If financial analysis were done during medieval times, the technical analysts, particularly "chartists", would probably have been burnt at the stake. Their methods of successfully predicting future price movements are not always easily understood, making it seem like some form of black magic. Even today the science of charting is reluctantly called a "science", more commonly referred to as an "art". Yet it is a fast growing method of analysing the increasing amounts of stock price data globally available.
Charting refers to technical analysis that is done through the painstaking investigation of share price data to identify predefined patterns emerging over time; for example, "head-and-shoulders", "channels" and "wedges". Chart patterns form the foundation of technical analysis as it is known today. The formation of such patterns in stock charts help predict the future trends of the stock prices.
The patterns are formed between support and resistance lines, which are imaginary barriers that the price stubbornly refuses to cross during the lifetime of the pattern. A support line can be explained as a price level at which most people in the market would purchase the share, resulting in the price never going lower than that specific level. Conversely, a resistance line is a price level at which few people are still willing to buy - rather they would sell at that level, rendering the price incapable of climbing any higher.
Despite the fact that every book written on technical analysis speaks of such patterns, and that these patterns form the foundations of technical analysis, charting remains a manual task that requires a human to visually inspect every chart, looking for any of a variety of predefined patterns.
Using human analysts to look for chart patterns among hundreds of price charts remains an expensive and time-consuming route to follow, and it has its drawbacks. For one, a person is not equally cognitive of all patterns at the same time - they might notice a Channel while failing to also see the Rising Wedge. Furthermore, the time it takes for one person to analyze one chart is too great to make analysis of entire markets viable.
Yet, many investors are very comfortable using chart patterns as they offer an easily understood set of signals that are well documented in technical analysis literature. There is a large overlap amongst different views on charts, which form the basis of the technical analysis knowledge known by the public.
Recently a South African company has come up with their own technology for automating the technical analysis process. Their web site, jse.autochartist.com, offers daily reports on chart patterns that were found by the software application that they had spent the last two years developing. It currently scans all JSE listed companies, unit trusts, precious metals, commodities, JSE indices, international indices, major foreign currencies, futures, bonds and warrants. A total of more than 2500 price charts are scrutinized daily for any one of 14 different patterns â€“ a task that would take human analysts weeks to complete.
A full description of each pattern is available for quick reference from any report, and a graphic image of the price chart is presented for each pattern, with the support and resistance lines highlighted on the image. This presents the investor with enough information to decide whether they want to invest in the stock or not.
Autochartist.com's primary offering is in the time it saves investors from not having to manually find chart patterns. Having a summary of anything that could be interpreted as a pattern allows the investor to focus on the interpretation of the pattern, rather than on finding it. There exists a definite trade-off between finding three or four extremely well-formed patterns among 80 stocks, and finding 30 possibly-well-formed patterns among 1000 stocks. In the former case, you would need the cognitive ability of a human with a certain level of experience and at least an hour or two in time. In the latter case you would need basic knowledge of chart patterns, and less than an hour to decide which of the 30 possibilities you would invest in.
Although Autocartist.com monitors over 2500 price charts for 14 different patterns every day, not all these patterns are equally reliable in their predictions. Furthermore, every pattern found by Autochartist.com has two further conditions that highlight the strength of the pattern found; these are a Start condition and a Volume condition.
The Start condition is an indication whether or not the pattern that has been found conforms to theoretical trends prior to pattern formation. The Volume condition is an indication whether or not an increase in the volume existed when the pattern crossed the support or resistance lines.
Despite the fact that charting theory stipulates the necessity of an increase in volume for a definite pattern completion signal, the developers of Autochartist.com are doubtful to the usefulness of this condition, but emphasize that the existence of a Start condition (initial trend of the pattern found) increases the reliability of the pattern tremendously.
For beginner users of Autochartist.com, or beginners to Charting, one should keep a close eye on the four different Wedge patterns that are found. Of four wedge patterns, 3 have a prediction reliability of over 75% over the past two years. On average however, statistics on the site indicate that the patterns detected by Autochartist.com correctly predict the direction of the market at least 2 out of 3 times - proof that the age-old art of charting has much to offer in terms of reliability.
Autochartist.com offers a fresh new look at traditional technical analysis techniques and should prove to be an invaluable tool in the hands of any serious investor.