October 6th, 2009

Market Extremes Turn Coppock Curve Unreliable

It’s been a while since we touched base with Mr. Coppock and his Curve. If you remember, it was one of the indicators that signaled that the market was turning over in November, 2007 and one of the indicators that gave me confidence that we had hit bottom in April (along with the “Reversion to the Mean” indicator in both cases).

So if it was so accurate in calling the top and and the bottom, I was wondering how the indicator would perform if the market did take a tumble in the near future. Actually, the correct question to be answered is “what sort of a drop would it take for the Coppock Curve to again signal a market top?”

To find out, I projected forward from the September close (the Curve is based on monthly data only; see “The Coppock Curve, Another Constructive Indicator“) and came up with some disturbing results. They were disturbing but not in the way you might expect. What I discovered was that the extent and rapidity of the market’s recovery (50% in six months), either: 1) precluded any risk of a significant market top in the foreseeable future, or 2) the Coppock Curve is an inconsistent indicator and it’s being “on the money” concerning both ends of this past crash was merely a coincidence.

Here’s what the data reveal (click on image to enlarge):

  • A relatively flat market projected through June, according to the Curve, increases the chances a decline in the second half of 2010:
  • A decline to the March-April lows projected to occur by year-end followed by a modest recovery would not cause the Curve to turn down:

We could hypothesize hundreds of other future paths for the market. The bottom line purely from a technical perspective is, however, that last year’s decline and this year’s recovery have been so dramatic and extreme that a momentum indicator like the Coppock Curve appears to be less reliable for indicating significant changes in the long-term trend. Complacency due to the Curves current unward trend is, therefore, dangerous and to be avoided; other long-term trends should be considered.

I’m not an expert in the Coppock Curve and welcome alternative interpretations from others.

Subscribe below or click here to learn more about help for navigating turbulent markets.
  • Anonymous

    from a technical perspective, we are still in a bull run until the 200-day SMA (or 40-week) is breached.

  • Manoj

    Hi Guru
    Very good analysis. You are saying – that at extreme conditions (in mathmatics borderline cases) one indicator
    (even many)could be highly unreliable.

    What is your take on dollar?
    Looks like fed/gov..etc are determined to bring dollar down.

    Thanks
    Manoj

  • Guru

    Manoj,

    Rather than raising rates (which several Fed Board members have indicated won't happen for some time) the other recourse for drawing down deficits is to do it in nominal terms through allowing the dollar to deflate.

    It's protectionist, helps insulate against continued Chinese exports (and they're being pressed to increase their currencies value) and helps our exports. It's been a long time coming but was relatively easily foreseeable.

  • Dividend Inc. Team

    Greetings,
    My understanding of the Coppock Curve is that it is strictly for the purpose of giving buy signals. Sell signals are purely coincidental if they occur at all.

    Drawing from Mr. Coppock's own words in Barron's October 15, 1962 article, Mr. Coppock states that,"It [Coppock Curve] gives a so-called buy signal."(page 5) Mr. Coppock goes even further to state that, "Because well-timed buying is far more difficult for the nonprofessional investor than timely selling, it is best to think of the curve as a very long-term buying guide. Its formula was devised for that type of use." (page 5,16)

    In James Dines' "Technical Analysis" (page 377, 1972) There is not mention of the Coppock Curve as being able to provide a sell signal or eminent market slumps. Any mention of the Coppock Curve was with the ability of the Curve to "pinpoint the start of new trends and enable investors to select future market leaders." (page 378)

    There seems to be no evidence that would suggest that the Coppock Curve should be used to determine potential declines. Instead, the Curve should only be tested on its ability to accurately call the bottom in a given stock or index.

    Your consideration of the evidence provided may add value to additional tests of the Coppock Curve's value as a buy indicator.

    Best regards.