John Murphy said it best:

Remember that the chart only records the data. In itself, it has little value. It’s much like a paintbrush and a canvas. By themselves, they have no value. In the hands of a talented artist, however, they can help create beautiful images. Perhaps an even better comparison is a scalpel. In the hands of a gifted surgeon, it can help save lives. In the hands of most of us, however, a scalpel is not only useless, but might even be dangerous. A chart can become an extremely useful tool in the art or skill of market forecasting once the rules are understood.

What Do All Of These Letters And Numbers Mean?

The first thing we want to do is determine what all the numbers and letters represent. As you can see here, the y-axis is the vertical data set on the right representing price. The x-axis is the horizontal data set on the bottom representing time. As the time goes by, the chart is plotted to the right of the previous period.

Notice how I leave room between price and the y-axis. This is done purposely to allow for a visible area to determine the direction of the next price movements. These are called “Blank Bars” or “Extra Bars”. Look at a chart without any extra bars:

Doesn’t this one look so much better. It allows you to visualize the future progression of the trend:

Arithmetic vs Logarithmic Scale

Next, I’d like to discuss the differences between a semi-log chart and an arithmetic scale chart. When you are using an arithmetic scale chart, you are visualizing price in a linear format. This is traditional when looking at say, time and temperature. In this linear case that we refer to as an arithmetic scale represents absolute price changes. So the vertical distance between $5 and $10, although a 100% move, appears the same on the chart as the vertical distance between $50 to $55, which only represents a 10% move.

Semi-log charts show the logarithmic scale on the y-axis. These are therefore sometimes referred to as “Percentage Charts”. The vertical distance between $5 and $10 is the same as $50 to $100 since both represent a 100% move in price.

Logarithmic Chart

Here is a good example of 3D Systems, ticker symbol $DDD, using a logarithmic scale. Look how the price changes closer to the bottom of the y-axis are much further apart than the price figures near the towards the top. This is because the logarithmic scale shows percentage changes

Arithmetic Chart

Here is the same exact chart as above but except using Arithmetic price scale. Below is the Arithmetic scale, or linear chart format. Notice how you barely even see the price changes in 2008-2009. How much better does this look on a logarithmic scale? You can actually see what happened in 2008-2009. You see the problem with the linear charts using an arithmetic scale is that they improperly exaggerate the top end of the axis, but hide valuable information towards the lower end.

Supplementary Material Comparing Methods

I prefer to use Logarithmic charts as the default. This is especially the case for long-term charts with a lot of data on it and for charts that have very wide swings in price from high to low. I had the opportunity to interview one of the Founding Fathers of the Market Technicians Association Alan Shaw about this very subject. I definitely recommend watching this. He spent 46 years as an analyst on Wall Street so he talks about some pretty old school stuff. I think you’ll appreciate his perspective on things: Alan Shaw & JC Interview.

He makes a great point during the interview that for Day Traders Log-scale charts are irrelevant. They’re trying to make points on a stock. But for any kind of intermediate or long-term perspective, logarithmic scale charts are not only incredibly valuable but a necessity. How else can we really get the real perspective on the direction of the primary trend?

Over the years I’ve been lucky enough to become friends with some of the smartest technicians in the world. One time I reached out to them to hear their thoughts on the subject of Arithmetic vs Logarithmic scale charts. You can read that here: Arithmetic vs Log.

Using Multiple Timeframes

We always want to incorporate multiple timeframes in our analysis. This is part of the top/down approach. First, we start with a longer-term view, formulate a thesis, and then work our way down to a shorter-term timeframe. The first thing you need to do is define your time horizon. Are you swing trader looking at 30-minute and 65-minute charts? Or are you a longer-term investor looking at weekly and monthly charts? I personally look out weeks and months at a time, so I use a daily time frame. Therefore I need to go one level higher to get my longer-term view, so I use weekly charts. If you are a day-trader, you can use daily charts to get a longer-term perspective on price levels. If you’re a longer-term investor looking at weekly charts, you can go further and look at monthly charts to get longer-term perspective. The whole idea is to use multiple timeframes, preferably no more than 2 or 3.

So here we’re looking at the US Dollar breaking out in 2014 above a downtrend from the 2005 highs. We look to this as a bullish development and US Dollars are coming out of a huge base.
 

Then you work your way down to the daily timeframe and see the US Dollar not only breaking out above that downtrend line shown above but actually coming out of a shorter-term range as well between 79 and 84. Both timeframes are showing bullish developments.

Incorporating multiple timeframes is a vital part of my process. I would encourage everyone to regularly take a step back and take a look at the markets you’re focused on but from a bigger picture perspective.

I asked my friend Brian Shannon @alphatrends, who originally inspired me to consistently use multiple timeframes in my technical analysis, to send you guys a personal message about the subject. Here is what he had to say:

Multiple time frame analysis gives you an opportunity to “see inside” a larger trend to improve your timing which means less risk and better profit potential.

Using a minimum of three time frames (in this example for swing trades) we would;

 

1- IDENTIFY THE OPPORTUNITY on a long term chart (200 days or more) to identify the bigger trend.

2- ANTICIPATE ENTRY on the intermediate term trend (generally 30-50 days using 30 or 65 min time frame) to establish our risk/reward.

3- PARTICIPATE (BUY OR SELL SHORT) on a shorter term time frame (10-15 days using a 15 minute time frame or lower) to fine tune our entry.

 

Multiple time frame analysis can (and SHOULD) be used for any market participant, regardless of your time frame!

I think Brian explains this concept perfectly. He has a shorter time horizon that I do, but the psychology is exactly the same. We want to work from the top and work our way down. I like starting 10-20 years back, then working our way to what is happening the past couple of years and finally to the past few quarters. This helps identify the direction of the primary trend and can put us in a position to trade in the direction of the long-term, intermediate-term and short-term trends. Having all that wind behind us is a huge advantage. Without using multiple timeframes, it becomes impossible to do that.