Lecture 2: Technical Analysis &

Indicators
Daniel Macauley

Lecture Outline

Technical Analysis Overview

Patterns
Moving Averages
Indicators
Momentum & Reversal Trading Stategies
De Mark

Technical Analysis
Ignores company & fundamental
information
Focus on historical price or volume
information
Industry use? Theory vs Practicality
What does the literature tell us?
Nonlinear dynamic models

Industry Use Brewin Dolphin

Background to TA

Dow in 1890s WSJ

Hamilton 1910s - Dow theory
Gann 1935 Fans predicted 1929
Elliot 1938 waves
Grew in popularity in 1950s
Then efficient markets
chart reading must share a pedestal with alchemy

Bollinger 1980s - bands

However in the 90s was TA was effectively
ignored.

Technical Analysis

Seminal paper in 2000

Lo et Al at MIT
This paper changed the way that TA was viewed.
Used statistical techniques to analyse the subjective shapes
and patterns found in TA.
ie a systematic approach to pattern recognition.
Focused on 10 patterns used in TA
H&S Inv H&S Double Bottoms & tops, etc

Discovered that incremental information can be used to add to

the investment process.
Did not focus on profitability of outcomes

Momentum Strategies
Price momentum Strategies
Buy in the direction of the trend

Price Reversal Strategies

Buy against the direction of the trend

Industry use?
Theory vs practicality

Momentum Trading Strategies

This involves buying stocks that are rising
and selling stocks that are falling
Works best in trending markets
These strategies lose significant capital in
sideways markets.
Major attraction is no fundamental research
required.

Directionless Market

Momentum Strategies II
Profitability of Momentum Strategies
Costs - spreads, dealing, market impact etc
Turnover
Market over reaction
All of the above are issues for M S.

NASDAQ Trendline

Types of Charts for TA

Points - Close price

Bar OHLC most common
Candle Stick
Point & Figure - rarely used
Time-scales - tic to yearly

Types of Charts - Candle stick

Technical Analysis
Support levels - buying in sufficient
quantities sufficient to halt a downtrend for
an appreciable period.
Resistance levels selling in sufficient
quantities sufficient to halt an up-trend for
an appreciable period.

Technical Analysis - resistance

Support Trendline

Power of Support & Resistance

Which are strong and which are weak?

Want to try to define!
Rem: Support is a concentration of demand.
How much demand?
Can be quantified mathematicallyto a
degree.
Time, Width & Volume

Power of Support & Resistance

CISCO

Pivot Points
This is the series of points forming the
trendline.
Must have three for a trend
Four is confirmation - geometric

Up trendline - support

Technical Analysis - Trendlines

Significance
Number of touches
higher is better

Slope
steeper is better ?

Length
longer is better

Trendline reversal

Channels

Channels

Continuation Patterns
Patterns that occur on a trend
Pennants
Flags
Triangles

Gives an indication of the length of the trend.

Used to set price objectives
Easy to spot in hindsight

Continuation Patterns Pennant

Continuation Patterns Trade

Continuation Pattern Flag

CRH Continuation Pattern

Reversal Patterns Head &

Shoulders

Reverse Head & Shoulders

Reversals

Major reversals tend to happen at

round numbers.

TA - does it work?

Research - Behavioural analysis

Self Fulfilling
Depends on the markets
Purity & Liquidity
Used unconsciously by traders
Support & resistance

Fibonacci Sequence

No. Sequence
1, 2, 3, 5, 8, 13, 21, 34..
Each no is the sum of the previous two
Common in nature
For Charting purposes ratios are key

Fibonacci Ratios
Xn / Xn = 1 =100%
Xn / Xn+1 = 1 =61.8%
Xn / Xn+2 = 1 =38.2%
Xn/ Xn+3 = 1 =23.6%
In technical analysis 50% is also significant
but is not a fibonacci ratio.

Fibonacci Ratios

Often incorrectly used

Trend must be correctly identified
Used for Retracements
Entry points in an uptrend
Useful for Stops placement
Good in fast markets

Magnitude of Retracements

Projection is directly correlated to length of

A-B and inversely correlated to B-C

Fibonacci Retracement

Fibonacci Extension

Stops - what are they & why are

they important?
Limits the size of a loss
Can calculate the trades risk and reward
ratio
To achieve a profit or MPO while letting the
trade run.
To stop a substantial hit to the P&L

Stops Placement
Should they be placed at all?
Depends.....
How tight should they be?
Risk reward of staying in the trade as opposed
to losing significant capital

What is the goal?

Problems with Stops

Stop Loss
Never cancel
Never extend
Can be moved,
but only in one direction
concept of a trailing stop loss

Stops
Percentage, Dollar,
arbitrary, driven by MM

Fundamental analysis
Technical analysis
Market consolidation
Gaps
Fibonacci

Volatility
Bollinger Bands, ATR

Trailing

Stops
10% for example, must be asymmetric with
the reward.
Classic beginners error.
Avoid round numbers.
Clustering of markets at key levels, round
numbers.
Double top on the ISEQ almost exactly 10,000

Stops
Fundamental basis
have a view on an outcome and place stop in
relation to this.

Stops at Resistances

Stops & channels

Gaps

Stops & Triangles

Place in the middle

Fibonacci Stop Placement

Volatility Stop Placement

Measure current volatility of instrument
Place stop outside the price range implied
by this.
Number of methods
ATR average true range
Historical Volatility

Trailing Stop loss

No of methods
Lowest low
Highest high
Creates a band
Used by hedge funds

Trailing Stop loss

Orders - Gunning the Market

Taking out all of the stops illegal
developed markets such as US, UK but
hard to identify.
Barclays
Market was weak

Barclays

Sell Disciplines
Stop loss mechanism disciplines portfolio re-balancing
1.The initial stop loss barrier is set at purchase price minus a
risk-adjusted discount of ~15%
2.The stop loss barrier is adjusted daily to the stock price.
3.A flag is triggered once the stock price hits the stop loss
barrier
4.The flag does not lead to automatic sell action but triggers
a team discussion and a fundamental ad hoc analysis with a
respective investment conclusion (hold, increase, reduce,
sell)

Sell Discipline

Indicators
These are calculations based on the price and the
volume of a security.
Two types Bounded & Unbounded

Moving Averages
Relative Strength Index (RSI)
Moving Average Convergence Divergence (MACD)
Bollinger Bands
Stochastics

Indicators - Do they work? Are

they used?

Some of them used extensively

e.g. Moving Averages, RSI
Depends on Types of Markets (Assets)
MACD more in FX
Self Fulfilling as with all TA

Moving Averages
Moving average of the price
Chart of,
(A + B + C)/3 , (B + C + D)/3, (C + D + E)/3,
etc
Above is a three day moving average.

Various periods
Simple, Weighted, Stepped, Exp, etc
Used to identify a trend

Moving Averages

Used to identify a trend

Different periods
Moving average crossovers, two periods
Can be used to create envelopes
Used extensively as easy to comprehend
Citi, MS, BCA, BD etc

Simple Single Moving Average

Moving Averages

Moving Average Crossover

Moving Average Envelope

200 Day Moving Average

RSI

Relative strength Index

Oscillator (ranges between 1 and 100
RSI = 100-1/(1+RS)
where RS = Average of Days up
closes/Average of Days downcloses
Various Periods but standard is 14 days
Very extensively used.

RSI

Over 80 is overbought
Under 20 is oversold
Not hard and fast rule 70:30 sometimes
Using shorter periods makes it more
volatile.
Divergences are the key

RSI

RSI Divergence

MACD
Difference between a fast and slow EPA
This is the MACD line
No.s used are 12 and 26
MACD = EMA 12 EMA 26

Uses a signal line which is the EPA of the MACD

Signal = EMA 9 OF MACD

Based on MA so is a lagging indicator

Histogram to show the crossovers.
HISTOGRAM = MACD - SIGNAL

MACD
Usual signal is,
Signal line crossing over the MACD line
or
MACD line crossing zero
Up being bullish and vice versa

Bollinger Bands
Middle Band
N period moving average (simple / exp)

Upper & Lower Bands

K times N period Std Dev above / below band
Typically n & k are 20 and 2 respectively

Two indicators
%b (from Stochastics)
Tells you where you are in relation to the bands

Bandwidth
How wide are the bands
BW = (upper band lower band) / middle band

Bollinger Bands
Uses, either
Buy when price touches lower band and sell on
reaching the middle.
Or
Buy when price breaks above the upper band.
Also used in option trading sell options when bands are
historically far apart and vice versa
Also used as a confirmation indicator
Studies have shown them to be effective in some
markets after costs

Bollinger Bands example

Momentum Trading
Stock that do well in one period tend to do well
in subsequent period.
Stocks that underperform in one period tend to
underperform in the next period.
So the way to play these trades is in a hedged
portfolio.
Research has shown that hedge portfolios
based on this tend to provide positive returns
on average after transaction costs.

Reversals
Longer term effect than for momentum trades.
Rank returns over a 3 year period and then
typically buy the decile with the lowest
returns and vice versa.
Holding periods for this strategy is of the
order of 3-5 years.
Returns are positive after TCs but not as
good as short term momentum.

Short term and medium term

Momentum based strategy

Long term reversal strategy

Momentum Strategies
J-month / k-month strategy
Winner / Loser portfolio
Modifications,

Skipping a week between j and k

Using quartiles, etc
High idiosyncratic volatility
Volume
Company size

Reversal Strategy
Debondt et al (85) did some of the earliest work
on this.
They looked at 36/36 reversals
Found that the loser portfolio generated most of
the return.
Good basis in psychology people tend to
overreact to unexpected and dramatic news
events.
i.e. Extreme losers become too cheap and
bounce back

Momentum
Backed up by a number of other researchers.
Poterba et al (88) essentially stocks in the short term
experience positive autocorrelation and in the long term
experience negative autocorrelation.
Similar outcome in 17 other markets.
Jegadeesh et al (93) - Significantly better outcome.
Looked at 16 different combos - robust outcome
Eg 6/6 9.3% annualised after 0.5% TCs

Momentum
New study in 01 by same team found that
the pattern continued thru the 90s.
Then Hong et al (00) 6/6 momentum 30th
percentiles,
Create a long short portfolio and screen for,
Market cap momentum highest in midcap
Analyst coverage lower better

Momentum Strategies Improvements

Research focus on improving returns from these
strategies.
Stocks with high Idiosyncratic Volatility IVOL
Arena et al 08 6/6 momentum portfolio
First sort the stocks into IVOL baskets,
Low, medium & high

All returns are statistically significant.

But increases with increasing IVOL

Also checked a number of other factors not

significant

Momentum Strategies Improvements

Vassalou et al 04 document a strong relationship between
price momentum and what they call corporate innovation.
CI defined as the proportion of a firms change in gross
profit margin that is not explained by the change in
growth in labour and the capital it utilises.
Stocks sorted on the basis of CI and employ a 6/6
strategy.
Get a return of 0.76% per month over six months.
Also found that the momentum profits tend to dissipate
over longer term time horizons.

Moving Averages Strategies

Early work buying stocks that are
substantially above their moving averages is
very profitable on a risk adjusted basis.
The use of MAs became common place.
Reilly et al (03) trading rules based on MAs.
Stocks are bought when the short term moving
average is above the long term moving average.
Park (10) shows that the relationship between the
short and long term MAs is important.

Moving Averages Strategies

Parks trading rules

50/200 day MA
Rank stocks in deciles depending on ratios
top decile are deemed winners and bottom losers
6 month hold period
Strategy generates 1.45% per month
Greater than that for pure momentum strategies
Also works if other MAs are used both for short and
long .

Strategies based on 52 Week Highs

& Lows
George et al (04) looked at a ratio of the current
price to its 52 week high.
Sort stocks based on this ratio.
Top thirtieth percentile are the winning portfolio
and the bottom thirtieth percentile are the losing
portfolio.
Create a zero cost portfolio by shorting the
losers and buying the winners
Six month hold strategy.
Generates about 0.5% per month.

So what is causing momentum

Some studies show that this is caused by industry effects.
The momentum effect is stronger when viewed by
industry as opposed to individual stocks.
Moskowitz (99) ranked industry portfolios based on their
6 month returns and then create long short portfolio.
This involved buying the top 3 performing industries and
selling the bottom 3 performing industries.
Also found that industry is more important for the long
portfolio.

Sector Momentum
Similar findings by other researchers
ONeal used fidelity sector funds and found a
similar outcome for a 6/6 momentum strategy.
He found a of return of 8.6% annualised.
Slightly higher for 12 months.
Importantly he factored in trading costs.
Clearly shows that industry momentum is
prevalent over the medium term.

Can momentum be used as a trading

strategy
Costs are an issue.
Recap
Large stocks more liquid, lower price impacts lower spreads

Some of the earlier studies had costs up to 0.5% but,

Deciles contained illiquid stocks in the main and lesmond et
al (04) argue that these are too low.
Estimate costs of 1 to 2% for large caps.
And up to 9% for less liquid stocks.
So some of these strategies could impacted severely by
transaction costs.
This is backed up by other researchers

Can momentum be used as a trading

strategy
Korajczyk et al 04
Look at 11/1/3 and 5/1/6 momentum strategies.
Also four strategies EW, VW, LW and 50:50
VW:LW
Only looked at the long side.
Factoring in realistic costs for the first strategy,
0.45% return pm for the VW
0.6% for the EW

And slightly lower for the second strategy,

Are momentum strategies used by

institutional investors

IIs tend to herd.

They follow positive feedback (momentum) strategies.
They also window dress.
They go for stocks with high liquidity and low TCs.
So yes if not explicitly
Also research shows that a significant portion do but
far from a majority.
Also opposite contrarian strategy employed by a
portion may cancel out some of the purchases.

Are momentum strategies used by

private investors
Tend to sell winners to early
Tend to hold losers
So,
No in the main are contrarian investors

Why are there returns from

momentum and reversal strategies
Compensation for risk? Probably not
Investor behaviour
Newswatchers & momentum Hong et al (99)
Confirmation bias Friesen et al (09)
Changes in the business cycle Chordia et al (02)

International momentum

Not just the US

39 international markets Griffin (03)
17 markets Huang et al (06)
Not so much in Japan
More evidence of information diffusion
internationally

DeMark Analysis

De Mark Analysis

Retracements
Generation of trade ideas
Identifies exhaustion of a trend
Used extensively
TD Setup + TD Sequential

TD Sequential

DeMark for ISEQ

FTSE

TD Set-up
Start of a Buy Set-up is Close today
greater than the close of 4 bars earlier.
Start of Buy set-up..

TD Buy Set-up Continued

Then you require a minimum of 9 consecutive bars close
lower than previous close to complete the set-up.
There can then be tradeable retracement but only short
term.

TD Set-up Continued
TD Set-up trade qualifer is that the low at
bar 8 or 9 needs to be below the low of 6
and 7

TD Sequential Buy Countdown

Need 13 bars in the same direction
Each close must be equal or below the low
of two bars earlier.
Unlike set-up dont need to be consecutive
The first of the 13 can be the same as the 9
bar in the set-up trend
Once you have 13 bars TD Sequential is
completed

TD Sequential Buy Countdown

TD Sequential Buy Countdown

Qualifier
The low of bar 13 must be below or equal to the
bar of close 8
If not get a plus sign on BBerg

TD Sequential - CRH

TD Setup & Sequential example

TD Sequential

TD sequential

Mitchell & Butlers

Ryanair

RIO

FTSE

Smith & Nephew nearly a 13!

But

Analyst is very positive

Whispers of corporate activity
Consensus expectations increasing
Signs of topline growth as economy
improves.
Rerating of the stock possible

De Mark Failures?