back start next
[start]           [ 11 ]                                         
oversold when the Percent R level touches the 0 percent line. As soon as any subsequent bar registers a reading greater than 0 the buy signal is immediately issued. Sell signals are created when the prices first generate a 100 percent reading and then give a value of less than 100 percent on any following bar.
The second method by which trading orders are created by Percent R is quite similar to the method used by the Relative Strength Indicator. Buy signals are created when the Percent R plot crosses above the oversold threshold, while sell signals are similarly given when the indicator plot passes through the overbought line on the way down from the higher zones on the chart. This is the more commonly used interpretation and is the scenario on which the trading system we will use to demonstrate the activity of the Percent R formula is based.
Figure 5.14 details specific buy and sell signals as generated by the Percent R oscillator. Paralleling the activity of the RSI system, the signals are generated as the Percent R plot crosses below the over bought line for a sell and above the oversold line for a buy.
Figure 5.15 shows this trading system on the same October 18, 2000, chart. This graph covers trading for the entire day. The gray box marks the portion of the chart detailed in Figure 5.14.
| || |
cMVTUST-2mm -i PdR
| || || |
ruiT 4 i:
..............i .V *.*"
| || |
| || |
| || |
| || |
| || |
| || || |
XJ x f
| || |
| || |
| || || |
| || |
: \Buy Crossover
| || |
| || |
9:52 10:12 10:32
Figure 5.14 Buy and sell signals from Percent R areissued as the indicator plot crosses below the oversold area or above the overbought area. Arrows on the actual price bars define precise entry points.
Figure 5.15 Percent R demonstratesits effectiveness during a sideways day. A length settiag of20 was used to generatethese signals on a two-minute chart. Chart created with TradeStation® 20001 by Omega Research, Inc.
Although there are a few more signals generated by the Percent R formula, you will again notice that the signals generated by yet a third oscillator-type indicator are able to place buy and sell signals in a position from which a profit could have been taken from a significant number of entries. The system has given these trades from a length input of 20 bars. Increasing this value to 30 would have the effect of decreasing the sensitivity of the system, thereby reducing the number of trades and possibly improving trade placement.
Figure 5.16 reflects the changes in the trading pattern as the system input is changed from 20 to 30. Corresponding changes are also made on the accompanying indicator so that a valid comparison may be made.
Notice that there are indeed fewer trades placed by these parameters, in some cases resulting in more precise trade placement. Unfortunately, this setting has resulted in one of the better sell signals at approximately 11:00 a.m. not being generated.
Again, for purposes of comparing data on the same chart, the chart from October 12, 2000, is reproduced as Figure 5.17, this time with the same Percent R system as used on the last two charts returning to the original length setting of 20.
Again note the relatively satisfactory performance of an oscillator-
<:hnulrifv Iru; LASl-2
F!?...?0?" =9«3<! -20.687 -,7.4«C, 0=1,5750 HM19L760 lt.563 V=,e65,00 Pera,ntR<30eo 20)
: pciR - "- .........!...........;.....----------------
....... ....... % ........i....... . , .
,!.........fı A Ai......:............... i =...R5
-100.500 99.500 -98.500 97.500 S6.500 95.500
Figure 5.16 Changing the length inputforthe Percent R system from 20 to 30 significantly reduces the number of trades generated while also increasing accuracy of trade placement. Unfortunately, one of the better signals is lost with this setting.
Chart created with TradeStation® 2000i by Omega Research, Inc.
Figure 5.17 Note the signals generated by the Percent R system with a length setting of20 during a sideways-to-lower day. Chart created with TradeStation® 2000i by Omega Research, Inc.
type system on the chart after it settles into a sideways mode beginning around 9:15 a.m. When we use this chart again later in the book to demonstrate the activity of a further refined system, keep in mind that this day will be classified as a downward-trending day in the market. This definition will come from the Directional Day Filter and will specify that only trades from the short side should be taken. The buy signals from the system will be ignored on these days.
Figure 5.18 demonstrates the performance of the same system when it is applied to a trending day.
As with the two previous formulas, Percent R consistently identifies oversold areas in an uptrending market. Consistent with our system rules explained earlier, this overbought designation leads to the entry into a short trade. Since this trade is against the major trend of the day, most of the short positions shown would have resulted in losses. In his book, Larry Williams is adamant about using his indicator only when you have established the major trend of the time frame in which you are trading. This chart and the one that follows display ample evidence to further substantiate Mr. Williams insistence on trading with the major trend. Note that the long positions established during the noon hour, this time in the direction of the major trend, would have in all probability become winning trades,
cfSnCSraMffKHsl 19.600 1=118.626 V=24300 Percsrrt R(20.80.20}
.,:- R » .ll
............ ............ -i...................i........pifHr -\ "4"
Pd3BP*V llAu.-.X : P12R «"
.P-tf.R....> ...... ..........;--......................... ............................................
J* yf Percent R Uptrending Day
-116.0 DO -112.0 DO 106.000 1C4.00C 1CO.00O
Percent R(20.20.80,black.blacKl «1.79 20.00 80.00 ; .................... ..... ............
| || |
60 00 40.00 -20.00
W \ < /
\I\aI- : W
S-32 l 0.02 10:32 11.02 "11:32 12:02 12:32 1:02 1:32 2 02 2:32
Figure 5.18 Note the effectiveness of Percent Rin identifying the proper
entry point in the direction of the major trend.
Chart created with TradeStation® 2000iby Omega Research, Inc.
Figure 5.19 In Chapter 10 you will learn to identify such days as these as downtrending days early in the tradiag session. Cnderstandiag the major trend will allow you to take only the sell signals from Percent R on a downtrending day and ignore the unprofitable buy signals. Chart created with TradeStation® 2000i by Omega Research, Inc.
Figure 5.19 again uses the data from October 6, 2000, for CMVT. Here we demonstrate the activity of the identical Percent R system on a moderately downtrending day. Once again we see an oscillator-based system entering trades against the dominant trend. Note that the trades entered in the direction of the major trend, which is lower, would have generated profitable results.
1. Oscillator indicators and systems driven by them can be quite effective on sideways, trendless days.
2. The major weakness of these tools is their overwhelming tendency to continually enter positions against the major trend on trending days.
3. Oscillators are very effective entry tools when they can be used to identify corrections in the major trend, thereby allowing a system to enter the market in the direction of the dominant trend.
MULTIPLE INDICATOR SENSITIVITY SETTINGS
In the previous chapter several charts were presented showing the actual trading signals generated by the oscillator-type indicators that are the basis for a large portion of our trading strategy being developed in this book. A great number of buy and sell arrows were placed to demonstrate the accurate location of these points. Observing these charts its obvious that there are many more signals generated by these formulas than would be practical for any reasonable trading program to use effectively. It is therefore necessary to filter out many of these generated signals, hopefully retaining the most profitable ones for use in our developing system.
One of the more effective techniques I have used to accomplish this task is the use of two rather diverse sensitivity settings for each indicator. It will frequently be noted that a relatively low sensitivity setting for an indicator will give buy or sell signals very quickly after the anticipated event has occurred. The problem here is that the signals are usually too great in number and will regularly occur well in advance of the significant high or low turning points we want to capture. Although selected signals generated by these settings are useful when observed on a historical chart, a method by which the inappropriate
[start]           [ 11 ]