In technical analysis we are primarily interested in price, and secondarily in volume (and close relatives like open interest). However, we make use of many derivatives of price (like moving averages, momentum and big name indicators derived from these, including MACD, RSI and SSTO). There are some indicators based on volume or volume and price, and often these focus on discerning patterns of accumulation and distribution (although not all accumulation/distribution indices make use of volume). This contributions discusses the advantages of combining volume into our indicators and illustrates some of the possibilities.
Most technical analysis focuses on trends, which represent consistent direction of price movement. Graphically, our eye picks up consistent trend lines and channels, which correspond to a consistent speed of advance or decline. In Physics, the term velocity is used to denote this combination of speed and direction, whilst momentum is defined as the product of velocity and mass (or informally weight). In Physics, an increase in volume results in a proportionate increase in weight. In technical analysis too, volume is recognized as a major factor in assessing the strength of a market.
Thus using the term momentum for assessment of the speed and direction of the market irrespective of price is something of a misnoma, so we will discuss common indicators that do not take into account volume in terms of physical analogs like speed, velocity and acceleration, and then look at indicators that correspond to physical properties like momentum, force and weight, energy, work and power. What are the underlying characteristics of the market that correspond to these?
Disraeli's aphorism about statistics is perhaps nowhere more apt than in application to the financial markets. In a context of political and financial expediency, misinformation and manipulation, technical analysis recognizes that it is important to focus on the truths that the market itself shows you. All indicators have some basis in mathematics and statistics, but averages and differences are only first order statistics. Higher order statistics can be even more revealing. Of course, all statistics can be misused, and technical analysis indicators and products are no exception.
Bollinger bands, Black and Scholes' pricing formulae, and Steidlmeyer's Market Profile and Distribution Theory are all grounded in second order statistics. These thus relate more to the physical properties of force and acceleration. In the same way that the mass of the earth, keeps the moon circling, and the mass of the sun and the other planets influence its orbit, the mass of past transactions influences the various market cycles. MACD is also a second order indicator, and reflects the forces that cause the market to slow down and speed up - but it measures only the acceleration and does not take mass or volume into account. Parabolic SAR and parabolic trend lines are also second order indicators.
Neural Networks and Machine Learning techniques from Artificial Intelligence and many other Statistical optimization techniques also tend to optimize using higher order statistics, but as with any other indicators, different techniques and parameterizations will give rise to totally different predictions; and black, white and grey boxes alike are in danger of falling into well known traps for the unwary and statistically naive.
Trend lines on semi-logarithmic charts can also be viewed as higher order, but most parabolic trend curves will NOT map into straight lines on a semilog chart. Charting log prices is inconvenient, not supported by all the common packages, and precludes recognizing many trend channels, but nonetheless it is essential to take into account the fact that it is the relative change that is important in calculating gains and risks. So what is the best way of doing this?
As discussed above, some indicators are measuring the speed of the market, others give a truer idea of momentum, while certain rare indicators measure acceleration. We will review some of the common indicators, and introduce some new ones, in the context of exploring the physical analogy which is implied but not correctly captured in current terminology.
Market profile is one of the lesser known but more effective approaches to assessing likely market action. None of the major software packages are capable of displaying market profile, and those that do are very limited. The factoring in of volume leads to a much more precise tool. ChartAnnotate and CapFlow software are specialized systems for this kind of distribution theory profile, with ChartAnnotate allowing you to see a Time, Tick or Volume Profile at the same time as a conventional bar chart, a Point and Figure chart, and a standard indicator.
Many writers and speakers, from WD Gann to Larry Williams, have claimed that there are particular days of the week or month that are better than others for going long or short, or which are more likely to be a market top or bottom. We will examine two such claims, explain why one of them is plain wrong, and assess whether the other is worth trading.
The correct approach to testing a trading model, and the pitfalls which lead to incorrect conclusions when this established training methodology is not employed. Also we look at statistical significance, which tells us what the chances are that our nice profitable pattern is just a coincidence - it will you how often it is likely to occur just by chance irrespective of any underlying law of the markets.
The statistics which are used to price options according to the expected payoff are also useful trading the underlying security, being closely related to Bollinger Bands. The relationship between time and volatility is at the heart of this approach. Index futures, since they incorporate the activity of many traders and investors in many difference securities, are especially appropriate for this technique.
I am not going to be discussing any of the standard charting software, although I have found Chart4Free and MetaStock useful, and I understand that MarketAnalyst II 2.1 is a conventional charting package that has a Market Profile module available. But for maximum flexibility I prefer to roll my own, and I will introduce the free software, Perl and GnuPlot, that gives you more flexibility than the expensive commercial progams. I also discuss how you can preprocess and adjust data using Perl scripts that you then use in commercial charting software or with the charting capabilities of a spreadsheet - I use Sun's free StarOffice. The Unix, Linux and Mac afficionado's will be pleased to know that none of this free software is restricted to Windows - it originated under Unix.
The data I use is all available for free too: from the various internet brokers and exchanges as well as various other data suppliers. Currently ChartAnnotate displays and annotates free 10-15 minute delayed charts - usually charts from FutureSource are used with various indicators and time scales being selectable from minutely to monthly bars.
For those new to trading, Share Trading and its sequel Trading Tactics by Daryl Guppy are recommended as the most readable introductions. Guppy's Trading Tactics columns in various magazines and broker websites is also worth reading regularly, and he publishes his own electronic newsletter.
Alexander Elder's book is recommended for those experienced traders who need to develop more discipline in money management and want to learn a bit more about all the standard indicators, whilst John Murphy's book is recommended for its encyclopaedic breadth in introducing all the standard and many of the more esoteric technical analysis techniques, although it does not penetrate very deep.
Finally, Dalton's book is a good introduction to Distribution Theory and it is worth reading this, and having a general familiarity with trading methods and technical analysis, before trying to make sense of Steidlmeyer's own writings.
David Powers is Associate Professor of Computer Science and Head of the Artificial Intelligence Laboratory at the Flinders University of South Australia. He is a world renown leader in the area of Unsupervised Machine Learning - techniques for finding useful patterns in readily available data - with Data Mining and Knowledge Discovery applications focussed in the area of Speech and Language Processing. Other recent applications include WWW search, analysis of biomedical data and technical analysis of financial markets. Dr Powers is an active futures, share and option/warrant trader and technical analyst.
![[Free Newsletter shows you how to build a $100,000 Home Business]](../Banners/Banner2.gif)
Charts reproduced on this site are JPEGs that have been reduced to about 10% their original length by a lossy compression process and are thus may not be quite as clear as the originals. The original charts provided by FutureSource are annotated using ChartAnnotate, are reproduced with permission, and remain copyright by FutureSource (www.futuresource.com). ChartAnnotate is a trademark of David Powers and the ChartAnnotate program and annotations are copyright by David Powers.
ChartAnnotate and my various articles and webpages should not be construed as giving trading recommendations or as an encouragement to trade. They are intended to give you a better understanding of the financial markets from the technical and statistical perspective of a researcher into automatic analysis of numeric data, as well as from the practical perspective of a share investor and futures trader. ChartAnnotate only presents existing data in a different form and, as discussed in the license, the accuracy of neither the annotations nor the original charts can be guaranteed.
Daryl Guppy (1996), Share Trading, Wrightbooks
Daryl Guppy (1997), Trading Tactics, Wrightbooks
Alexander Elder (1993), Trading for a Living, Wiley
John J. Murphy (1999), Technical Analysis of the Financial Markets, New York Institute of Finance
James F. Dalton, Eric T. Jones and Robert B. Dalton (1993), Mind over Markets, Traders Press
Peter Steidlmayer (1989), Steidlmayer on Markets: The Information Revolution, Wiley