Option Investor
Trader's Corner

More about Keltner Channels

Printer friendly version

This week's Trader's Corner continues a series of articles on nested Keltner channels. As noted in the first article in the series last week, I use them to determine where support or resistance might lie, make judgments about the strength of that support or resistance, and ask if/then questions that help determine when assumptions about the market might be confirmed or violated.

I also use them to determine when markets are in a runaway mode, times when countertrend or so-called fading-the-move trades should not be attempted, and to set upside or downside targets. Several of those topics were addressed in last week's article. This article builds upon that discussion.

When I prepared the Wrap for Thursday, March 15, I included this nested Keltner chart in the "What about Tomorrow?" section.

Annotated 30-Minute Chart of the RUT:

Determining that support looked strong enough to propel the Russell 2000 into a test of resistance was easy enough, wasn't it? The RUT was already moving up into the close, already testing resistance. However, my annotation had been made prior to the last 30-minute candle. I had begun annotating charts several hours previous to the close with the intention of editing them if needed. My original annotation wasn't based on what happened near the close but instead had to do with the way Keltner lines were configured, helping me to determine where support or resistance lay and which appeared stronger.

That chart's annotations also pinpointed potential targets. Determining the relative strength of support and resistance and setting targets are two of the ways that I use nested Keltner channels. They'll be discussed in this week's article.

Last week's introduction on Keltner channels employed one channel at a time, showing how effectively Keltner channels can pinpoint likely support or resistance. These channels, built around a multiple of a central moving average, can be set so that their boundaries contain most price movements of a security. Breakouts--periods when momentum controls price movements--can be identified as strong momentum carries prices above or below channel boundaries, converting upper channel lines to support lines and lower ones to resistance lines for the duration of the breakout.

However, I find Keltner channels most useful when several are nested together on a single chart. QCharts and QuoteTracker both allow the channels to be nested in this way, although the QuoteTracker guidelines say that they cannot be nested. I do it every day! ESignal also allows it, although traders have to import the formulas themselves or they did when I investigated eSignal several years ago.

I didn't invent the idea of nesting Keltner channels. I'm indebted to Jerry Wawrzeniak ("The Case for Keltner Channels") for the idea as well as for the settings I employ. Those settings are as follows:

Widest Channel: Based upon a 120-ema, 7.2 multiple
Next Widest Channel: Based upon a 45-ema, 3 multiple
Narrowest Channel: Based upon a 9-ema, 1.4 multiple

Nesting the channels together on a single chart produces a chart like the first one in the article or the one found below.

Annotated 7-Minute Chart of the SML:

The confusion of basis lines and outer channel lines proves disorienting for some traders, but there's much information to be gleaned from this chart once traders become accustomed to the confusion of lines.

First, upside moves tended to be stopped at the outer channel, showing that this channel line did have relevance to the then-current trading pattern. As prices were rebuffed from upper-channel resistance and then rebounded again, they tended to stay in the upper half of the channel, showing an overall bullish tenor during the time period covered. However, that bullish tenor was weakening slightly toward the right-hand side of the chart as the black-channel's support slipped just below the 120-ema, the basis line for the largest channel. These two lines tend to be important support when prices are above them.

However, support lines were still converging thickly enough that they should have held, especially since those two lines mentioned tend to be important support when prices are in the upper half of the widest channel. One of two occurrences would likely have been needed to break that support, and neither happened that next Monday morning. Either SML's prices would have to be gapped below that converging support or else prices would need to batter at the converged supporting lines until they separated further, weakening support enough that prices could slip lower.

That slight weakening had shown up in another way, too. Up until about 11:00 Friday, 3/16, candles were mostly closing 7-minute periods above the pink line, the 45-ema that is the basis line of the black channel. Not only had prices been staying in the upper half of the widest channel, but they had also been staying mostly in the upper half of the next widest channel. However, about 11:00 on Friday, candles began closing 7-minute periods under that 45-ema. The pink line began serving as resistance rather than support. Candles were still in the upper half of the widest channel, but these two changes noted were alerting traders of the possibility of some weakening, a possibility that wasn't realized when trading opened the next Monday.

However, that warning about potential weakness could have been gleaned from other sources, too. The second test of the upper channel line, at about 10:20-10:30 on 3/16, had been accompanied by bearish price/RSI divergence. When prices rose again, they rose into a lower high and flattened. Pretty obvious stuff, isn't it?

What wasn't as obvious, however, is where the SML might have been headed when it broke above the black channel, as it had done on the far left of the chart. That is, it wasn't obvious unless traders were studying these nested Keltner charts. The breakout above black-channel resistance immediately set an upside target at the widest-channel's resistance.

What happens if prices break out of that widest Keltner channel, as they were ultimately to do that next Monday morning? How is a potential target set then? In that instance, it's time to dial up to a longer time period. Depending on the security being watched and the type of trade being contemplated, I often transition from the 3-minute to the 7-minute charts and then to 15-minute, 30-minute and 60-minute charts. These intervals might be used for scalping, day or sometimes even swing trades, but the 15-minute and up intervals also might be used to perfect the entry or exit on a longer-term trade such as a position trade that might be held a couple of weeks. When considering longer-term trades or when wanting to put short-term trades into the context of the overall tenor of the markets, I look at 240-minute, daily, 3-day and weekly charts. It may not be possible to view all those intervals with all charting services. My current service does not work well with 60-minute charts; my previous one did. My current service allows 3-day charts, which turn out to be surprisingly useful, but it doesn't allow the 240-minute ones that I used to favor when using another charting service. Experiment.

Let's look at some other examples.

Annotated 15-Minute Chart of the XAU:

After being rebuffed from the resistance found at that target channel level, prices fell, and candles began closing 15-minute periods back inside the black channel again. That was a signal that the upside move had likely concluded. What was the downside target? Since it was the widest channel's resistance that had rebuffed the move, the logical target was the basis line of that channel, the aqua-colored line on my charts. Although prices could of course have continued moving lower, a test of that channel's basis line is all that can be presumed when setting targets. As this article was edited Thursday evening, March 22, that's all that had been hit, too, as the XAU continued bouncing from this central-channel support on each test. That 3/16 low that tested central Keltner channel support had not been breached as of the close on 3/22.

What judgments can we make about the strength of support and resistance from that chart? Let's take another look.

Annotated 15-Minute Chart of the XAU:

The converging of various Keltner charts allows traders to make some determination about whether support or resistance is stronger.

I moved the candles to the left to make the annotations so that traders could confirm that this chart was snapped on 3/16, so that the judgments voiced there were not made up after the fact. At that time, it would have been possible to make some statements about support and resistance and likely targets for the XAU. An end-of-day move had barely edged the XAU above the thin red line seen above, the 9-ema that is the basis line of the smallest channel, the blue one. Until then, that line had been providing resistance as the XAU fell inside the black channel again. I tend to distrust last-minute movements such as these. Essentially, the XAU was still mired in the converging channel lines that were providing resistance and hadn't truly broken free.

Support and resistance looked about equally weighted, with support having enough of an edge that the XAU might at least go on testing resistance if not breaking through it. This conclusion comes because the channel lines forming support lines tend to be a bit more important than those forming the resistance that was in play at the time the chart was snapped.

Therefore, at the close that day, it appeared that support would likely hold unless one of two things happened. Either a strong downside move would be needed to blast prices through that support or even gap them below it, or a period of choppy movement would be needed to separate the channel lines providing support. Otherwise, support would likely hold.

Short-term bulls had something to worry them at the close that day, however. Notice that there's nothing between those converging support lines and the lower channel support at 127.9998? Support was weak below the level being tested at the close. A push through that support on Monday morning could have produced a quick decline toward that support. That didn't happen but knowing what could happen could help traders prepare trading plans for that Monday morning.

More traditional technical analysis would have already noted the potential head-and-shoulder on this chart as of the close 3/16, complete with bearish price/RSI divergence as the head was formed. Although I don't trust H&S formations, and particularly don't trust them to meet their targets, they are useful to watch. It is notable that if this one had confirmed, the downside target would have been near the bottom Keltner channel line.

If prices continued climbing Monday morning, as they did, the chart seen above predicted that first short-term resistance would have been found near 134.63 on 15-minute closes, although any strong rise in prices would have shoved that black line up toward 135.51 resistance at the top channel line, too, so it would have appeared possible that resistance wouldn't have been found until that level.

Let's look at what did happen.

Annotated 15-Minute Chart of the XAU:

Nesting the Keltner channels together has allowed for some judgments as to whether support or resistance is stronger. It's provided for a way to make judgments about the likely next action and to set targets if resistance or support were broken.

Do Keltner channels always work? Of course not. However, they do give traders ways to judge whether and when their assumptions about the next action are being violated. That's the topic for the next article, next weekend.

Trader's Corner Archives