Friday, December 24, 2010

Generating Signals from RSI

RSI signals that are not derived from overbought or oversold

Ever wondered how you can generate signals from RSI apart from the usual overbought and oversold levels? It can be done although personally I wouldn't trade on signals that do not include an input from price. Let's face it RSI can be very frustrating when it fails to reach overbought or oversold for long periods of time. Well, let's just consider what RSI is telling us and whether we can use that concept in another way.

In fact RSI is calculated by measuring the sum of the higher closes and also the sum of the lower closes and normalizing the ratio of the result within a band of zero to 100. Clearly, when price rises then the sum of the positive close movement is larger than that of the negative close movement and thus RSI moves higher. The opposite is also obviously true.

Therefore RSI is reacting to sustained directional moves in one direction. What we need try and obtain from that information is when the RSI moves sufficiently in one direction is there risk of follow-through. We actually do this type of thing with price by placing Bollinger Bands around price and looking for breaks of the upper or lower bands. It is possible to do just that with RSI. It will look like this:

The bands are based on a 9 period average with a standard deviation of +/- 1. The idea I am investigating is whether breaks of either band by the RSI constitutes a valid signal. However, without even looking at price I can see that the number of occurrences where the band is broken but does not sustain a movement in that direction is quite frequent.

One of the most important factors, in my opinion, when trading is to match a signal generated by a technical indicator with price. Too often the movement in price that generates a break in an indicator fails to follow through. Because of that I like buying breaks of resistance or selling breaks of support. Therefore I can include both conditions in my trading.

In this way, when RSI breaks above the Bollinger high then I will only trade if price also breaks above resistance. What I tend to look at are swing bars - that is where a peak is formed and has at least 1 or 2 lower highs surrounding it (and vice verse for a trough.) So when RSI breaks above the upper band I will want to buy at the level of the swing high. If RSI breaks below the lower band I will want to sell at the level of the swing low.

The following shows how this may look:

Let's look at what we can generate from these combined signals one by one. I have marked swing highs and lows with a small red horizontal line:

  1. RSI dips below the lower Bollinger Lower Band. However, by the time it has confirmed this (at the close of the bar) price has already dipped below the swing low - it is too late to trade.
  2. RSI breaks above the Bollinger Higher Band. However, price fails to follow through above the swing high. No trade is seen.
  3. RSI breaks above the Bollinger Higher Band and on the next bar price confirms this move by penetrating the swing high. We buy at the swing high level. Price continues higher and then pulls back. In order to provide a trailing stop I have also entered a fairly fast SAR/Parabolic and we square on the 9th bar of the trade at the Parabolic level for a modest profit.
  4. RSI breaks below the Bollinger Lower Band. However, price does not break below the swing low. No trade.
  5. RSI once again breaks below the Bollinger Lower Band. However, price does not break below the swing low. No trade.
  6. RSI breaks above the Bollinger Higher Band and on the next bar price confirms this move by penetrating the swing high. We buy at the swing high level. Price continues to rally and to the right of the chart the position is still open having remains above the Parabolic.

Quite clearly the benefit of combining a signal from an indicator and from price appears to be working well, filtering out the bad indicator signals.

It is quite easy to have Bollinger Bands plotted on the RSI by writing a new indicator in the Chart Studio using CTL. To do this open up your Chart Studio and then in the system modules (click on the tab at the bottom left of the screen) and then scroll down to find “Relative Strength” and double click on it. This will open the coding for the RSI which is shown below. You will need to make some minor changes or just copy and paste the following into a new page:

indicator RSIandBollingerBands;
input price = close, period = 14, SD = 1, BollPeriod = 9 ;
draw line("RSI"), BBhigh("Bollhigh"), BBlow("BollLow");
vars i(number), u(series), d(series), au(series), ad(series), dif(number), f(number),RS(series);
begin
f := front(price);
u[f] := 0;
d[f] := 0;
for i := f + 1 to back(price) do begin
dif := price[i] - price[i - 1];
if dif > 0 then begin
u[i] := dif;
d[i] := 0;
end else begin
u[i] := 0;
d[i] := -dif;
end;
end;
au := mma(u, period);
ad := mma(d, period);
RS := 100 * au / (au + ad);
line := RS ;
BBhigh := SMA(RS,7) + SD * stddev(RS, Bollperiod) ;
BBlow := SMA(RS,7) - SD * stddev(RS, Bollperiod) ;
end.

  • From the menu bar select “Build” and then “Verify Module.”
  • You will be prompted to save the file. Name this “RSI and Bollinger Bands”
  • Against select “Build” and then “Verify Module.”
  • Finally once again select “Build” but this time “Install Module.”

You should find that the indicator appears in the “User Modules” window at the left of the screen. When you go back to your charting area and insert an indicator you will find “RSI and Bollinger Bands” will be available to insert into the chart.

Tuesday, December 14, 2010

Using Fibonacci Levels to Detect Range Bound and Trending Markets

Why we shouldn't buy breakouts

The FX-market oscillates on a regular basis between range bound and trending markets. In range bound market conditions, traders typically adopt a simple buy low, sell high approach, where as trending market climates call for traders to trade with the trend. However detecting whether the market is currently in a range bound or trending environment can be tricky, and costly if applied inaccurately. With that said, the Fibonacci levels can provide a valuable insight to the current market climate and the appropriate trading approach.

The first chart below shows a significant rally to the upside, as the trend reversed direction, the market then passed through all 3-commonly used Fibonacci levels; 38.2%, 50%, & 61.8%. Due to the fact that not one of our Fibonacci levels established our new support, we can extrapolate that a trend is not probable. It is important to keep in mind that trends exist when there is an uneven distribution of buyers and sellers, forcing the market to new high/low prices. However due to the fact that the market fell back below every Fibonacci line, indicates that the buyers were not in fact in control of the marketplace. Finally note how the market then accomplished a 'slightly' new low before reversing once again back to the upside. If we were to sell short the market at its slightly new low price, we would have certainly exited the trade at a loss.

We may now see how the market rallied back to the upside, and found a new (lower) resistance at the 50% Fibonacci level; measured now from our recent high to low prices. Due to the fact the market's progress was then halted at a specific Fibonacci level tells us that at that moment in time, the sellers had in fact gained control of the marketplace as they will not allow the buyers to force us back to recent highs, and an ensuing downtrend is now more probable.

These observations also teach us a valuable lesson against the practice of buying new highs or selling new lows. Conventional wisdom dictates that if the market accomplishes a new high price, the short (stops) will be triggered, in turn will compel the market to even higher highs. However upon careful study of the two aforementioned charts, we learn that if a new trend is intact, the market 'should have' found a new support/resistance level at one of the previous Fibonacci lines. After the first rally to the upside was complete, and as the market retreated back to recent lows, the fact that the market did not find a new (higher) support at one of our Fibonacci levels tells us that the buyers were not in fact, in control, and a relatively equal distribution of power remained between the buying and selling forces which have a greater chance of keeping us in a perpetual range bound market environment. On the other hand, due to the fact that the market found new lower resistance at a Fibonacci level indicates stronger selling pressure, and a greater likelihood of a new trend to the downside will now emerge.

Saturday, December 4, 2010

Enhancing Currency Trading with the Elliott Wave Analysis

One of the most common topics of conversation for traders is Elliott Wave analysis. Ironically, few traders actually apply this method because many are unsure about the intricacies of the Elliott Wave. In truth, correct analysis and counting of the waves can be a daunting task. However, even without the help of electronic wave analysis, traders should be able to enhance their profitability with a disciplined use of Elliott's method.

How It All Started

In the first half of the 20th century, Ralph Nelson Elliott concluded that financial markets have a striking resemblance to a basic harmony found in nature and postulated that price movements in financial markets follow patterns but not necessarily in time or amplitude. Elliott set up the Wave Principle on empirically derived rules for interpreting the price action.

Basics of Wave Analysis

Elliott postulated that financial prices unfold according to a basic rhythm or pattern: five waves in the direction of the trend and three waves counter trend. He named the five-wave upward movement an impulse wave, and the three-wave counter trend- a corrective wave.

Within the five-wave bull move, waves 1, 3 and 5 are called impulse waves. They are subdivided into five waves of smaller scale. The subwaves of impulse sequences are labeled with numbers. Waves 2 and 4 are corrective waves, subdividing into three smaller waves each. The subwaves of corrections are labeled with letters.

Waves of any degree in any series can be subdivided and then subdivided again into waves of a smaller degree, as follows:

IMPULSE CORRECTIVE CYCLE
Waves 1 1 2
First subdivisions 5 3 8
Second subdivisions
Subdivisions 21 13 34
Third subdivisions
Subdivisions 89 55 144

Elliott's Rules of Interpretation

Elliott identified three essential rules of interpretation of his wave principle:

Wave 2 may never retrace more than 100% of wave 1.
Wave 3 is never the shortest; most of the time the longest.
Wave 4 can never enter the price range of wave 1.

Characteristics of the Waves:

  • Wave 1.
    Wave 1 is difficult to identify because it appears to be more of a correction. It is often the shortest of the impulse waves.
  • Wave 2.
    Wave 2 should be easier to identify due to its three-subwave structure. It tends to retrace by about .618% of the first wave.
  • Wave 3.
    Wave 3 is usually the longest and it is never the shortest. It has a dynamic move, and the penetration of the top of Wave 1 attracts more demand in a bull market. Naturally, this makes for a good volume and fundamentals support the move.
  • Wave 4.
    According to the rule of alternation, if wave 2 is complex, then wave 4 tends to have a less complex pattern and vice versa. Of course, Wave 4 can never enter the price range of Wave 1.
  • Wave 5.
    Wave 5 can be dynamic and extended, as Figure 8 shows. By now everyone has figured out the long-term trend and these conditions can create a good overshooting scenario.
  • Wave A.
    Wave A is difficult to catch through all the euphoria. A good hint comes from the break into five subwaves.
  • Wave B.
    Wave B may be of different complexities and lengths since the last bulls are making their final mark in a previously rising market and the bears are testing the waters and starting to go short.
  • Wave C.
    Wave C puts the stamp on the end of the trend. Following a bull market, wave C should fall below the bottom of wave A.

Impulse Waves - Variations

Elliott Wave identifies several extensions, which are not easy to use. In the five-wave sequence, one of the three impulse subwaves tends to generate an extension. These subdivisions are of nearly the same amplitude and duration as the larger-degree waves of the main impulse sequence, giving a total count of nine waves of similar size rather than the normal count of five for the main sequence.

Extensions can be useful guides to the lengths of future waves. Most impulse sequences contain extensions in only one of their three impulsive subwaves. Therefore, if the first and third waves have about the same magnitude, the fifth wave will probably be extended. Extensions may also occur within extensions. Although extended fifth waves are not uncommon, extensions of extensions occur most often within third waves.

FAILURES OR TRUNCATED FIFTHS

Elliott called failure any impulse pattern in which the extreme of the fifth wave fails to exceed the extreme of the third wave.

Corrective Waves

Market swings tend to move easier with the trend of a larger degree than against it. Therefore, corrective waves can be highly complex, choppy and often difficult to interpret before completion.

The most important characteristic about corrections is that they never consist of five subwaves. Only impulse waves consist of five subwaves.

Elliott identified the following four corrective patterns:
1. Zigzags (5-3-5)
2. Flats (3-3-5)
3. Triangles (3-3-3-3-3)
4. Combined structures

Zigzags are simple three-wave patterns, subdivided into 5-3-5 structures, in which the extreme of wave B remains a significant distance from the beginning of wave A. Occasionally, a double zigzag formation may occur.

FLATS - Flat corrections have a 3-3-5 structures. The original movement of wave A lacks the momentum to develop into a full five waves, as in a zigzag. Wave B also lacks countertrend pressure and often ends at or beyond the start of wave A. Wave C usually finishes near the extreme of wave A, rather than significantly beyond it, as in a zigzag. Elliott identified four types of flats: regular, expanded, irregular and running.

TRIANGLES - Triangles tend to occur just before the final rally in the direction of the trend. They tend to be extended because of low volume and low volatility during a consolidation period. Triangles consist of five waves, labeled A-B-C-D-E, subdivided into three waves each. The four types of triangles are symmetrical, ascending, descending and expanding. After completion of a triangle, the final impulse wave of the larger trend is usually swift and has a price objective equal to the base of the triangle. This movement is called a thrust.

COMBINED STRUCTURES - In Elliott Wave analysis, zigzags and flats are often called threes. These combined structures consist of two or more threes separated by smaller three-wave movements, labeled X waves. For example, a double three may consist of a flat, a smaller zigzag-forming wave X and a second flat, or it might contain a zigzag , a smaller flat in wave X and a second zigzag. The combined structures are generally sideways formations reflecting market hesitation.

Conclusion

A disciplined analysis and a trained eye should help you get a good grip on Elliott's method. While it can be complex and time consuming, it is also an excellent approach to true forecasting.

Enhancing Currency Trading with the Elliott Wave Analysis

One of the most common topics of conversation for traders is Elliott Wave analysis. Ironically, few traders actually apply this method because many are unsure about the intricacies of the Elliott Wave. In truth, correct analysis and counting of the waves can be a daunting task. However, even without the help of electronic wave analysis, traders should be able to enhance their profitability with a disciplined use of Elliott's method.

How It All Started

In the first half of the 20th century, Ralph Nelson Elliott concluded that financial markets have a striking resemblance to a basic harmony found in nature and postulated that price movements in financial markets follow patterns but not necessarily in time or amplitude. Elliott set up the Wave Principle on empirically derived rules for interpreting the price action.

Basics of Wave Analysis

Elliott postulated that financial prices unfold according to a basic rhythm or pattern: five waves in the direction of the trend and three waves counter trend. He named the five-wave upward movement an impulse wave, and the three-wave counter trend- a corrective wave.

Within the five-wave bull move, waves 1, 3 and 5 are called impulse waves. They are subdivided into five waves of smaller scale. The subwaves of impulse sequences are labeled with numbers. Waves 2 and 4 are corrective waves, subdividing into three smaller waves each. The subwaves of corrections are labeled with letters.

Waves of any degree in any series can be subdivided and then subdivided again into waves of a smaller degree, as follows:

IMPULSE CORRECTIVE CYCLE
Waves 1 1 2
First subdivisions 5 3 8
Second subdivisions
Subdivisions 21 13 34
Third subdivisions
Subdivisions 89 55 144

Elliott's Rules of Interpretation

Elliott identified three essential rules of interpretation of his wave principle:

Wave 2 may never retrace more than 100% of wave 1.
Wave 3 is never the shortest; most of the time the longest.
Wave 4 can never enter the price range of wave 1.

Characteristics of the Waves:

  • Wave 1.
    Wave 1 is difficult to identify because it appears to be more of a correction. It is often the shortest of the impulse waves.
  • Wave 2.
    Wave 2 should be easier to identify due to its three-subwave structure. It tends to retrace by about .618% of the first wave.
  • Wave 3.
    Wave 3 is usually the longest and it is never the shortest. It has a dynamic move, and the penetration of the top of Wave 1 attracts more demand in a bull market. Naturally, this makes for a good volume and fundamentals support the move.
  • Wave 4.
    According to the rule of alternation, if wave 2 is complex, then wave 4 tends to have a less complex pattern and vice versa. Of course, Wave 4 can never enter the price range of Wave 1.
  • Wave 5.
    Wave 5 can be dynamic and extended, as Figure 8 shows. By now everyone has figured out the long-term trend and these conditions can create a good overshooting scenario.
  • Wave A.
    Wave A is difficult to catch through all the euphoria. A good hint comes from the break into five subwaves.
  • Wave B.
    Wave B may be of different complexities and lengths since the last bulls are making their final mark in a previously rising market and the bears are testing the waters and starting to go short.
  • Wave C.
    Wave C puts the stamp on the end of the trend. Following a bull market, wave C should fall below the bottom of wave A.

Impulse Waves - Variations

Elliott Wave identifies several extensions, which are not easy to use. In the five-wave sequence, one of the three impulse subwaves tends to generate an extension. These subdivisions are of nearly the same amplitude and duration as the larger-degree waves of the main impulse sequence, giving a total count of nine waves of similar size rather than the normal count of five for the main sequence.

Extensions can be useful guides to the lengths of future waves. Most impulse sequences contain extensions in only one of their three impulsive subwaves. Therefore, if the first and third waves have about the same magnitude, the fifth wave will probably be extended. Extensions may also occur within extensions. Although extended fifth waves are not uncommon, extensions of extensions occur most often within third waves.

FAILURES OR TRUNCATED FIFTHS

Elliott called failure any impulse pattern in which the extreme of the fifth wave fails to exceed the extreme of the third wave.

Corrective Waves

Market swings tend to move easier with the trend of a larger degree than against it. Therefore, corrective waves can be highly complex, choppy and often difficult to interpret before completion.

The most important characteristic about corrections is that they never consist of five subwaves. Only impulse waves consist of five subwaves.

Elliott identified the following four corrective patterns:
1. Zigzags (5-3-5)
2. Flats (3-3-5)
3. Triangles (3-3-3-3-3)
4. Combined structures

Zigzags are simple three-wave patterns, subdivided into 5-3-5 structures, in which the extreme of wave B remains a significant distance from the beginning of wave A. Occasionally, a double zigzag formation may occur.

FLATS - Flat corrections have a 3-3-5 structures. The original movement of wave A lacks the momentum to develop into a full five waves, as in a zigzag. Wave B also lacks countertrend pressure and often ends at or beyond the start of wave A. Wave C usually finishes near the extreme of wave A, rather than significantly beyond it, as in a zigzag. Elliott identified four types of flats: regular, expanded, irregular and running.

TRIANGLES - Triangles tend to occur just before the final rally in the direction of the trend. They tend to be extended because of low volume and low volatility during a consolidation period. Triangles consist of five waves, labeled A-B-C-D-E, subdivided into three waves each. The four types of triangles are symmetrical, ascending, descending and expanding. After completion of a triangle, the final impulse wave of the larger trend is usually swift and has a price objective equal to the base of the triangle. This movement is called a thrust.

COMBINED STRUCTURES - In Elliott Wave analysis, zigzags and flats are often called threes. These combined structures consist of two or more threes separated by smaller three-wave movements, labeled X waves. For example, a double three may consist of a flat, a smaller zigzag-forming wave X and a second flat, or it might contain a zigzag , a smaller flat in wave X and a second zigzag. The combined structures are generally sideways formations reflecting market hesitation.

Conclusion

A disciplined analysis and a trained eye should help you get a good grip on Elliott's method. While it can be complex and time consuming, it is also an excellent approach to true forecasting.