Keltner Channels

Discussion in 'Technical Analysis' started by John9999, Feb 22, 2018.

  1. John9999


    I have come across some interesting info on this indicator. I trade NQ Index Futures, but I am sure the methodology could be applied to any trading asset. Go Long at bottom of channel and short at the top. I am sure that is way too simplistic, but maybe simple is what we all need

    Who is using these? thoughts?
    murray t turtle likes this.
  2. Xela


    So do I ...

    Yes, that methodology can indeed be applied to any trading asset, and in itself it will produce a similar result with any: an overall loss of money.

    As you rightly say, it's way too simplistic, and - as you can easily verify simply from backtesting it - it doesn't confer an edge.

    Keltner Channels are basically a variant of Bollinger Bands: the similarity between the two is that both show volatility-based deviations around a moving average of prices; the difference between the two is that with Bollinger Bands, the deviations are SD-based, while with Keltner channels they're ATR-based. That's about it.

    Here's a simple explanation of why what you've described above isn't profitable overall: the principle is one of "mean reversion", as many traders (perhaps unwisely) call it, and it lets them down on the positive-expectancy front because the mean itself is a moving one. In other words, the price can be at the bottom of the channel (lower line of Keltner, or lower Bollinger band: the principle is exactly the same), and "revert"/"rise" to the midline ("mean-reversion principle") but actually drop further in the process, because the moving average midline is itself also moving, while the price is. So, yes, it "works" in the sense that when the outer lines are touched/breached, eventually the price must revert to the midline ... and that's what many people find attractive about it, but the point they're typically missing is that that reality doesn't actually predicate which way the price will have moved, overall, by the time it's actually done so! [​IMG]

    (And that's without even mentioning stop-losses and risk management at all, which complicate it further.)

    Like so many things, it sounds good until you try to make profits from its "positive expectancy", the inverted commas in this case signifiying that it doesn't in itself have one. Not on NQ, and not on any other instrument, either.

    The specific mistake to avoid, with such things, is imagining that it "must" somehow be profitable, if only one "confirms" its "signals" by using it conjunction with another indicator as well. This line of "reasoning" is often an attractive and appealing one to aspiring traders who attribute to indicators predictive powers they don't really have - which tends to be something of a triumph of hope over experience.

    Of course, that's not to suggest that it has no possible use or benefit at all - but it's not capable of providing positive-expectancy trade entries, in itself, in the simplistic way described above ... perhaps not least, overall, because "entries" themselves are by no means the most important component of trading with steady profitability, anyway ... but again, the aspiring traders who attribute predictive powers to combinations of indicators do, in my experience, tend to be broadly the same ones who also confuse "entry methods" with "trading systems"; they often harbor, also, the illusion that "if they can just get the entries right, everything else will somehow, magically, manage to work itself out successfully". [​IMG]

    Hope these few thoughts help ...
    Last edited: Feb 22, 2018
  3. Metamega


    I use it myself more of a way to measure volatility, and to help filter bad setups(when moves have gone parabolic). Do some test yourself and see what you come up with.

    His book go me into using Keltner Channels. He puts some nice stats together in the link above. Take some ideas based off Keltner Channels and backtest.

    One thing I took from those stats( I’ve performed similar tests in AmiBroker with similar results), stocks tend to mean revert, and theirs a natural upward drift over the long haul. But you wouldn’t need Keltner Channels to come to that conclusion.
    murray t turtle likes this.
  4. Every indicator works and is logical and perfectly rational in hindsight.

    You can draw and trace lines and see pictures within them and create unique names for them since the dinosaurs roamed someone just discovered how to split the atom, or discovered electricity. Name a Law after each of those so-called magical subjective discoveries.

    You need to just isolate the day, or time frame you're trading,...and consider all the fixed and dynamic variables affecting its potential movement within that matrix.
    Trading is truly part art, part science. -- So it pays to keep an open, patient, malleable mind, o_O, :confused: and viewpoint at any given moment in time along that fixed chart movement environment box,

    But more importantly, it's 2018...High-Five`, ET extraterrestrial trader
    Chinese New Year of the some Sweet and Sour Beef/pork/chicken
    Make this year...the year, you become a SuperTrader in full effect.
    Put some sauce on dat thang,
    vanzandt and Pekelo...where you at...make a Killing in the Shock market.
    Johnny Rock...are you still Rocking -- or rocking in your chair, fading away,
    Last edited: Feb 22, 2018
  5. %%
    Like them every now + then:caution:[Edit note, the only disadvantage of posting something like this, seems to mess them up for while.LOL]I measure the main trend+ use that more than buy @ KC bottom or short @ KC top
    Last edited: Feb 22, 2018
  6. RedDuke


    Very easy to back test. Why even ask a question, test and see for yourself. I can tell you that end result will be in negative.
    Xela likes this.
  7. Handle123


    Any thoughts of what is trend? Or you using different timeframe?