Traders often turn to charts when deciding whether it is time to buy or sell. Over the years, they have developed a wide array of charting techniques in pursuit of increased profits. Some traders believe candle stick charts provide them with an edge over traders while others turn to point and figure charts and some remain loyal to traditional bar charts. No matter what type of chart they use, traders invariably face another decision which is whether the chart should be scaled arithmetically or logarithmically. This decision, whether to use arithmetic or logarithmic scaling, can actually lead to emotional debates among technical analysts because some have developed a strong opinion about the superiority of their preferred method.
Before looking at the debate, and revealing the answer of which style is best, we need to define the terms. The scale we are talking about is the y-axis of the chart, the axis where prices are recorded.
Most graph paper uses what is known as the plain or arithmetic scale. With this scale, equal distances on the vertical axis represent equal amounts of dollars. With this chart, the distance between $10 and $20 is the same as the distance between $110 and $120.
Logarithmic scaling, or semilog scaling, shows equal percentage changes with equal distances. So, the distance from $10 to $20 would be the same distance as $100 to $200 since both changes represent a 100% increase. The distance between $120 and $110 would be about one-tenth the distance from $10 to $20.
- Screw Up All Of Your Trades And Still Bank Monthly Gains The Perfect Trading Strategy for risk-averse conservative traders who want consistent, predictable and reliable weekly and monthly income from trading stocks… even when… they are 100% WRONG on every trade. Over a recent 30-day period, a well-known trader used this conservative trading technique to earn a substantial $13,241.50. He explains everything (and shows you the PROOF) in his just-released video report. I won’t leave this video up forever. So watch now because you’re about to discover some things about active trading for weekly and monthly income you’ve never seen before.
The chart below provides an example of the differences. Arithmetic scaling is used in the chart on the left and log scaling is shown in the chart on the right.
In the short-term, the differences between the two scaling styles is almost imperceptible. Many traders would see little difference between the two charts. The differences do become apparent at longer time frames. The next chart is a yearly look at the Dow Jones Industrial Average.
Now the difference is notable. The chart on the right, the one with log scaling, is more compressed. The market crash of 1929 stands out in that chart while it is lost in the noise in the chart using arithmetic scaling on the left side. The 1987 crash, on the other hand, is visible in the chart on the left but the detail of that event is lost in the chart on the right.
There are obviously differences between the two charts in long time frames but do those differences matter to traders? That question is at the center of the debate between proponents of the different charting styles.
The origins of the debate can be found in the work of Robert Edwards and John Magee who cited the difference in the first editions of their classic book, Technical Analysis of Stock Trends. Edwards and Magee preferred log charts. They explained:
Percentage changes, it goes without saying, are important in trading securities. The semilogarithmic scale permits direct comparison of high and low priced stocks and makes it easier to choose the one offering the greater (percentage) profit on the funds to be invested. If facilitates the placing of stop loss orders. Area patterns appear much the same on either type of paper but certain trendlines develop more advantageously on the ratio (logarithmic) scale. Almost anyone can quickly become accustomed to making entries on semilogarithmic paper. We recommend its use.
It’s important to note the fathers of technical analysis added, “However, its advantages are not so great as to require one to change.” That sums up the current state of the argument seventy years later – it simply doesn’t matter which scale you use.
Interestingly, many traders are using log charts without realizing it. StockCharts.com, for example, is a popular web site for charts that uses log formatting in its charts by default. Other web sites use arithmetic scaling by default. No matter which scale is used, several important features of the chart remain unchanged. Patterns, and in particular support and resistance, are the same no matter what scale is used.
Traders completing a pattern analysis, which is of course the primary reason for using a chart, will see the same patterns in any scale. Edwards and Magee agreed with this point. A rectangle pattern highlighting a trading range or a head and shoulders pattern indicating a potential reversal will appear the same, in general terms, on both charts. The trading rules for the patterns will be the same on either chart. The call to action, a break of support or resistance in the rectangle or a break below the neckline in the head and shoulders pattern, will be exactly the same in dollar terms. The price targets, derived solely from the depth of the patterns, will also be the same on either chart.
For their argument, Edwards and Magee pointed to the superiority of trendlines on log charts. This shows their belief that trendlines carry importance on their own. We now understand trendlines are important only because they are followed by a large number of traders.
A trendline, or other chart pattern or indicators, only carries significance and profit potential if it is followed by other traders. Trendlines provide support only if buying develops when the trendline is touched. A trendline break is significant only if the price action continues to push prices away from the trendline. This happens only when a number of traders see the trend developing.
In the current markets, it doesn’t matter if the trendline is drawn on an arithmetic chart or a log chart. If other traders see the trendline and believe it is significant, it is important. If only one trader sees the trendline, it isn’t important. Since traders are using log charts on some popular sites, even if they don’t realize it, and arithmetic charts on other sites, trendlines tend to be of equal importance.
Now, back to that debate over which scaling method is better. The truth is, no single technique is ever the best one for all traders. Some will find log scaling better and others will find it distracting. Some traders will profit from charts drawn with arithmetic scales and others will suffer losses with those charts. Rather than debating how to draw the chart, it is better to remember what charts represent.
When looking at a chart, traders should be looking for important price levels including areas of support and resistance and patterns. These areas are important because of the psychological importance of the prices. Resistance is an area where selling pressure overwhelmed buying pressure before. Once resistance is broken, it’s likely traders who spotted the price move will be buyers. Resistance is important because it represents a call to action for other traders. The same is true of support and patterns like the head and shoulders. They are all important only if other traders observe them.
A trendline break using a line you drew using some secret technique that isn’t visible to anyone else is unlikely to be significant in the market.
Since a large number of traders use both chart styles, patterns on either style of chart are significant. You should use the chart that makes it easier for you to spot important price moves.
To demonstrate how irrelevant the choice really is, consider the fact that if you are using a site like StockCharts.com, you are using log charts by default whether you knew it or not. The fact that many traders don’t realize they are using that scaling method shows how unimportant the debate really is. Charts work because of the patterns they reveal, not because of the scale they are drawn with.