Tips on Technicals - Less is More
Many times, the data presented on a chart can get in the way of the analysis. The wiggles seen even on long-term charts obscure the trends we are trying to find so we need a way to cut through the clutter. Computerized filters using such complex concepts as fractals, wavelets and noise reduction can help, but only if you have the algorithms and understand how to apply them. Rather, let's use the tools at hand to simplify the process to "see the forest" rather than "getting lost in the trees."
Line Charts
Most technicians use either bar, candle or point and figure charts in their work. These types use open, high, low and close data (the former two) or tick data (the latter) so they are subject to counter-trend movements throughout the trading day. These movements can be knee-jerk reactions to news or market conditions rather than illustrative of the actual trend or technical pattern in force at the time. Using only a closing line chart, or even a low interval moving average of the closes, reduces the resolution of the chart but it also reduces the visual clutter. This makes trends and patterns easier to spot.1
Figure 1 shows the NASDAQ Composite Index from April 1995 to December 1996 in both bar and line formats. On the bar chart, the corrective pattern from August 1995 to January 1996 is a triangle. On the line chart, the pattern is a rectangle. For comparison, the bottom of the bar chart's pattern is drawn on the line chart. The difference is clear.
More important than the shapes of the patterns are the depths of the two retracements. The height of the rectangle when projected up an integral two times from the upside breakout forecast the top of the ensuing rally as well as later resistance in October 1996.
Of course, this is all hindsight. How would we have known to use a line chart for forecasting at the time? The answer is in the price action of October 10, 1995; the date of the big intraday spike down. When the market moves in such big swings, the likely cause is a panic-led overshoot of a technical target. A line chart ignores this condition and should have been employed immediately.
Trends and Waves
As mentioned above, reducing the noise on a chart can help in spotting trends. Elliott Wave analysis is a specific example of how this can be useful.
With wave analysis, we are attempting to spot the trends and counter trends in a market so that we can classify them as impulsive or corrective. Spike highs and lows as well as volatile trading patterns can hide these trends and their turning points. As seen in Figure 2, the rally in the Australian Dollar/ US Dollar exchange rate began in late 1993. The wave count from the start of the move is critical in determining if it was impulsive (in the direction of the trend) or corrective. The weekly bar chart makes this a difficult endeavor.
However, a weekly line chart with a 2-week interval removes enough of the market wiggles to allow us to see the wave structure clearly. The three waves highlighted are subdivided into a 5-3-5 structure. Without going into the details, this can only be part of one pattern -- a zigzag correction higher within a larger decline. Other Elliott rules need to be applied to create a forecast and the end result is not pertinent here. We can leave this topic by saying that the rising and falling trends were more readily seen with less data on the chart.
While this technique is a valuable tool, it should not be used exclusively. Not all spike highs and lows nor other wiggles should be ignored. The majority represent actual turning points in the markets. Taking a step back from the data can help in seeing patterns in otherwise noisy data. Once these patterns are found, all the data are needed to make trading decisions.
1 Moving averages are especially helpful for filtering out spikes in volume data. A three or five day average smoothes the data without shifting the result too far to the right.
