The McClellan Oscillator

During the October 6 Real-Time Trading Session, member Jock asked, “Wouldn’t the McLellan Oscillator and Summation give a broader indication of market sentiment than the [Mamis] New High/New Low which reports only stocks at extreme levels?”
I went to the official website to get today’s chart:

A careful review of the calculation method of the McClellan Oscillator reveals that it is not a true +/- 100 oscillator, so here’s my take on it, done properly with a 63-day (one quarter) window:

Better yet — because so many of these “indicators” look good from far but are far from good — we make a paintbar version of it and color the price bars. This is one of the secrets that The Quant told me years ago for testing trading ideas: paint it first to make it easy for the eye and brain to evaluate the signals objectively rather than let them try to approximate the turns with the price bars.

Click the chart below to see it going back to 2006.

Does anyone want me to add this to the daily Market Barometers?
RELATED: Engineering Better Bollinger Bands
RELATED: Thoughts on the Kase Dev-stop


Submit a comment
  1. Wow, T – our questions about other indicators (McLellan in my case) seem to have kicked off “tour de force” night chez Lo. Thanks.
    One of the best aspects of your site is that it makes you really question what is worth looking at. On the surface, it seemed that advancers vrs. decliners would HAVE to tell you more about the overall tenor of the market than just 52 wk. highs vrs. lows.
    Frankly, I had forgotten that the McLellan oscillator looks so far back 19MA and 39MA of advancers and decliners. Where is the logic in THAT telling you anything about yesterday, today or tomorrow? There may be something “modern” and with less-lookback to be done with advance/decline, but McLellan doesn’t cut it. The charts look SO similar to the price chart itself, just truncated at the tops.
    I think people put a lot of credence in dubious cycles in the 60s. Reminds me of SKI’s gold timing model (from about the same period) with cycles upon cycles.
    Anytime we get to the son continuing to publish Dad’s indicators, I wonder if a re-think isn’t long overdue, and whether it’s more a matter of carrying on Dad’s “bidness”. People who become comfortable with those tools just keep on using them, it seems.
    The Better Bands are intriguing. Will you polish and publish them? I’ll certainly study them in depth. Thanks again for this most interesting exercise.

  2. Leave it to you to make something better. I have checked out the McLellan Oscillator for quite some time. Up to this time, I have doubted it was of much practical use to me other than for showing short term crowd mood. With your improvements it may have become useful.

  3. Thanks for the feedback. I’ve cleaned up the daily market barometers page and added this updated McClellan Oscillator because I think it is useful to look at advancers and decliners. Of course, their 60s style calculations have gone out the window. 😉

Comments are closed.