The Blending of Technical Analysis & the Top-Down Approach
– To find profitable trading strategies not only in the medium-term but also the short term;
– To find the new trend or to follow the prevailing trends within an industry;
– Furthermore to find a certain single stock within such industry.
We take a top-down approach using technical analysis, in order to achieve our objectives. There are three major elements:
1. Trend of an industry/a single stock: that’s where the technical analysis is performing the best;
2. Strength of the trend of an industry: we use the relative value between the industry and the benchmark of full markets, in order to find a sector which is stronger or weaker than the benchmark. E.g. The DJ STOXX 600 Basic Resource vs. the STOXX 600 index.
3. Strength of the trend of a single company: we use the relative value between a stock and the benchmark of the sector to which the company belongs, e.g. BHP Billiton (BLT LN) vs. the DJ STOXX 600 Basic Resource.
In Top-Down approach:
2. Trend of the sectors (channel, moving average etc.)
3. Relative value markets-sector;
4. Relative value sector-stock.
5. Inter-markets analysis (EUR/USD, T-note, Commodities, especially Crude oil, Gold).
2. Trend of the stock on a daily basis;
3. Oversold/overbought (weekly and daily RSI);
4. Technical patterns (head & shoulders top/bottom, M-top, W-bottom);
5. Technical indicators (Volumes, Stochastic, RSI, Bollinger bands)
1. (Simple) Moving averages
3. Bollinger bands
4. Head & shoulders top/bottom
8. Chart patterns: range, triangle, flag, wedge
1. (Simple) Moving averages (MA)
(Simple) Moving average is the simplest technical indicator, used for determining the trend of the prices of a security. MA is the mean of the previous N consecutive closing prices, normally described as MAt. On a daily basis, we call it the t-D MA; on a weekly basis, t-Week MA.
MAt = [Xt + Xt-1 + … + Xt-N+1] / N
You can build your favourite trading system with the MAs:
For ST trading, common practice utilizes the 5D, 10D, or 5-week MA.
For MT trading, common practice utilizes the 20D, 30D, 50D, and 60D, 10-week or 20-week MA.
For LT trading, common practice utilizes the 100D, 200D, and 30-week, 50-week or 100-week MA.
The short-term moving average, for example, the 10D MA is more dynamic and less smooth than the mid-term moving average, for example, the 50D MA.
How to use:
a. Trading strategies based on a single MA.
Buy above the MA, and (short) sell below this MA.
Please see the daily chart of DJ EUROSTOXX 50 index (below). If we trade using MA10, the signal is given more quickly than using the MA60; but the using the MA10 gives more false signals than using the MA60.
b. Using the crossover between two or more moving averages as a filter, in order to find signal with high-conviction.
Golden cross: Buy signal. A crossover involving a security’s short-term moving average (such as 10-day moving average) breaking above its mid/long-term moving average (such as 30-day moving average).
Deaths cross: Sell signal. A crossover resulting from a security’s short-term moving average breaking below its short-term moving average.
In the FTSE 100 example above:
Golden cross (buy signals) were given on Apr. 07, 2008 and on Aug. 08, 2008;
Death cross (sell signals) on Mar. 10 and Jun 02, 2008.
Channel is a kind of popular technical pattern, used in determining the trend of a security’s prices.
A channel is formed of 2 parallel lines: one is the basic trend line, another is the channel line.
Ascending channel is formed by a basic uptrend line and an upper channel line.
Descending channel is formed by a basic downtrend line and a lower channel line.
How to use:
a. Normally, we buy when the prices are approaching/reaching the basic trend line, and sell when the prices are approaching/reaching the channel line.
b. We will trade oppositely against the channel’s trend: buy (sell) if the prices have broken above the downtrend line (below the uptrend line), expecting the former channel is not more valid.
See the buy/sell signals in the example with the weekly FTSE100 chart below.
3. Bollinger Bands
Bollinger bands consist of:
– 2 trading bands (an upper and a lower) around a moving average.
– A middle band being an N-period simple moving average
– An upper band at M times an N-period standard deviation above the middle band
– A lower band at M times an N-period standard deviation below the middle band
Typical values for M and N are 2 and 20, respectively.
How to use:
In bullish markets: the prices fluctuate between the upper band and the 20D MA.
a. For ST traders: the downside breakout of the 20D MA will generate a sell signal, with target at the lower band;
b. For MT trader: accumulate the long position between the lower band and the 20D MA, expecting a recovery towards the upper band.
In bearish markets: the prices fluctuate between the lower band and the 20D MA.
c. For ST traders: the upside breakout of the 20D MA will generate a buy signal, with target at the upper band;
d. For MT trader: accumulate the short position between the upper band and the 20D MA, expecting a drop towards the lower band.
In neutral markets: the prices fluctuate frequently between the lower band and the upper band.
e. Traders can buy between the lower and the 20D MA, but sell between the higher and the 20D MA.
f. The upside breakout of the upper band will generate a buy signal;
g. The downside breakout of the lower band will generate a sell signal.
4. Head & shoulders top/bottom
The head and shoulders is one of the most efficient trend-reversal patterns.
The head and shoulders top (bottom) has 4 elements:
– One left shoulder, which is a relative high (low);
– One right shoulder, which is almost the at the same height than the left shoulder;
– One head, which is the highest peak (lowest trough) level between the left and the right shoulder;
– One neckline, which serves as a short-term support (resistance) line, and will be broken shortly.
Please note that a head and shoulders top (bottom) must be validated by a downside breaking (an upside breaking) of the neckline.
How to use:
The minimum target of a head and shoulders top = the height of the breaking point of the neckline – the distance between the head and the neckline.
H ≥ h
H ≥ h
If we consider the neckline is almost horizontal, we can simplify the formula into
For example, let’s take a look at the DJ STOXX 600 index, which had formed a head and shoulders top during all the year 2007. In the beginning of 2008, the downside breakout of the neckline validated the top pattern, has triggered a down move towards its theoretical target at 296.85 (= 348.84 x 2- 400.89), then 276.98 (1.382x projection) and 264.71 (1.618x projection).
M-tops and W-bottoms are also called double-tops/bottoms.
The double top (bottom) has 3 elements:
– Two peaks, which are almost the at the same level;
– One middle peak (trough) point, which stands normally on the previous uptrend/downtrend line;
– One breakout point, which is at the same level than the middle peak (trough) point.
Please note that a double top (bottom) must be validated by a downside breaking (an upside breaking) of the neckline; and that many times, the 2nd top (bottom) is slightly higher/lower than the 1st top (bottom) – we call this occurrence a false breakout.
For example, TESORO CORP. (TSO US)
The US refiner posted a false upside breakout in Oct. 2007, reinforced by the bearish RSI divergence. In Jan. 2008, a downside breakout of the middle trough point at USD42.67 (in Aug. 2007) has enabled the M-top and trigged a down leg towards 20.58.
Sometimes, the double top/bottom is transformed into triple-top/bottom, of which only the breakout timing is delayed, but the trading rules and the target measuring are basically same.
V-bottom (top) is the strongest reversal pattern.
This reversal pattern appears normally in an extreme case, when the markets are seriously overbought (oversold).
See below the chart below of the NASDAQ 100 index in during the 2000’s internet bubble as well as the airline companies that bounced upon the drop in the oil prices.
V-top: The NASDAQ 100 index in 2000’s internet bubble: the index lost 35% in 3 weeks.
V-bottom: UAL CORPORATION (UAUA US): after the shares lost 93% (oversold) from Feb. to July 2008, it took only 1 month that the airline company jumped sharply from its July 15 low at 2.8 to 15.84 high on Aug. 19.
Gaps are the price areas between 2 consecutive sessions where no trading has taken in place.
4 types of gaps:
i. Breakaway gap: usually appears during the completion of an important (reversal) pattern. A breakaway gap can remain open for days, weeks, or even years.
ii. Runaway or measuring gap: usually appears after the breakaway gap, or after a short-term or mid-term trend which has already been underway. We can consider it to be in the middle of the underway trend. A runaway gap is often not filled until the underway trend is over (ideally confirmed by the filling of the exhaustion gap).
iii. Exhaustion gap: usually appears after the breakaway gap and the runaway gap, and occurs by the end of the underway trend. An exhaustion gap is always filled very quickly.
iv. Common gap: appears often when the market prices have no a significant trend or are ranging. It can be filled very quickly.
Upward gap example: EADS during Mar.-Dec. 2003 (see below)
Breakaway gap on May 28, 2003
Runaway gap on July 28, 2003, with target at EUR18.61 (=12.47 x 2 – 6.33)
Exhaustion gap on Dec. 01, 2003 at EUR18.2 (almost equals to the theoretical EUR18.61 target)
The filling of the exhaustion gap triggered a 4-month consolidation between EUR20.2 and EUR16.37.
There was an upward gap on Feb. 27, 2004, which is a common gap. It was filled very quickly in the next session. If a trader initiated a short selling position above the gap at EUR18.23 (Mar. 01 2004 closing price) before closing this position at EUR16.82 (Mar. 16, 2004 closing price); he could’ve generated a gain of 7.7% in 11 trading days.
Upward gap example: VOLKSWAGEN (VOW GY)
Downward gap example: PAGESJAUNES (PAJ FP)
Breakaway gap on May 04, 2007
Runaway gap on May 09, 2008, with target at EUR8.86 (=13.05 x 2 – 17.24)
Exhaustion gap on July 15, 2008 at EUR8.2; which was filled on July 17, 2008.
The filling of the exhaustion gap triggered a 27% rise to EUR10.42.
8. Continuation Patterns: rectangles, triangles, flags, wedges
We can class the technical patterns into 2 groups:
a. Reversal patterns which indicate the trend change
b. Continuation patterns, which simply means a pause in the prevailing trend.
How to use:
We use the reversal patterns in anticipation of a new trend and use the continuation pattern in following the trend.
The major continuation patterns are rectangle (range), triangle, flag and wedge. In a bullish trend, we call the continuation pattern a bullish pattern (i.e. Bullish Flag); in a bearish trend, we call the continuation pattern a bearish pattern (Bearish Flag).
However its shorter-term trend is ascending or descending. E.g. a triangle in a bullish trend is called a bullish triangle.
All continuation rectangle (range), triangle, flag and wedge must have 3 elements:
1. a prevailing trend
2. two limits: one upper and one lower
3. a breakout point: a continuation pattern must be completed by a breaking out on the same side of the trend. An upside breakout point for the bullish pattern, and a downside breakout point for the bearish pattern.
Calculation for Measuring:
The width of the continuation pattern is W = upper limit – lower limit
The height of the preceding trend is H = upper limit – beginning of the bullish trend
= beginning of the bearish trend – lower limit
Target 1 = height (breakout point) +/- W …….. + for bullish trend, – for bearish trend
Target 2 = height (breakout point) +/- H ……..+ for bullish trend, – for bearish trend
Please note that a continuation pattern can be unexpectedly invalidated. The invalidation of the pattern signifies that the prevailing trend might be over, or the continuation pattern will be transformed into a more complex consolidation pattern. Let’s see the example with Gold.
The gold had been forming a bullish triangle during May 2006-Aug. 2007, the upside breakout of the triangle triggered an up move towards USD1000.
W = 730.75 -542.4 = 188.35
H = 730.75 – 417.65 = 313.1
Height of breakout point was at 687.35.
Target 1 = 188.35 + 687.35 = 875.7 (reached)
Target 2 = 687.35 + 313.1 = 1000.45 (reached)
After the target 2 at 1000.45 has been reached, another “bullish rectangle” was shaping between 1000.45 and 845.4. Unfortunately this “continuation” pattern has not worked, invalidated by a downside breakout of the lower limit at 845.4. Therefore, we should close the long position below 845.4 and initiate a short selling position immediately.
The trading activity is composed of prices and volume. Therefore we can not ignore the trading volumes as a factor while we make a direct research on markets move.
Ideally, during a bullish (principal) trend, a significant increase in volume while the prices rise; and a decrease in volumes while the prices correct against the principal trend, on a shorter-term basis, can reinforce our bullish view.
In a bearish trend, an increase in the trading volume is not necessary for confirmation of our bearish view; but it is better to see the volumes increasing with the principal bearish trend.
At the trend-changing point/area, especially for an upside breakout of the bearish trend (E.g. the upside breakout of a descending channel, upward breakaway gap, V-bottom etc.), normally we need higher volumes to confirm the changing trend.
Please see the example with the DJ STOXX 600 index: Before July 2003, the markets had been bearish for a long time; a bullish trend (principal trend) was in place during July 2003 and May 2007; the former bullish trend has been invalidated in Dec. 2007, by a downside breakout of the uptrend line and of the neckline of a head and shoulders top.
In July 2003, while the index broke above the 60-week MA (the trend was changing), the volumes increased (see green upward arrow);
During Mar. and Aug. 2004, there was a slight consolidation against the principal bullish trend. The view was confirmed by the declining volume (see red downward arrow).
After then, during a long time, the volume was almost flattening, while the prices were increasing.
In Dec. 2007 and Jan. 2008, the former bullish trend was invalidated by a downside breakout of the uptrend line and of the neckline, reinforced by an increase in volume (see red upward arrow); But before this breakout, the volume had not increased significantly.
During Mar. and May 2008, the decreasing volume can’t help the rebound (correction against the principal trend) go further (see green downward arrow).
During May and July 2008, the prices accelerated on the downside, helped by the increasing volume (see red upward arrow).
Now, the decreasing volume just shows investor’s caution …
10. Oscillator Indicators: RSI and MACD
There are two major types of technical indicators: leading and lagging.
The lagging indicators, such as Moving Averages, Bollinger Bands, which follow the market prices move, are commonly referred as trend-following indicators or systems.
While the markets lose momentum, for example in a trading range, the lagging indicators provide a trade signal quite late, after the change of trend has been confirmed. The oscillators are invented in order to find better solutions in that case.
The Relative Strength Index (RSI) is one of the most popular oscillators.
RSI = 100 – 100/ (1+RS)
• RS = Average of N days up closes / Average of N days down closes
How to use:
1. The RSI oscillates between 0 and 100. The area above 70 is considered an overbought area, and the area below 30 is an oversold area;
2. The most used N is 14 or RSI (14).
3. The fist time when the RSI reaches 30 or 70, can be considered an oversold or overbought warning. However, it does not stress the necessity of buying or (short) selling the security;
4. The RSI can stay in the overbought or oversold area for a long time, while the security is accelerating on the upside or downside;
In order to generate optimal signals, we use bullish or bearish divergences:
5. Bullish divergence: in a descending trend, when the underlying sees a new low and its RSI does not hit a new low; we can say the underlying is oversold.
6. Bearish divergence: in an ascending trend, when the underlying sees a new high and its RSI does not hit a new high; we can say the underlying is overbought.
Why did the Basic Resource sector fall 30% during May-Aug. 2008?
Because it was seriously overbought, confirmed by a bearish divergence for 2 times on the14-week RSI.
Why did the FOOD & BEVERAGE sector rebound sharply (by 16%) during July-Aug. 2008?
Because it was oversold, confirmed by a bullish divergence on the 14-day RSI.
The Moving Average Convergence/Divergence (MACD) indicator combines some characteristics of oscillators and lagging indicators.
MACD is composed of 2 lines: One is the MACD, “faster line”, showing the difference between a fast and slow exponential moving average (EMA) of closing prices, usually the last 12 and 26 units.
The other is the Signal, “slower line”, which in effect smoothes the “MACD” with a further EMA, usually the last 9 units.
How to use:
1. The golden cross (MACD > Signal) between the MACD and the Signal generates a buy signal;
2. The death cross (MACD Cautiously Bearish; pending clearer signals
Entry 1: On Jan 08, the neckline of the H&S top and the 100-week MA were broken out, the bearish view is confirmed.
On Jan 21, the 14-week RSI (blue horizontal line on RSI) joined the low level on Mar. 03, 2003. This is followed by a ST range (sideways trading action) within a MT bearish channel. Within this ST range, a ST rebound occurred, but was capped by resistance (30-week MA/upper boundary of the declining channel in May 2008)
ST range between 332 & 296
Entry 2: Short selling opportunity, suitable in both ST and MT.
Entry 3: The downside breakout of the 200-week MA triggered acceleration.
Entry 4: The downside breakout of the lower limit of the range triggered acceleration.
In July 08, the 14-week RSI was posting a bullish divergence the markets were oversold cautiously bearish.
STOXX 600 Index (Short-term view) – (See Chart below)
In daily (Short-term view): a range had been shaping between Aug. 2007 and Dec. 2007.
=> Neutral/Bearish (in line with cautiously bearish view in the MT)
Entry 1 On Jan 03, the 10D, 30D & 60D MA maintain negative bias (1st clear bearish signal).
Entry 2 (= MT entry 1) HIGH CONVICTION
On Jan 07, the neckline of the range was broken, the bearish view is confirmed.
A range between 331 and 296 was shaping between Jan. and May 2008. In May, the prices were capped by the upper limit of the range, and of the upper boundary of the declining channel.
Entry 3 (= MT entry 2) HIGH CONVICTION
On June 06, the downside breakout of the 60D MA reinforced our bearish view.
Entry 4, suitable in both ST and MT HIGH CONVICTION
On June 19, the 30D MA passed below the 60D MA
Entry 5 Early July, the 14-day RSI was seriously oversold.
=> Cautiously bearish, prepared to take profit
On July 17, the sharp rise with bullish gap confirmed our cautious view for remaining short.
=> We closed our short sell position in the ST, to play ST rebound
Single Stock Recommendations
As we mentioned, early June, while our trading system generated a multiple high-conviction ideas on both the ST and the MT basis, we were interested in the Basic Resource sector, which had been overbought in weekly.
In early June, the price of Crude Oil was increasing (while it had been being oversold), both the Basic Resource and Oil & Gas sectors couldn’t post new highs.
The bearish divergence between the commodities-related sectors and Crude oil called for caution.
On July 5, the STOXX 600 Basic Resource (chart above) broke below the neckline of a ST Head & Shoulders top, enabling us to initiate our 1st short selling position on Dana Petroleum (DNX LN) at 1789. The trade was stopped out unfortunately at 1897 (due to a tighten stop loss), but the share lost finally 33% at 1202 from our initial recommendation.
On June 12, we short sold Royal Dutch Shell Plc (RDSA NA) at 26.3, closed at 24.98 on June 23 with a gain of 5%.
On June 12, we also short sold ArcelorMittal SA (MTP FP) at 62.95, closed at 56.19 on Jun 11 with a gain of 10.7%.
On June 17, we short sold Compagnie Generale De Geophysique Veritas (GA FP) at 30.57, closed at 25.35 on Jul 03, with a gain of 17%.
On June 25, we short sold ThyssenKrupp AG (TKA GY) at 39.96, closed on Jul 02, with a gain of 10%.
On June 25, Lonmin PLC (LMI LN) at 3032, closed at 2762 on Jul 11, with a gain of 8%.
On June 25, STOXX 600 Basic Resources (SXPP) at 735, closed on Jul 25 at 608.12, with a gain of 17%.
On June 30, SALZGITTER (SZG GY) at 116.28, closed at 97.68 on Jul 03, with a gain of 16%.
On Jul 09, Vallourec SA at 196.25, closed at 108.7 on Jul 24, with a gain of 8%.