With (arguably) more and more efficiently priced markets, as technological advances have led to high-frequency flash trading, are we moving closer and closer to the world of Fama’s efficient market hypothesis (EMH)? My personal analogy to the Grossman-Stliglitz paradox, i.e. zero-arbitrage equating to arbitrage everywhere, is as follows: imagine you’re on a date, just you and the girl, no one else is around to hear your conversation; if you’re in a club, without space to breathe, again, no one can hear the two of you speak. So once all opportunities for arbitrage have been exhausted, there can be no more arbitrage; likewise, if you’re surrounded, it’s as if you’re on your own, in a paradoxical way.
I personally don’t believe that we’ll ever reach Fama’s EMH world, but that’s not to say I completely ignore all predictions that follow on from EMH reasoning, in some cases, they may hold as an approximation. So, what’s this got to do with the ‘head-and-shoulders’ pattern?
The simplest way of outlining a ‘head-and-shoulders’ pattern may well be as follows: “a chart pattern with 3 successive rallies, tracing out a ‘head’ with ‘shoulders’ either side.” It is certainly a classic technical set-up. As long as there are well-defined rules for market entry, risk management and profit-taking, it can easily be an attractive set-up for any technical trader’s repertoire.
The middle, ‘head’, peak is necessarily the highest, with the two outside, ‘shoulders’, peaking lower. The reaction lows, following the first ‘shoulder’ peak and the second ‘head’ peak, may be connected to form a neckline support.
A few things to note: firstly, there must be an initial up-trend for a ‘head-and-shoulders’ pattern to work (though a down-trend may in some cases allow for the formation of an inverted ‘head-and-shoulders’); the first reaction low immediately following the first ‘shoulder’ should remain above the trend line, such that the up-trend carries forward; the second reaction low following on from the ‘head’ should then break the up-trend line, such that the up-trend is no longer carrying momentum; the ‘right shoulder’ may be thought of as a ‘false move’, once the neckline support is broken past the ‘right shoulder’ in a convincing manner, the anticipated market movement would then be a significant fall in the price of the asset being traded.
The slope of the neckline support may typically influence the expectations of the extent of the decline in the market price following the pattern; a downward slope would be a more ‘bearish’ signal than an upward (or horizontal) slope.
The neckline break is typically deemed to be ‘convincing’ if it is accompanied by an expansion in volume, as measured by OBV or Chaikin Money Flow. Ideally, volume during the advance of the ‘left shoulder’ should be higher than during the advance of the ‘head’. This fall in volume, along with new high that forms the ‘head’, serves as a signal. The next signal comes when volume increases on the decline from the peak of the head. Should this be followed by a further increase in volume upon the decline past the ‘right shoulder’, this would then be the final confirmation to short, short, short.
The projected price decline is typically based on the distance from the neckline support to the top of the ‘head’. This distance is then subtracted from the neckline support to reach a price target. This should serve as a rough guide, amidst other factors such as previous support levels and long-term moving averages.
EMH believers would tell you that the ‘head-and-shoulders’ pattern would never work in reality, and tell you to forget about all technical set-ups in fact. Though markets are being priced more and more efficiently, I believe that from the basis that market participants are human, and to the extent that human nature has characteristic traits for irrational exuberance, bubble-behaviour is inherently imbedded to the functioning of markets, i.e. non-fundamental market movements with potential for profitable trading will almost certainly exist out there! So, why not give the ‘head-and-shoulders’ a go?
Comments