Remember when you were a child and you would be shown pictures and play “which one of these doesn’t belong?” I suppose it is an early introduction to pattern recognition and the examples weren’t very difficult.  Look at the picture below and ask yourself “which one doesn’t belong?”  It is not a trick question… the lone apple is clearly different than the 6 oranges in the picture.  So before I lose your attention I will explain what this has to do with trading.

As you may know, I like to look at multiple timeframes when I trade.  Multiple timeframe analysis allows us to better visualize and put into context the conflicting messages the market may be broadcasting.  Let’s say you started your analysis on a daily timeframe and you wanted to take a deeper look at the market action, many people will look at an “hourly chart” or one where each data point (candle or bar) is equal to 60 minutes of trading.  In my opinion this timeframe is a flawed timeframe to look at, especially if you add any indicators or oscillators to your analysis.

Here is why the 60 minute timeframe is flawed.  Each day the equities markets are open from 9:30AM to 4:00PM Est.  From the opening bell to the closing bell, the market is open for 390 minutes each day.  If we are to divide a 60 minutes into 390 minutes you would get 6.5 periods.  In other words, you would not have an equal amount of data in each bar or candle.

Another way to think about it is to recognize that the bars on an hourly chart are completed at the end of each hour, so the 9:30-10 timeframe would build one full candle on a sixty minute chart even though it only has 30 minutes of data…  On a chart with 60 minute candles there would be seven candles for each day: six of them would be constructed with 60 minutes of data and one would be constructed with just 30 minutes.  We would be comparing apples and oranges.

You may be thinking that it is not a big deal, but once you start adding basic technical tools, such as a moving average, you will recognize that the 30 minute timeframe having equal weight in the calculation of the average as the six- 60 minute periods doesn’t make sense.  As we know, technical analysis is more art than science, but little subtleties like keeping our data consistent can make a difference.  If we want a true “average” weight of a bunch of oranges, we cannot accurately determine it by averaging six oranges and one apple.

So here is the simple solution.  Either switch your chart to a 65 minute timeframe so you have six candles of equal length each day (390/65 = 6.0) or further reduce your timeframe down to 30 minutes of data.  With 30 minute data periods each day will have 13 individual candles of equal length.