Kpos is the abbreviation for “Keltner Position” and summarizes many elements of the price movement in a single number. I want to share some ideas for using this powerful tool and give you some insight into how we think about it.
Channels in general
The idea of pulling some kind of band around the prizes is so good that there are roughly a ton of different variations of this tool! The most common are Bollinger Bands, where bands place a standard deviation (of price, not yield) around a moving average. This has the advantage of responding to a certain type of volatility in the market: the bands widen as the volatility increases. However, we haven’t found Bollingers to be the best at our job.
The problem with Bollingers is this standard deviation of the price is a strange measure. Better ideas could be the standard deviation of daily changes, daily returns, or possibly the distance price from the moving average. If you’re looking for ideas on how to look for improved bands, these are some great places to start.
In our work we have found that Keltner channels are the most useful. (Historic Footnote: The naming of this tool is kind of confused in history. It should be called STARC (Stoller Average Range Channel), but nobody will know what you mean by saying that. Chester Keltner was a grain merchant in the 1930s Who used a channel tool when they didn’t claim to invent it, and their channel doesn’t really look much like what we use today. This is a historical curiosity, but something you should probably be aware of! )
Keltner channels place bands with multiples of the average true range above and below a moving average. Entries to be considered are:
- What price should I use to calculate the average? (We use close for most applications.)
- Exponential or Simple Average? (We use exponential.)
- What period on average? (We use 20.)
- Which ATR multiple? (We use 2.25.)
The graphic above shows the channels represented on an active stock and you can see that the widths of the channels expand as the stock becomes more volatile. We find this behavior a little more stable than Bollingers and have developed a number of trading systems and noted many clear tendencies in relation to these channels.
What is KPos?
Kpos is a measure of where the price is on the channel, as a whole percentage (the percentage multiplied by 100: 100% = 100). So a market in the upper channel has a Kpos of 100, the moving average 50 and the lower channel 0. Kpos can be greater than 100 (above the upper channel) or less than 0 (below the lower channel). This value is available intrabarically or can only be calculated on closing, depending on what you want to do with it.
Ideas for Using Kpos
We’re doing a lot with this measure, but just let me share a few ideas with you.
First, it’s a quick overbought / oversold indicator. In general, we are very suspicious of overbought / oversold markets as these are usually markets that you would rather trade towards overstretch than look for a medium reversal. However, simply capturing the most extreme Kpos values of a group of markets will give you a brief overview of the most common ones.
It’s also a relative measure of strength, but maybe a strange one. Since the moving average adapts to the short-term trend of each market (which increases more rapidly in stronger uptrends), a market has to be even strong to register a higher Kpos. One thing to play with is ranking a group of markets by Kpos to see if this is a reasonable measure in your universe. “
You can also play games such as B. Track markets that are> 100 or <0 and then wait until they are back in the channels. These sometimes lead to nice withdrawals that can be quantified using the Kpos history.
Another cool idea is to organize subsets of stocks by Kpos. In the table above, we took the S&P 500 components and averaged the Kpos values for each component. This gives us unprecedented insight into the trend, extent and relative strength in each sector. In our weekly high-level analysis we use many variations on similar concepts. (A weekly review is so much more than just looking at charts!)
Hope these ideas are helpful, interesting, and maybe give you some ideas for your own work.