital
Register online at www.ta-conferences.com
Ron William,
IDEAglobal
Widen your horizons
Choosing a dynamic and innovating bank provides you with new perspectives and widened
horizons, to help you achieve your ambitions today and tomorrow. With SG Corporate
& Investment Banking all our clients - from corporate clients and financial institutions to public
sector clients and investors - benefit from sound analysis, reliable advice and the best financial
solutions. With a growing worldwide leadership in our areas of excellence, we will do our best
to respond to your financing, capital management or investment requirements by combining our
areas of expertise, our innovation and our cross-product approach ■ SG CIB, your partner in Euro
Capital Markets, Derivatives & Structured Finance
■
www.sgcib.com
Societe Generale is authorised by Banque de France and the Financial Services Authority, and is regulated by the Financial Services Authority for conduct of UK business.
EURO
CAPITAL
MARKETS
DERIVATIVES
STRUCTURED
FINANCE
Interview
VOLATILITY ARBITRAGE STRATEGIES
Rami Habib, quantitative analyst at Fimat in London, discusses
volatility arbitrage fund strategies and performance across the hedge
fund industry.
TA: Volatility arbitrage remains something of a specialized
area within the hedge fund community. What is Fimat's
involvement in the market and your role in the firm?
RH: We are in the Prime Brokerage group of Societe
Generale where I look after the quantitative analysis of
hedge funds for investors. Our product is the FVAM (Fimat
Volatility Arbitrage Median) which is an equally weighted
portfolio of volatility arbitrage funds. This is part of our
ongoing research at Fimat where we study the returns of
fund managers running volatility arbitrage funds. The
FVAM was created as a performance benchmark for the
hedge fund market - it is not an active fund as such.
TA: What funds make up the FVAM?
RH: There are seven volatility arbitrage funds within the
FVAM: er Global, Lynx, Quadix, SGAM, Shooter MultiStrategy, Titan Global and Turtle. These hedge funds treat
volatility as an asset class of its own.
TA: How widespread are volatility arbitrage funds within
the hedge fund industry?
RH: Over the last few years we are seeing more and more.
The first volatility arbitrage funds emerged in Europe early
in 2002 and slightly earlier than that in the US. Fimat now
know of about 20 volatility arbitrage managers around the
world.
TA: Does this mean global funds under management have
reached significant levels?
RH: It is difficult to say because volatility arbitrage is traded in many multi-strategy funds but as for pure volatility
arbitrage managers, it is probably somewhere in the region
of $3 billion globally.
→
July/August 2005
Rami
Habib
33
Interview
TA: Why trade volatility? What are the return characteristics
of volatility arbitrage funds?
tive value trades have been popular. In these markets,
many traders are uncomfortable with selling volatility as
the profits are small but the risks are large. Some traders
are also avoiding buying volatility as it has remained low
over this year.
RH: Volatility itself has some attractive characteristics:
• It is independent of the direction of the underlying price
move.
• It is uncorrelated to the underlying securities.
• It increases when uncertainty increases
• It is mean reverting
• When most asset classes go down, volatility tends to go
up
Volatility based strategies can offer uncorrelated risk-return
profiles that most alternative investment and long-only
portfolios require. Returns are also created without the
credit risk element that is inherent in conventional trading
such as convertible arbitrage.
TA: What are the typical strategies traded by volatility arbitrage funds?
RH: There are many volatility trades that arbitrage funds
can enter into. Each fund manager will use his own techniques to extract value from movements in volatility, typically by trading short-term, liquid and mostly exchange traded
financial securities and derivatives.
• Specific volatility trading strategies include directional
trading where managers can take directional views on
volatility, i.e. take an outright view (long or short) on the
level of volatility (strikethrough). The trader may believe
that options market expectations are wrong or that
imbalances in supply and demand for options have driven implied volatilities to levels where they no longer
reflect consensus expectations of future realised volatility. Some traders can switch from net long to net short
volatility positions depending on their views on the markets. Others will maintain a net long or short bias believing the market systematically misprices tail risk.
Buying options in general means buying volatility. Delta
hedging should be profitable if realized volatility minus
transaction costs is higher than the implied level at which
the options were purchased. Selling options in general
means selling volatility. Delta hedging should be profitable
if realized volatility minus transaction costs is lower than
the implied level at which the options were sold.
• Relative value volatility trading involves taking relative
value positions on volatility. The trader may take a view
that the options on one stock or index are mispriced relative to another. With volatilities remaining low, the rela-
34
• Term structure is where the trader trades on the shape
of the implied volatility term structures.
• Skew trading involves a strategy of trading the relative
levels of implied volatilities across strikes - across the
implied volatility skew structure.
• Dispersion/correlation trades. Dispersion trading allows
traders to profit from price differences using index
options and offsetting options on individual stocks.
Dispersion trading can be explained by the fact that historically index volatility has traded rich, while individual
stock volatility has been fairly priced. The dispersion
strategy typically consists of short selling options on a
stock index whilst simultaneously buying options on the
component stocks. Timing is key to a successful dispersion trade. It is also important to carefully select the basket of stocks for the offsetting dispersion basket. At the
simplest level they should account for a large part of the
index to keep the net risk low, but at the same time it is
crucial for the trader to make sure that they are buying
cheap volatility.
• Finally, a new area that has only recently been looked at
is cross-market volatility trades. Historically, volatility
traders have been heavily focused on equities but we are
now seeing hedge funds looking at other asset classes for
volatility trading opportunities. For example, traders are
looking at the correlation between currencies and equities or at taking volatility positions in the commodity
markets. While, there won't be volatility in all the markets all the time but there should be volatility is some of
the markets some of the time.
TA: How have the strategies traded by fund managers
changed over the last few years?
RH: In the late 1980s and early 1990s the focus was on
theta and directional macro strategies using options instead
of stop losses. By the mid-1990s this had shifted to classical
volatility arbitrage with delta hedging. The late 1990s and
early 2000 saw volatility arbitrage strategies trading across
the term structure, between plain vanilla and exotics, using
volatility and variance swaps and more recently we are seeing cross market volatility trading.
TA: What are the risks associated with volatility arbitrage?
July/August 2005
Interview
RH: One of the risks associated with vol trading is the
impact of changing volatility, this can affect the performance of delta hedging. This is why it is important for traders
to monitor the vega as well as delta and gamma. The vega
of an option is the rate of change of the price of the
option with respect to the volatility of the underlying asset.
For example, if everything remains constant and the market's view suddenly changes that future volatility is going to
fall, then the option price will sit on a lower curve. This can
be good or bad news for the volatility player. Whatever the
situation, the portfolio will suffer an immediate mark to
market loss due to the option price fall. However, if the
underlying price is near the exercise price then the increase
in gamma can actually mean that there is more scope for rehedging than before.
closing of positions caused a liquidity crisis. In 1998 we had
the Russian debt default and the Asian crisis which produced implied vol levels of 50%.
From 1999 to 2001 there was the internet boom when
implied volatilities ranged between 20-30% and 9/11 saw a
spike at 45%. After that the Worldcom and Enron accounting scandals and the Iraq War saw vols exceed 50% again.
From 2003 to 2004 they declined rapidly as a result of low
activity in the markets and a lack of major events.
Moreover, interest rates were falling and there was renewed
corporate balance sheet strength.
TA: How have implied volatility levels changed in the equity
markets over the past decade or so? Have recent volatility
levels been unusually low?
RH: Like any strategy with an arbitrage component the
returns may diminish as the number of factors increase. A
crowded trade remains a crowded trade. However, there is
ample capacity for asset class, instrument and geographical
diversification in the years to come.
RH: Current equity market implied volatility is sitting
around 11-13%. By comparison to recent memory, this is
historically low. but if one looks back further the current
levels are above 'the 91 - 97' levels.
If one goes back to 1990 during the global recession,
volatility for the S&P 500 index ranged from 20%-35%.
During this period interest rates were high and low profits
meant that debt coverage was low and corporate leverage
was high. Added to this is the lead up to the first Gulf War
which resulted in implied volatilities being highly priced.
Between 1991 and 1997 there was a market recovery.
Interest rates were decreasing and this period was uneventful. As such implied volatilities spent a long period of time
at low levels, around 10-15%. From 1997 to 2003 there
were several events that resulted in spikes and high levels of
implied volatilities. 1997 saw LTCM's collapse where the
“PURE VOLATILITY
ARBITRAGE MANAGERS
PROBABLY HAVE
SOMEWHERE IN THE
REGION OF $2 BILLION
UNDER MANAGEMENT
GLOBALLY.”
- RAMI HABIB,
FIMAT INTERNATIONAL
July/August 2005
TA: Do you see returns and opportunities in volatility arbitrage diminishing because of the increased number of managers?
TA: What have the recent returns been like for volatility
arbitrage funds?
RH: Since the summer of 2002, volatility in the equity markets has been in decline. As a result of the declining implied
volatility levels and the fact that implied volatilities were
trading above realized market volatility levels, many of the
long biased volatility hedge funds experienced some losses.
Some net short volatility funds managed to generate consistent returns over this period, however, care needs to be
taken as volatility has the tendency to spike quite rapidly.
Losses can be experienced as a result of prices gapping and
the market becoming illiquid at exactly the wrong time for
option sellers.
TA: What is your volatility outlook for these markets?
RH: Volatility is thought to be mean reverting. It is heavily
dependent on consumer demand, interest rates and the level
of corporate investment leverage. As such, volatility could
easily remain below 20% for quite a while. Consequently,
the strategies that are expected to perform in this environment are the long/short relative value volatility traders, the
macro volatility traders and the dispersion traders in some
of the non-arbitraged markets.
Rami Habib is quantitative analyst within Alternative
Investment Solutions at Fimat International Banque
SA (UK Branch), in London.
35
Subject Matters
IDENTIFYING COMMODITY CURRENCIES
by Paul Cashin, Luis Céspedes, and Ratna Sahay
Which currencies can we accurately call "commodity currencies?"
F
or decades, economists
have tried with little success to model long-run
movements in real - that is,
adjusted for inflation exchange rates. Almost all of
the studies have focused on
industrial countries, trying to
pinpoint whether fundamentals
such as government spending,
current account imbalances,
and differences in productivity
and interest rates hold the key
to explaining exchange rate
movements. But the results
have been disappointing,
with many models that are
based on fundamentals
failing to provide a convincing explanation of the
behavior of real exchange
rates in industrial countries.
this relationship for a handful of commodity-exporting industrial countries,
such as Australia, Canada, and New
Zealand. But the biggest hurdle in
extending these studies to developing
countries has been the lack of countryspecific data on commodity export
prices.
That is why we undertook a study of
the relationship between the real
exchange rate and real commodity
prices for all commodity-dependent
economies. We asked the question: Do
real commodity prices and real
exchange rates move together?
In contrast, studies of the behaviour of developing country real
exchange rates are scarce. The
few studies that have examined
the determinants of these rates
have focused largely on Latin
America and have emphasized Figure 1.
the role of terms of trade movements in driving the real exchange rate. Identifying commodity
However, a natural assumption for currencies
developing countries is that fluctua- Our study was based on the constructions in real commodity prices have the tion of new monthly indices of nationpotential to explain a large share of al real commodity export prices and the
changes in real exchange rates, given gathering of monthly real exchange
that so many of these countries are rate data for 58 commodity-exporting
highly dependent on commodities — countries for the period January 1980
in some cases, a single commodity — to March 2002. Each country's (nomifor the bulk of their export revenues. nal) commodity export price index is a
Indeed, several studies have explored geometric weighted average of world
36
July/August 2005
prices for 44 individual nonfuel commodities using country-specific export
shares (averaged over 1990-99) as
weights.
The 58 commodity-exporting countries include 53 developing countries
and 5 industrial countries, all of which
rely on commodity exports for a major
share of their export income. Indeed,
during the 1990s, the cross-country
mean share of total export receipts
derived from primary commodity
exports was about 48 percent.
Commodity exports typically exceeded
50 percent of the total exports of several sub-Saharan African countries, especially Burundi (97 percent), Madagascar (90 percent),
and Zambia (88 percent).
The share of primary commodity exports in total exports was
quite high even for the industrial
countries (Australia, 54 percent;
Iceland, 56 percent). In addition,
many countries remain overwhelmingly dependent on export
receipts from their dominant
exportable commodity — the
dominant exportable exceeded 90
percent of commodity export
receipts in Dominica (bananas),
Ethiopia (coffee), Mauritius
(sugar), Niger (uranium), and
Zambia (copper).
Armed with the real exchange rate
and real commodity export prices for
each country, we then checked to see
whether these two series displayed a
close relationship. We found that, for
many countries, such as Australia and
Burundi (both of which have flexible
nominal exchange rates), this was
indeed the case (see Figure 1), while
others appear to display a relationship
once a onetime movement in the real
Subject Matters
exchange rate (such as the 1994 devaluation for the CFA franc zone countries
of Mali and Togo) is accounted for.
Next, we used regression analysis to
formally examine whether there was a
stable, long-run relationship between a
country's real exchange rate and the real
price of its commodity exports - in
other words, which of the commodityexporting countries qualified as having
"commodity currencies"? In the analysis, we allowed for a structural shift in
the relationship between the two series
to account for onetime movements in
either series (which typically involve
rapid movements in the nominal and
real exchange rate). We found just such
a long-run relationship for 22 of the 58
commodity-exporting countries (see
Table 1).
It is not unexpected that sub-Saharan
African countries, given their dependence on commodity exports, account
for half the commodity-currency countries. Moreover, for these 22 countries,
over 80 percent of the variation in the
real exchange rate can, on average, be
accounted for by movements in real
commodity prices alone — a surprisingly strong result. As for those commodity-exporting countries where such
a long-run relationship could not be
found, it is likely that factors other than
real commodity prices played a key role
in real exchange rate movements.
How large an impact do real commodity price movements have on the
real exchange rates of commodity-currency countries? We found that the
elasticity typically ranged between 0.2
and 0.4, with a median of 0.38. Thus, a
10 percent drop in the real price of the
commodity exports of countries with
commodity currencies is typically associated with a 3.8 percent depreciation
of their real exchange rate. Further
analysis also indicated that, when deviations from the relationship between
exchange rates and commodity prices
occurred in countries with commodity
currencies, they were caused primarily
by changes in real commodity prices.
SubSaharan
Africa
AsiaPacific
Middle
East and
North Africa
Western
Hemisphere
Europe
Burundi
Cameroon
Central African
Republic
Côte d'Ivoire
Ethiopia
Ghana
Kenya
Madagascar
Malawi
Mali
Mauritania
Mauritius
Mozambique
Niger
Senegal
South Africa
Sudan
Tanzania
Togo
Uganda
Zambia
Zimbabwe
Australia
Bangladesh
India
Indonesia
Malaysia
Myanmar
New Zealand
Pakistan
Papua New
Guinea
Philippines
Sri Lanka
Thailand
Morocco
Syrian Arab
Republic
Tunisia
Turkey
Iceland
Argentina
Norway
Bolivia
Brazil
Canada
Chile
Colombia
Costa Rica
Dominica
Ecuador
Guatemala
Honduras
Mexico
Nicaragua
Paraguay
Peru
St. Vincent and
Grenadines
Suriname
Uruguay
Table 1. Identifying the commodity-currency countries. Countries found to have commodity
currencies are in bold. Two-fifths of commodity-exporting countries have commodity currencies: there is a long-run relationship between their real effective exchange rate and their real commodity export prices. Source: Cashin, Céspedes, and Sahay, 2002.
July/August 2005
Following a movement in commodity
prices, it is typically the real exchange
rate that then adjusts to restore its longrun relationship with real commodity
prices.
A handy crystal ball
Our study found evidence in support
of the co-movement of national real
exchange rates and real commodity
prices in a group of commodityexporting countries. For these commodity-currency countries, the world
price of their commodity exports has a
stable and important effect on their real
exchange rate. This empirical regularity
is surprisingly robust, given the repeated failure of previous attempts to use
models that are based on fundamentals
to explain exchange rate movements.
For policymakers in commodityexporting developing countries, understanding the effects of commodity
price movements on exchange rates
should be of great interest in guiding
the conduct of monetary and exchange
rate policies, particularly as such countries liberalize their capital markets and
increase the flexibility of their
exchange rate regimes. For countries
with commodity currencies, commodity prices are the key determinant of
their real exchange rate and can be used
as a benchmark in determining when
exchange rates have deviated excessively from their equilibrium value.
Paul Cashin is a Senior Economist
and Ratna Sahay is an Assistant
Director in the IMF's Research
Department. Luis Céspedes is an
Economist in the IMF's European I
Department.
Extracted with permission from Cashin, P., Céspedes,
L. and Sahay R., (2003), "Commodity Currencies",
Finance and Development, Vol 40, No. 1, IMF.
37
Software
Dresdner Kleinwort Wasserstein's
PRICE INFORMATION ADVANTAGE
B
anks and financial institutions are increasingly
looking to leverage
their in-house market research
in the most effective way possible. The growth of the hedge
fund industry and competition
amongst prime brokerage services means that greater value is
now attached to research output. This has meant that more
institutions are now developing
their own proprietary research
tools in order to give their
front- end services an edge over
the competition.
At Dresdner Kleinwort Wasserstein
in London, Max Knudsen, head of
capital markets technical strategy, has
developed PIA - Price Information
Advantage. PIA is DrKW's proprietary interactive sentiment forecasting service designed to optimise the
timing of trade execution. The service provides technically derived timing and trading recommendations via
a daily interactive email service. PIA
is based on an assessment of market
sentiment in the equity, FX and interest rate markets made by Knudsen
and his three man team. Sentiment is
assessed using a combination of traditional charting techniques including
Bollinger Bands, momentum and stochastics.
Knudsen says the role of his team
is to translate their assessment of
market sentiment (majority bullishness and bearishness) in each market
and to convert this into specific trading and timing advice consisting of a
forecast for the direction of sentiment and the most important prices.
"This is 75% a statistical and technical
process" says Knudsen, "with the
remaining 25% drawn from practical
market experience. Each member of
my team has around 20 years →
PIA Team, from left: Pavel Gronbjerg, Max Knudsen seated, Steve Lucas, Alan Collins
July/August 2005
39
Software
Fixed income/MM
10-yr Treasury
BOBL
Bund
Euribor
Eurodollar
Schatz
Short Sterling
T Bond
Eurodollars
FX
Equities
EUR/GBP
EUR/JPY
EUR/USD
GBP/USD
USD/CHF
USD/JPY
FTSE
DAX
S&P 500
Eurostoxx
Nasdaq
Nikkei
Table 1. PIA market coverage
market experience in either sales or
trading and I consider this crucial in
being able to accurately assess market
sentiment."
PIA was developed in response to
comments from clients that they
received too much research, much of
which looked indistinguishable.
Simplicity seems to be the key to
PIA. There is very little text; just
straight forward recommendations of
when to buy or sell and the 3 most
important support and resistance lev-
els. "When we started developing
PIA 5 years ago" says Knudsen, "it
became clear that clients felt they
were already receiving too much
research. Our goal was to design a
research application that was both
reliable and very easy to interpret and
use".
PIA covers 6 major currency rates,
6 stock indices and 9 interest rate
markets (see Table 1.) DrKW may
extend the services to cover the credit and commodities markets some-
Figure 1. The PIA fixed income page for weekly Bunds
40
July/August 2005
time next year.
"By looking at price action and
price development we are able to
identify how sentiment is changing"
adds Knudsen. "We discern what current sentiment is from a technical and
statistical analysis of price action
alone. This provides us with 75% of
the information we need. Our
assumption of the impact differing
price changes will have on current
market sentiment makes up the
remaining 25%. This allows us to
publish an implied confidence level
with each recommendation." The
confidence levels range from 65% to
80% and reflect Knudsen's and his
team's assessment of the reliability of
each call. However as Knudsen
stresses, "This service is all about
market timing, we don't pretend to be
fortune tellers. The service is
designed to interpret, track and spot
changes in majority market sentiment
and to assess the likely impact on the
direction and timing of investor
activity".
PIA now has 1300 registered users.
Among them is Nick Gartside, a
European government bond fund
manager with Schroders in London.
"We get an awful lot of technical
research from various sources" says
Gartside, "although with each recommendation that's published, the
degree of bullishness of bearishness
is seldom mentioned. PIA publishes
confidence percentages with its levels
and this helps compliment my own
technical research”.
Using candlesticks
Time periods covered by PIA are
intraday, weekly and quarterly. This
suits the DrKW client base served by
Knudsen's team. Their clients range
from bank proprietary desks and
hedge funds to real money accounts,
corporates and central banks. Each
market and time period features up to
3 candlestick charts containing a minimum of conventional chart analysis
such as the plotting of trendlines and
Software
“PIA PUBLISHES
CONFIDENCE
PERCENTAGES WITH
ITS LEVELS AND THIS
HELPS COMPLIMENT
MY OWN TECHNICAL
RESEARCH.”
NICK GARTSIDE, SCHRODERS
Figure 2. The PIA Cheat Sheet
pattern identification. This allows
price action to be clearly highlighted
without the clutter often associated
with price charts, although non-candlestick aficionados may still yearn
for conventional bar chart analysis.
However, Knudsen is confident that
candlesticks provide the most practical charting method. "With candlesticks we are able to visually differentiate the positive and negative periods
plus highs and lows by using contrasting colours. Clients tell us this makes
our research that much easier to disseminate than our competitors
research" he says. "We cannot achieve
this with bar charts. Furthermore, the
open and closing price of a market
have an important relationship to
market sentiment. They allow us to
identify and track this relationship
more closely than would bar charts."
Figure 1. is a screenshot for a PIA
report on weekly Bunds showing the
trading recommendation (sell below
123.78) with its confidence level
(70%). Figure 2. is a PIA 'Cheat Sheet'
which provides an at-a-glance summary of all markets covered by the
service and its basic recommendations.
Not real-time
PIA is not a real-time service. Each
report is prepared the night before
the market opens and is published
and delivered by email that night.
Although the US FX market is still
open by the time PIA reports have
been written, the trading recommen-
21/05/01 –
12/07/05
Fixed income
FX
Money markets Equities
No. of calls
3742
6042
799
1325
No. correct
2638
4498
573
955
Hit rate
70%
74%
72%
73%
dations are still valid for the next day's
trading. As Knudsen explains, "over
the past 4 years trading in the US and
Asian markets rarely has a detrimental
impact upon the calls we have made
for the following day in Europe. FX
activity is really centred in Europe so
it would take a significant event in the
US or Asia for us to alter our forecasts the following morning.
However when this does occur we reanalyse sentiment and send out an
updated forecast as quickly as possible, usually by 8am London time.
Knudsen points out that PIA has
not been designed as a tip sheet. "I
have never considered this as my role
- that is trading." He concludes "My
goal from the outset was to offer
clients a consistently reliable call on
just two things: the direction and timing of sentiment changes. The results
of our efforts in the 4 years since we
started are a 73% hit rate from a total
of 11,908 calls. With these results we
can make a valuable contribution to a
client's decision of when and what to
trade."
For more information contact:
max.knudsen@drkw.com
Table 2. PIA hit rate
July/August 2005
41
The Cube
The browser-based
Market Data Solution.
Real-time Data, News, Analytics,
charts and the TraderMade
database available globally 24/7,
fully customisable by you.
Contact us at:
tel: +44 (0)20 8313 0992
email: sales@tradermade.com
Book Review
MECHANICAL TRADING SYSTEMS
Mechanical Trading Systems
Pairing trader psychology with
technical analysis
By Richard L Weissman
John Wiley & Sons, Inc.
217 pages, £60.00
ISBN 0-471-65435-3
Mechanical Trading Systems is
available from the Technical
Analyst Bookshop at the reduced
price of £51.00 plus £2.00 P+P. To
order please call 01730 233870
and quote "The Technical Analyst
magazine" or order online at
www.global-investor.com/technicalanalyst. This book is usually
posted within three working days
of your order.
When I picked this book up and leafed through its content it rang all the alarm bells I
use to detect groan-inducing trading books. Each chapter starts with a quote from the
likes of Aesop, Sun Tzu, and those great giants of trading, Shakespeare and Buddha.
There is a summary of technical analysis in chapters 1 to 3 that is essentially a re-hash of
existing knowledge (although I understand the need), words like "utilization" are always
preferred to their simple alternative such as "use", and there is a promise of "pairing trading psychology with technical analysis". It all looked very familiar. So I certainly wasn't
expecting it to be the outstanding book that it is.
Weissman's book does what it says it's going to do in the title. That is, it tells you how
to use mathematically-based technical analysis such as moving averages, RSI and
Bollinger Bands to create a profitable mechanical trading system.
The trading systems are divided into two types - trend following strategies and mean
reversion strategies - across four time scales (long-term, intermediate, swing, intraday).
Helpfully - for users of CQG at least - it also includes simple programming language for
CQG backtesting and optimisation software, although I wouldn't let this put you off if
you use another charting provider.
This division into trend-following and mean-reversion strategies is not simply an organisational device. With each category, he talks about the mentality needed to stick with a
strategy, making reference not just to profit and risk/reward but to all the statistics that
really help understand what it would be like to trade the strategy - number of consecutive losing trades, maximum drawdown amount, time out of the market, percent winners,
average trade duration, etc. This may sound dry, but with Weissman's authoritative style
and understanding of what it means to be a trader, he makes the statistics come alive and
you can see the trader enduring the awful losses one after another. This is no gimmick he really has paired trading psychology with technical analysis and presented it in such a
way that you can properly assess what trading style is best suited to you.
Furthermore, Weismann's clarity makes this one of the most thought provoking and
enjoyable books on trading. Its advice is practical and simple to apply and it will spur you
on to ask deeper questions (regarding trading at least) that will probably lead to experimentation and success.
One question that emerges is: are the strategies outlined in this book just ways of making the trader apply sensible risk management discipline, rather than any great tool to
exploit market anomalies? The winning ratios, particularly for trend-following techniques, are very low and suggest poor forecasting ability, despite their profitability. Many
traders might shrug their shoulders and say that technical analysis was never about making forecasts.
However, other techniques such as chart patterns and Elliott Waves and even trendlines, do not feature in this book, (nor in Lars Kestner's equivalent book "Quantiative
Trading Strategies" reviewed in the last issue). Yet these methods are still mathematical,
albeit they are difficult to model. How much more profitable could strategies be if they
were able to draw on this other non-linear side of technical analysis that arguably offers
more predictive capability?
This book is highly recommended for almost anyone involved in trading, forecasting,
or studying the markets. For the trader in particular, reading this book should help identify and emphasize the psychological traits that will be needed to apply a coherent strategy. It will also provide a framework for assessing strategies, and cfor reating new and
better ones.
July/August 2005
43
Commitments of Traders Report
COMMITMENTS OF TRADERS REPORT
6 July 2004 - 12 July 2005
Futures only (open interest) non-commercial net long positions and spot rates
10-year US Treasury
Source: CBOT
200000
4.80
5-year US Treasury
Source: CBOT
5
300000
Non-commercial net long
Spot
Spot
150000
250000
4.5
200000
4
150000
3.5
100000
3
4.60
100000
4.40
50000
4.20
0
2.5
50000
-50000
4.00
2
0
-100000
3.80
-150000
1.5
-50000
-100000
1
-150000
0.5
3.60
-200000
-250000
06/07/2004
3.40
28/09/2004
21/12/2004
15/03/2005
Dow Jones Industrial Average
07/06/2005
Source: CBOT
20000
11000
-200000
06/07/2004
0
28/09/2004
21/12/2004
15/03/2005
Swiss franc
50000
07/06/2005
Source: CME
1.35
Spot
Spot
40000
1.30
15000
10800
30000
1.25
20000
10000
10600
10000
1.20
5000
10400
0
1.15
-10000
0
10200
-20000
1.10
-30000
-5000
10000
1.05
-40000
-10000
06/07/2004
9800
28/09/2004
21/12/2004
15/03/2005
Pound sterling
07/06/2005
Source: CME
50000
2.00
-50000
06/07/2004
1.00
12/10/2004
04/01/2005
29/03/2005
Yen
21/06/2005
Source: CME
60000
114
Spot
Spot
112
40000
1.95
40000
1.90
20000
110
30000
108
20000
1.85
0
1.80
-20000
1.75
-40000
1.70
-60000
106
10000
104
0
102
-10000
100
-20000
-30000
06/07/2004
44
98
1.65
28/09/2004
21/12/2004
15/03/2005
07/06/2005
-80000
06/07/2004
July/August 2005
96
28/09/2004
21/12/2004
15/03/2005
07/06/2005
Commitments of Traders Report
Euro
Source: CME
70000
1.40
3-month eurodollar
20000
11000
Spot
Spot
60000
Source: CME
1.35
15000
10800
1.30
10000
10600
1.25
5000
10400
1.20
0
10200
1.15
-5000
10000
50000
40000
30000
20000
10000
0
-10000
-20000
-30000
06/07/2004
1.10
28/09/2004
21/12/2004
15/03/2005
Nasdaq
07/06/2005
Source: CME
4.00
600000
-10000
06/07/2004
9800
28/09/2004
21/12/2004
15/03/2005
Nikkei
07/06/2005
Source: CME
12500
10000
Spot
Spot
3.50
400000
8000
12000
3.00
6000
200000
2.50
11500
4000
0
2.00
2000
-200000
11000
1.50
0
-400000
1.00
10500
-600000
-800000
06/07/2004
0.50
0.00
28/09/2004
21/12/2004
15/03/2005
07/06/2005
Gold
Source: CEI
500
160000
Spot
480
140000
-2000
-4000
06/07/2004
10000
28/09/2004
21/12/2004
15/03/2005
US dollar index
07/06/2005
Source: NYCE
25000
118
Spot
20000
116
460
15000
120000
114
440
10000
100000
420
112
5000
400
80000
0
110
380
60000
-5000
360
108
40000
340
106
20000
0
06/07/2004
-10000
28/09/2004
21/12/2004
15/03/2005
320
-15000
300
-20000
06/07/2004
07/06/2005
July/August 2005
104
28/09/2004
21/12/2004
15/03/2005
07/06/2005
45
Long-Term Technicals
LONG-TERM TECHNICALS
Provided by Thomas Anthonj, ABN Amro, Amsterdam
EUR-USD
USD-JPY
Retreating from a projected Fibonacci-target for a major top (1.3650)
the market entered a bigger consolidation pattern that has already
come very close to the decisive support zone between 1.1760 and
1.1590. Above this we expect a minimum rebound up to 1.2930 to
potentially form the right shoulder of a bigger head-and-shoulders
pattern. Only above 1.2930 would we see evidence that the old uptrend has resumed. However, a break below 1.1760/1.1590 would
call for a much bigger setback towards 1.1000/1.0764 and
1.0380/00.
Forming almost a double bottom close to the historic 101.25 bottom
and breaking decisively above trend line resistance two months ago,
the market already indicated a bigger rebound. But whether the upmomentum is strong enough to clear key-resistance at 114.46 is too
early to say. A break above would reverse the earlier 3 year long
bear-trend whereas a failure to do so would most likely lead to
another sell-off targeting 101.25 and the head-and-shoulders target
at 95.75. .
GBP-USD
Gold
Stalling right under the projected target zone for a potential top
(1.9566-88), the market retreated and is currently sitting down on
key-support at 1.7308. This should provide support in should prices
begin to rally. Breaking below would only leave minor support at
1.7025 and 1.6904 (Fib.support/old top) to prevent a much bigger
setback towards 1.5923 and 1.5066 (61.8/76.4 %). So above
1.7308 we still see a good chance for another advance towards
2.0115 and 2.0165 (old top/Fib.projections) which should definitely
cap the market for a while.
The triangle consolidation observed lately is still leaving the backdoor open for a continuation of the up-trend (target 490) what would
be confirmed on a triangle breakout at 441. In case triangle support
at 416 should fail, we'd have to expect a minimum setback towards
379 before the bulls might get another chance.
46
July/August 2005
Long-Term Technicals
Dow Jones
S&P 500
Retreating from a potential target zone for the first bigger bull-cycle
up between 11036 and 11400, the market is still in a vulnerable
stage as the correction of this potential accumulation phase might
have just started. However, as long as the market manages to stay
above trend line support at 10165, the bulls are still in control
shooting for the next two major tops at 11350 and 11750 which
again are big hurdles to overcome. A break and close below trend
line support would be a cause for concern as we might be due for
a 61.8 % or 76.4 % retracement of the whole advance from 7197.
Having said that we'd still have room for a decline to 8644 and
potentially even 8091/62.
The latest break above the March top at 1229 is very positive for the
current chart. But still trading within a Fibonacci-target cluster for a
potential top of the bull-cycle from 768, we are on alert for signs of
weakness such as a trend line break at 1179. Such a break could
signal the end of this accumulation phase. If the last low at 1136 is
also taken out we'd have to expect a minimum setback to 1061 and
potentially even to 954/45 (old top/61.8 %/old low). In order to delay
the permanent risk of a bigger setback, the market would have to
clear strong resistance between 1246 and 1288
(Fib.projection/retracement/trend line).
Nasdaq
Nikkei
The setback experienced from 2192 down to 1890 in the first 4
months of this year is much too small to be classified as a correction based on the whole advance from 1108. The conclusion is
therefore twofold: A) The market is shortly going to break above
the last 2192 top which should trigger further buying, and if the old
2328 top can also be cleared, strong acceleration upwards. B) But
if 2192 can't be cleared we are still running a fairly high risk of
accelerating down with a minimum target at 1778/51.
Having spent more than a year in a sideways consolidation pattern,
the market managed to break and close above long-term trend line
resistance (now support at 11509). This supports the idea that it
wants to break out of an inverted head-and-shoulders reversal pattern that would project a target at 16850 once the neck/trend line is
broken on close at 12307. There under we are still running a risk of
missing the right shoulder of the head-and-shoulders pattern that
would imply another decline to around 9383.
July/August 2005
47
Events
EVENTS 2005
Organiser Date
Event
Venue
August
23
Technical Analysis in the
Commodity & FX Markets
Cape Town, events@technicalanalyst.co.uk
South Africa
+44 (0) 207 833 1441
August
25
Technical Analysis in the Johannesburg, events@technicalanalyst.co.uk
Commodity & FX Markets South Africa
+44 (0) 207 833 1441
September Technical Analysis in the
Dubai,
27
Commodity & FX Markets United Arab
Emirates
48
Contact
events@technicalanalyst.co.uk
+44 (0) 207 833 1441
October
25
Technical Analysis in the
Commodity & FX Markets
London,
UK
events@technicalanalyst.co.uk
+44 (0) 207 833 1441
September
14
Monthly
Meeting
London,
UK
info@sta-uk.org
September
22
Annual
Dinner
London,
UK
info@sta-uk.org
October
12
Monthly
Meeting
London,
UK
info@sta-uk.org
August
4-5
Technical
Analysis
New York,
US
info@ftknowledge.com
July/August 2005

Issue 11 - The Technical Analyst

Transcription

Similar documents

Trading floor

Marketplace - Traders.com

Small Business Development Fund

Read full article - MultiCharts Trading Platform

PPT - Service Leadership Education

0809 MARKETPLACE

0807 MARKETPLACE

Presentation Slides

Brochure - QuoteStorm