back start next

[start] [1] [2] [3] [4] [5] [6] [7] [ 8 ] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52]


high end of its scale, the underlying market on which the indicator has been applied is conventionally deemed to be in an overbought condition. Such markets are expected soon to begin a downward corrective move. Conversely, oscillators that are approaching the lower levels of their limited range are said to be in oversold territory, the implication being that a rise in price is imminent. All of the oscillator-type indicators have popularly defined levels within their definitive ranges that are said to be important for trading purposes. Additionally, some of these tools consider other aspects, such as crossovers of various moving averages of their base values. When these occur in certain areas of their ranges, they are important considerations for the generation of actual trading signals. Since it is necessary to understand the conventional use of a trading tool before exploring our alternative application of the same indicator, we will now take a more detailed look at the conventional use of each indicator we will be using later in the book.

For the next several pages I will be using multiple charts to demonstrate the activity of each oscillator under varying market conditions. In order to point out the similarities as well as the differences between several approaches, I will be using data from a single stock issue, Comverse Technology, Inc. (CMVT). The charts are created using two-minute bar intervals to show easily an entire day on each chart in a limited space. Please be aware that these indicators and systems can be used on any market, any time frame. I am using the same market and the same time frame here to be able to make valid comparisons between the strategies examined in this chapter.


Stochastic, first developed by Dr. George C. Lane, is arguably the most popular technical indicator in use today. Having assisted many people in the development of trading strategies, I find that the stochastic approach is clearly the indicator most often used as the base from which to formulate an overall trading plan.

The principal theory of stochastic revolves around the tendency of closing prices to trend closer and closer to historical highs in an up-trending market and closer and closer to historical lows in a falling market. The stochastic function has one principal input that designates the number of bars over which the calculations will actually oc-

cur. This input is commonly referred to as the length input, as it designates the length of the period in question. It is this number that determines the historical highs and lows just referred to. These highs and lows are determined by the equations as the highest highs and lowest lows of the period defined by the length input. The calculations compare the relative positions of the most recent close to these extreme highs and lows. These raw values are then smoothed by simple moving averages to create the familiar plots of the stochastic indicator.

The classical interpretation for stochastic is to consider the market overbought when the stochastic values are above 80 and oversold when these values are at 20 or lower.

Figure 5.1 details the four main plots that make up the stochastic indicator as is conventionally applied. Note the overbought line at 80 and the oversold line at 20. The lines labeled Slow D and Slow are the two moving averages of the raw stochastic values that actually create buy and sell signals from the 14-period slow stochastic plot shown in the chart.

Although these values are the ones popularly applied to the use

9:32 10:02 10:32 11.02 11:32 12:02 12:32 1:02 1:32

Figure 5.1 The conventionalinterpretation of the stochastic oscillator recognizes the market as being overbought when the plots oftheindicator rise above 80 and oversold when the values fall beneath 20. Several buying and selling routines can be formulated from the various crossovers occurring above and belowthesethreshold points.

Chart created with TradeStation® 2000i by Omega Research, Inc.

of this indicator, many combinations of these variables are frequently used. For instance, one can give an upward or downward bias to a trading strategy by independently altering these overbought and oversold levels.

For example, lets say that a trader is expecting a falling market for the next several days, for any of a variety of reasons, either fundamental or technical. In this situation the trader would want to make it easier to enter a short position and more difficult to enter a long position. This can be accomplished in a variety of indicator input variations. For instance, the overbought and oversold lines can be manipulated. Lowering the overbought threshold from 80 to 65 or 70 increases the amount of overbought territory in which the system is allowed to enter a desired short position, thereby allowing considerably more latitude to select a selling point. One of the characteristics of a declining market is to exhibit relatively weak rallies. In this case, it is to the traders advantage to sell these rallies easily by lowering the requirements for the identification of overbought segments. Lowering the overbought threshold accomplishes this purpose.

In a declining market we also would expect price declines to extend beyond the commonly observed levels. Any system increases profits by capturing the maximum amount of profit available from any price movement. We want to capture an increased portion of downtrends in a declining market. To accomplish this purpose the trader, in a downtrending market, will lower the oversold line from, say, 30 to 15 or 20. This forces the market to even lower levels before a buyback of the short position would be allowed under the system.

The opposite is obviously also true in the case of an anticipated rising market. In this case the oversold line is raised to allow the shallow corrections in a rallying market to trigger profitable entries in the direction of the major trend. Also, following the same theory as with the declining market, the overbought line is raised to allow the extended rallies expected in this type of market to grow to their full profit potential before profits are taken.

Obviously the optimum use of these varied settings for the overbought and oversold thresholds of the stochastic indicator is totally dependent on the identification of the dominant trend for the day or group of days being considered. This is the subject of Chapter 10, "Directional Day Filter."

There are many variations in the use of stochastic, especially if

one varies the length input with the overbought and oversold levels. Unfortunately, these manipulations of stochastic inputs, even when combined with other common indicators and strategies, rarely result in a long-term profitable system.

In the remainder of this section we will use a basic stochastic crossover system to demonstrate the buy and sell points identified by the stochastic indicator. This system will use the standard slow stochastic 14-period indicator, with an overbought level of 80 and an oversold threshold at 20. Also, this system is set up as a reversal system with no exits other than the opposite signal. In other words, a long position can be exited only when a new short position is signaled, and a short position will be exited only on the arrival of a new long position. There are no stop loss levels used in this system. Additionally, to adequately demonstrate the accurate placement of all possible buy and sell signals as generated by this system, the strategy has been enabled to take all generated positions rather than a single position either long or short. Admittedly, it is quite unlikely that a system such as this one would actually be traded. It is constructed as it is here simply as a tool to point out more clearly the exact entries and exits as dictated by the stochastic indicator. In other words, dont try this at home.

Popular systems using the conventional stochastic interpretation issue sell orders when the Slow line crosses below the Slow D line when both lines are above the overbought line. Buy orders are generated when the Slow line crosses above the Slow D line when both are below the oversold line when the crossover occurs. For the most part, crossovers occurring between the overbought and oversold lines are ignored for the traditional interpretations of stochastic. There are multiple other strategies that utilize the basic stochastic indicator, including those that treat the 50 level, placed equidistant between the existing overbought and oversold lines, as a balance point attempting to issue buy and sell signals as the various lines cross back and forth across this level. This 50 percent line also finds occasional use as a reference point for stop and/or exit placement. Various other applications rely on divergence between price and stochastic or concentrate on trend lines from stochastic and basic price to generate buying and selling signals.

Figure 5.2 points out the principal characteristics of a typical buy signal and a typical sell signal as generated by a conventional stochastic crossover system. Note the sell signal when the Slow line passes through the Slow D line after first rising above the Slow D.

3.52 10:02 ; 10:22 1Q;32 1Q;42

Figure 5.2 Classic stochastic sell signals arise as Slow crosses below Slow D above the overbought line. Buy signals are generated when Slow crosses above Slow D in the oversold area.

Chart created with TradeStation® 2000i by Omega Research,Inc.

Additionally, these levels must be at a chart level that is above the overbought line when this crossover occurs. Two successive sell signals are detailed on the chart. The appropriate crossovers are also labeled in the lower graph.

The single buy signal occurs when the Slow line crosses up through the Slow D line after both lines are first located below the oversold line.

Figure 5.3 is an expansion of the Figure 5.2 chart showing all the trades generated from this system during the day in question. The signals used in Figure 5.2 are contained within the gray box in Figure 5.3.

The downward-pointing arrows signify the sell signals arising from the stochastic plots in the lower portion of the chart. Note that the arrows correspond with the specific line crossings on the stochastic indicator. The upward-pointing arrows identify the buying points as signaled by the 14-period slow stochastic indicator.

The chart is from a sideways day in the market for Comverse Technology Inc. (CMVT). Note that the price of the stock wandered within a $4 range for the entire day after establishing its daily range only one and a half hours into the session. This is an unusually quiet

Figure 5.3 Multiple signals may be issued when Slow and Slow D are both above or below the threshold lines. The darkened box identifies the trades detailed in Figure 5.2.

Chart created with TradeStation® 2000i by Omega Research, Inc.

day for this rather active issue. The 14-day slow stochastic was fairly successful in its identification of overbought and oversold areas on the chart, leading to several profitable buy and sell signals for the system.

Figure 5.4 shows the reaction of the system the next day when CMVT found itself in a pronounced uptrend for the majority of the session. The system used to generate the signals depicted here is the same system as used before, using identical parameters.

Note that, in sharp contrast to the previous chart, the same indicator-driven system had much less success issuing profitable trading signals on this trending day. In fact, the first 12 signals, all being shorting trades, would have had a difficult task turning much of a profit at all. Most reasonable stop loss strategies would have created losses for many of these 12 positions. Of particular note on this chart is the relative success of the system as it accurately identified excellent points for entry on the long side of the market on an up day. Note that the three buy signals issued just before 1:00 p.m. would have easily resulted in profitable trades using almost any exit strategy one would care to apply. In Chapter 10 we will describe in detail the use of the Directional Day Filter to identify this type of day early on, allowing our system to take only trades in the direction of the major trend of the day.

This chart painfully identifies the major weakness of most

[start] [1] [2] [3] [4] [5] [6] [7] [ 8 ] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52]