By Vlada Raicevic*

I've been getting some email messages asking for more information about how I trade, what I look for and how I've set up my trading workspace. This article started out trying to answer that question, but, as usual, I was waylaid by my own curiosity when I started talking about stacking indicators. I kept seeing charts that I wanted to post showing what I had found, and it turned into a specific discussion on stacking the stochastic.

The first thing I'd like to talk about is the stacking of multiple indicators into one panel. I started experimenting with this two years ago and have found some interesting patterns. What I mean by indicator stacking is the placement of several different settings for an indicator, one on top of the other.

For example, many people use stochastics. And, there are some standard settings that are used without question. Each setting has its strengths and weaknesses, with some being too slow and others being too fast. These settings can, and should, be used according to the current market. Lately, the market has been in a rather tight range, and I changed one of my slower settings of (27, 9, 4) to (17, 7, 3).

The problem was that the slow setting was just sitting there, barely moving and giving me almost no useful information. In markets that whipsaw, a fast setting can also be rather useless because it continues to move from overbought to oversold too fast, again, not giving any useful information.

Just when it seems that stochastics are of almost no use at all due to market conditions, I turn to stacking several different settings on top of each other. Most charting services allow this stacking, and they all follow the same process: First, you create several individual stochastics studies, and, then, you just lay them on top of each other.

For example, in eSignal, you would hold down the SHIFT key while dragging one indicator on top of the other to merge them. To unmerge the studies, just right click on the indicator pane, and select the Un-Merge-Studies label.

When I first started out, the first thing that I tried was putting together the settings of (21, 3, 1) and (11, 11, 1), which is how eSignal displays it.

This produces a single line for each of the two indicators, and the two stochastics are now a single indicator with two lines (see Chart 1).

Rather than trying to use the crossovers of the %K and %D lines for a single setting, I'm using the interaction of the two indicators with each other for generating possible trade signals. The vertical purple lines are there just to show where the stochastics values are with respect to price.

First, look at the area labeled A. Here, you can see that both lines are rolling over in unison from the overbought area of 80 (yellow line). This uniformity of action is a fairly strong signal and can be traded. The area labeled B shows where both of the indicators are still in unison, and are crossing the centerline, yet another trade signal, or one that keeps you in the trade if you went short at “A”. Once the stochastics get into oversold territory, below 20, you can pull out of a short trade or wait to see what happens from here (or cover half of a position).

Now, look at the area marked C, where the (21, 3) setting (red line) crosses the centerline but does so without the (11, 11) stochastic. In fact, at this point, they cross the centerline 5 bars apart, and soon thereafter price fails in its move upwards. The pattern that I've noticed is this: If the two settings cross over together or within 2 bars of each other, it is a fairly strong signal. Once the crossing is 3 bars apart, the signal is weaker, and anything over 3 bars apart often fails as a tradable signal as shown in Chart 1.

This indicator is like any other in that there are false signals and exceptions to the rule, but, for the most part, the 3-bar crossover rule seems to work rather well. Let's take a look at another example (see Chart 2).

At the area labeled A, you can see that, even though both of the settings were not working in tandem, they still crossed the centerline together. The crossover at A ended up being a good signal for a long entry. At the area labeled B, you can see how the two settings are working at cross purposes, with one rising and the other crossing down. The start of the price rollover was bought and the dual stochastics kept you from a losing trade.

This dual setting works well with price that has some momentum behind it, but, in a very choppy, narrow range, the two stochastics tend to lag behind. In Chart 3, the vertical lines show where the stochastics cross over the centerline in unison. In both cases, the signal was of poor quality. The first one signals near the top, and the second signal is given just before a strong rise that would have stopped out anyone short.

So, my search for a better stochastic for the day’s market was not over. I needed a better setup. The next idea was to shorten the settings and make them a little more similar to each other to avoid the strange formations that the first attempt seemed to produce.

The next chart (Chart 4) shows a single pane with the following three stochastic settings: (17, 9, 1), (13, 7, 1), (11, 5, 1). This is in eSignal format.

I looked at months of data and noticed that a fairly persistent pattern emerged. The three lines give rather good signals when they act, for a better word, cleanly. Note how boxes A and B both show a bottoming, and then a clean crossover, as fast, slower, slowest lines reverse and cross over each other one by one.

The key in this pattern seems to be that, once they cross, they either continue to diverge from one another (pull apart), or they stay at an equal length apart. Even the large, red stick in the rise on the right-hand side doesn’t change the general slope of the three lines. Although the fast line does turn down, it still stays within the trend. Also notice how the price chart shows the curve of the rise in price increasing while the slope of the stochastics remains the same and actually starts to curl over as it moves into the overbought area. Soon after, price rolled over hard from such extremes.

The reversal patterns shown in the boxes in Chart 4 are some good signals, but they are not the cleanest or best signals. The next chart (Chart 5) is an example where a crossover, like the ones above, is actually messy compared to the base-and-reverse pattern.

The area marked “Messy Top” in the stochastics pane gives a reversal signal when all three lines cross and start to diverge. Still, price had some unfinished business because it tried to break above recent highs several times. The stochastics kept falling, and only the fast (black) managed to cross over the other two lines, with hardly any convergence of the other lines to show weakness. One might have stayed in the short, but would most likely be stopped out.

Look at the two arrows on the stochastics indicator. They point to lines that are both in overbought / oversold areas and have gone flat and are basing. Then, all three diverge from each other and point up, crossing the upper boundaries at 20 and 80 and giving nice signals for a long and a short.

This idea of “unison” with the three lines is an important aspect in the reading of these stacked indicators.

The first vertical line shows how the stochastics lines didn’t get too oversold, but they did get into that unison phase where they were all together, crossed up and diverged from each other. As long as they are diverging and moving in the same direction, crossover of the fast line over the centerline at 50 can be taken as a signal. The signal in Chart 6 gave a nice rise up, which actually formed a bearish divergence at exactly the same time that the three stochastics lines converged into a tight unison grouping in the overbought level, a simultaneous crossover and rollover occurred and gave a nice sell signal, which could have been taken because we have:

1. Divergence
2. Complete convergence and unison in stochastics
3. Rollover from overbought
4. Divergence between the three lines (they are moving apart from each other).

In the next crossover from the bottom, you can see that there was no unity in the actions of the three stochastics lines, and a crossover signal, while good for the overall trend change, came from a messy setup, and would most likely have stopped out a long.

It all looks great, and the results are much more consistent than in the first attempt at using twin stochastics lines. While it would probably be prudent to wait only for the best signals, such as the unison-and-reverse pattern mentioned previously in this article, you could lose out on some nice trades while waiting for that perfect setup.

Chart 7 shows a nice long rise in price that was not preceded by a unison pattern. Like all indicators, the multi-stochastic cannot be used alone, and, in order to have caught the move shown in Chart 7, we would have needed additional information. The next chart (Chart 8) adds RSI and a 78-period Regression Channel (standard deviation: 2) to the mix and will help us either filter out some bad signals or verify some good signals from the multi-stochastic.

In Chart 8, you can see a nice setup, where the stochastics merged and then started to roll over just as the RSI crossed over and verified the short signal. After a nice decline, you can see how price pierced the bottom of the regression channel. And, as shown in Chart 9, you see a doji reversal candle, and the stochastics start to converge in the oversold area, giving you a good place to exit the scalp trade.

The following chart (Chart 10) shows how even loose groupings of the multi-stochastic can give good signals when used in tandem with other indicators.

At loose grouping B, the vertical line is drawn just after price gapped above the top of the regression channel. I don’t like to take trades on indicator settings after gaps because they are skewed, so I thought it best to wait. As I watched, there was a slight selloff; another push was made and was again repelled at the upper channel.

Now, you have a gap, a failure, a push into regression resistance and a stochastic that didn’t react at all to the push-up after the gap was sold. Not only that, the RSI was still under the moving average and the MACD was moving down and did not cross over on the push-up. This second failure at the regression channel is a fine place to go short. Covering the short was easy as well, at the bottom of the regression channel when the stochastics had flat-lined, and the RSI was starting to turn up along with the MACD (labeled Loose Grouping C in Chart 10).

The bounce-up could be played long, or, because the slope of the regression channel is pointing down, you can just wait for another shorting opportunity. Once price reaches the upper channel, there is another rollover by the indicators, but the rollover at the area of Loose Grouping D is very sloppy, with no unison.

Even this leads to a decent selloff, but it is reluctant selling, as is shown by the rather sloppy movement of the stochastics and sideways motion of the RSI. If you chose to take the signal at Grouping D, you would most likely have been stopped out. Again, one can choose to trade these signals or wait for the best setups with the highest probability of success. It all depends on the risk one is willing to take.

With Chart 11, we will briefly discuss the centerline. The vertical line in Chart 11 marked “A” shows where the stochastics and RSI both give a long signal, but, because price is at the top of the channel, it is not a good idea to go long at that point. The stochastics pull back, but bounce at the centerline and start to move in unison upward just as the RSI crosses the moving average, giving a long signal.

With room to go up to the top of the regression channel, you can go long at vertical line “B.” The turn back down is a little sloppy on the stochastic, but the RSI crosses back down, and price has turned over again from the upper channel. With plenty of room to move down to the other side of the channel, a short could be taken at vertical line “C.”

Keep in mind that the centerline is a place of support / resistance. For some indicators, such as the RSI, it is not very strong support / resistance, and it is why I use a moving average and trendlines on the RSI. However, for the stochastic, the centerline plays a fairly key role, and it should be respected.

Chart 12 shows how the centerline will act as support as price continues to climb while pausing twice for a little profit taking.

I've tried to give some ideas on what to look for, and how to use this multi-stochastic setting. It can be a valuable tool, especially when used with other indicators for verification of signals or to keep you out of a trade due to a possible false signal. When you use it with support / resistance lines, trendlines and moving averages (for example, don't take a long signal when price is just below a pivotal moving average), you can reduce the number of imperfect trade signals and increase your odds of a good trade.

The examples in this article are all shown on 5-minute charts, but, as with any indicator or pattern, these carry over to any chart period that you prefer. Try a few different settings, and see how they act together at key reversals in price. If you see a pattern emerge, you may end up creating your own personal trading tool that matches how you prefer to trade.

*This article is reprinted from www.optioninvestor.com with modifications and permission.