What Counts? - And Can it be Counted?

"Not everything that can be counted counts, and not everything that counts can be counted."

-Albert Einstein

I always believed that there is nothing "technical" about technical analysis and the tenets of TA are essentially behavioural.

In TA, there are basically two dimensions, price and time. Price is considered all important while time is also looked at in many theories. Most TA systems would look at combinations of time and price and/or derivatives therefrom to identify an indicator that gives a signal to buy or sell.

Trend-following systems for instance try to identify the speed or relative speed of price action in either direction to decide the trend at a particular degree of time scale. A simple price based system would look at higher or lower highs and lows. A more complex system would look at say, relative speed now as compared to "x" periods back or a rate of change or a combination of moving averages or create a band of specified percentages around the previous price action etc.    

The Elliott Wave theory is based on the price structure. It proposes that human progress can be broken down into advance-correction cycles that in turn can be identified through repetitive wave-like patterns. Impulse or motive waves are followed by corrective waves, and further impulse waves. These happen at different degrees in a fractal fashion.

The EW construct in itself is not against any other theory of market behaviour. Thus Dow theory can be thought of as part of EW. Technical indicators can be used in conjunction with EW and EW itself can be used as part of time cycles. 

Its value comes from providing a context for market analysis more than as a forecasting tool. Though, correct identification of a market's general position can improve the forecasting ability, particularly at market extremes. Obviously, the popularity of EW suggests that EW helps in forecasting. If one could say that the market is in an impulse vs a corrective structure, then one can look to  buy on dip or on break-up or sell on rally or on break-down. And clear stops can be identified linked to what you think is the beginning of a particular wave.

Forecasting markets, in my view should not however, be confused with making money. In the first place, trade plan conception and execution require a high standard of discipline and not everyone's cup of tea. In a discipline like EW, it can get even worse, if the confirmation of a structure is far away, and you end up trading in anticipation or out of impatience. So you can get it wrong trade-wise even if you are eventually correct forecast-wise. Secondly, there is always an element of uncertainty arising out of the fact that TA is behavioural. So many a time you may have to disbelieve yourself as the most unlikely option to our logic oriented minds may be the only option to the behaviourally oriented markets. So a forecaster can get it wrong straight. I love Yelnick's theme (see yelnick)viz., "all forecasters will eventually be hoist by their own petard". Thirdly, experienced practitioners such as Neely say that EW cannot give you accurate predictions all the time, the middle of any pattern being the most difficult. Fourthly, in order to avoid multiple counts and bring about solidity to the EW, one has to use some "rules" (such as neo-wave)  that could be seen as probabilistic in nature (there cannot be any finality in human behaviour; can there be?) but enhance the overall predictability. Applying such rules and awaiting confirmations can be a huge exercise in itself. Lastly, it can be expected that as EW becomes more popular, new wave patterns and more complex behaviour will develop and to stay on top of such developments you need to invest significant time and effort.

To recap using EW, (i) you may be right but get stopped out, (ii) you may be plain wrong, since you are as much part of mass psychology as anyone else (iii) you don't know what's happening, the possibilities are too many, (iv) the rules to be applied are complex and too many and (v) new patterns emerge and complexity keeps increasing. 

With all these, if you make money systematically using EW, consider yourself as one of those very few who can count the waves that count.