Charting Traps for the Unwary
Arithmetic Versus Semi Log

From time to time we have discussions with traders and discover that they are using charts that are not serving their intended purpose very well. The chart is the key tool for the technical trader, and hence the choice of chart type is very important.

Most traders become very comfortable about using an arithmetic chart, and for most of the time this is the most appropriate chart to use. When they examine a market that has had a large percentage increase or decrease in price, however, it is more appropriate to use a semi-log (exponential) chart.

Most charts in use today are arithmetic charts. On an arithmetic chart equal vertical distances represent equal price ranges. For example, a market move from 10 to 15 on an arithmetic scale would be the same vertical distance as a move from 100 to 105. The 10 to 15 move is an increase of 50 percent. The 100 to 105 move is an increase of just 5 per cent. When we consider trading profits, it is the percentage increase that is important!

The following chart is an arithmetic chart of the Dow Jones Industrial Average from 1910 to June 2002. Each bar represents six months of market action.

On this chart the 1987 crash dwarfs the 1929 to 1932 pull back. The pull back from the 2001 high to the September 11 lows makes both the 1929 to 1932 bull back, and the 1987 crash, pale into insignificance.

Now let us examine a six-month semi-log chart of the same market over the same period of time. The pull back from the 2001 high to the September 11 lows now pales into insignificance when compared with the Dow's decline between 1929 and 1932.

On a semi-logarithmic chart equal vertical distances represent equal percentage increases in price. For example, a market move from 10 to 20 (100 per cent) would be the same vertical distance as a move from 1000 to 2000 (also 100 per cent).

The Dow fell by almost 90 per cent between the high in October 1929 and the low in 1932. In 1987 the Dow fell by 41 per cent. The fall to the September 11 low was slightly more than 30 per cent.

Clearly, the long-term arithmetic chart failed us, as the large increases in the magnitude of the Dow over the last 10 years completely overshadow the massive percentage fall in the late 1920s and early 1930s. It is only by using a semi-log chart that we can see the 1929 to 1032 decline, the 1987 decline, and the decline to the September 11 low, in perspective.

 

Conclusion

The arithmetic chart is the most appropriate chart to use for most technical analysis. Remember, however, that when you are displaying data that has had a large percentage increase or decrease in price, usually longer-term charts, it is more appropriate to use a semi-log (exponential) chart. Fortunately, modern software will convert an arithmetic chart to a semi-log chart with the click of the mouse.

[This article was reprinted from Your Trading Edge Magazine, Aug/Sep 2002.]

"For more information on how you can maximize your trading profits while strictly controlling your risk, click here..."

DISCLAIMER:

Every effort has been made to ensure that the content and conclusions presented in MasterStroke are complete and accurate.

No part of MasterStroke contains trading advice - stated or implied, nor is an invitation to trade. The directors and associates of Market Masters Pty Ltd are NOT licensed trading or investment advisors. Market Masters Pty Ltd is an organization designed to assist traders and investors to become more knowledgeable and independent.

The giving of advice is therefore contrary to the very objectives of Market Masters Pty Ltd.

Traders requiring trading or investment advice should contact a licensed advisor. Stockbrokers and futures brokers are licensed advisors.

Neither Market Masters Pty Ltd, nor anyone else involved in the production of MasterStroke, will be liable for any liability, loss or damage directly or indirectly caused, or believed to be caused, by MasterStroke.

Traders, to be successful, must take full responsibility for their own actions.

With respect to trading results, past performance is not necessarily an indication of future performance.

By maintaining your subscription to MasterStroke, you acknowledge that you understand and accept the contents of this disclaimer.