Developed by Donald Lambert, the Commodity Channel Index (CCI) is a momentum oscillator that can be used to identify a new trend or warn of extreme conditions. This strategy uses weekly CCI to dictate the trading bias when it surges above +100 or plunges below -100, which are key levels noted by Lambert. Once the trading bias is set, daily CCI is used to generate trading signals when it reaches its extremes.
Lambert's trading guidelines for the CCI focused on movements above +100 and below −100 to generate buy and sell signals. Because 70 to 80 percent of the CCI values are between +100 and −100, a buy or sell signal will be in force only 20 to 30 percent of the time. When CCI moves above +100, a security is considered to be entering into a strong uptrend and a buy signal is given. The position should be closed when CCI moves back below +100. When CCI moves below −100, the security is considered to be in a strong downtrend and a sell signal is given. The position should be closed when CCI moves back above −100.
By requiring an exit on a move back below +100 or a move back above -100, Lambert's original trading strategy produced lots of relatively short signals. This CCI Correction strategy offers a tweak to Lambert's original strategy, but maintains his general trading guidelines, which rely on a surge above +100 or plunge below -100. There are three steps.
1. Define the bigger trend and trading bias. A CCI surge above +100 on the weekly chart indicates that an uptrend is emerging and a bullish trading bias is adopted. This bullish bias remains until there is a surge below -100. A CCI surge below -100 on the weekly chart indicates that a downtrend is emerging and a bearish trading bias is adopted. This bearish bias remains until proven otherwise with a surge above +100.
2. Wait for a smaller countertrend movement. Use the daily chart to look for overbought pullbacks when the weekly chart dictates a bullish trading bias. A CCI plunge below -100 reflects a pullback within a bigger uptrend. Look for oversold bounces when the trading bias is bearish. A CCI surge above +100 on the daily chart indicates a bounce within a bigger downtrend.
3. Look for a reversal of this countertrend movement. When the trading bias is bullish and daily CCI moves below -100, a surge back into positive territory signals a reversal of the pullback. This indicates that the bigger uptrend is also resuming. When the trading bias is bearish and daily CCI moves above +100, a plunge into negative territory signals a reversal of the bounce. This indicates that the bigger downtrend is resuming.
The general idea is to trade in the direction of the bigger trend. Chartists must shorten the timeframe to look for signals based on the shorter trend. In theory, any combination of timeframes can be used. For example, daily charts could be used to identify the bigger trend and dictate the trading bias. Thirty-minute charts could be then used to follow the shorter trend and generate trading signals. Initiating positions after a correction improves the risk-reward ratio.
The first chart in this article shows the S&P 500 ETF (SPY) with 26-week CCI. A 26-week timeframe was chosen because it represents six months, which is a pretty good yardstick for a medium- or long-term trend. The yellow areas show when 26-week CCI was in bull mode, which means the most recent signal was a surge above +100. The white areas show when 26-week CCI was in bear mode, which means the most recent signal was a plunge below -100. The next three charts show daily bars and 26-day CCI to generate signals for 2008, 2009 and 2010. Let's start with 2008.
Weekly CCI moved to bear mode in November 2007 (blue line). This means we approach the daily chart with a bearish bias and only look for bearish signals. Bullish signals are ignored because the bigger trend is down. Keep in mind that a bearish signal is a surge above +100 and then a move below the zero line. The red dotted lines show five signals, one in late 2007 and four in 2008. The sell signal in late February did not work out that well, but the others aligned with SPY peaks quite well.
The chart above shows weekly CCI changing over from bear mode to bull mode at the beginning of May 2009. Prior to this changeover, daily CCI produced another good sell signal in early January. After the May changeover, CCI produced a buy signal in mid-July, dipping below -100 and then surging above the zero line. This proved to be quite a timely signal. There were two near-signals as CCI dipped to -97 in late June and early November.
2010 started with a bullish bias, switched to a bearish bias in late June and then back to a bullish bias in early October. This means there were three different trading periods: bullish signals were considered from January to mid-June, bearish signals were considered from mid-June to early October and bullish signals were considered from early October until year-end. It was a violent year. The first bullish signal in mid-February foreshadowed a nice advance. The bullish signal in mid-June would have been a loser after the sharp decline below 100. There was some followthrough after the bearish signal in August, but a good stop-loss was needed to either lock in profits or prevent a loss.
Using two timeframes, such as weekly charts for trading bias and daily charts for signals can be cumbersome. Chartists can emulate weekly CCI on the daily chart by using a longer look back period. The example below shows a daily chart for the S&P 500 ETF from February 2010 to February 2012, two years. Instead of a 20-week CCI on a weekly chart, this chart shows 100-day CCI to dictate the trading bias. A bullish bias is in force after a surge above +100 (yellow areas) and a bearish bias is in force after a plunge below -100 (white areas).
20-day CCI is used to generate trading signals in harmony with the trading bias dictated by 100-day CCI. There were bearish signals in late June and early August (red dotted line). A bullish bias took hold when 100-day CCI surged above 100 in mid-September 2010. A strong uptrend subsequently took hold, with 20-day CCI not dipping below -100 until March 2011, six months later. There was a near-signal in November as CCI reached -94 before turning back up. This is probably a time when some personal judgment is required.
The CCI Correction strategy offers traders the best of both worlds: trading with the trend and initiating positions after a corrective phase. As the examples show, finding the perfect setting would be pretty much impossible. Moreover, chartists should avoid curve-fitting with designing a trading strategy. Also, note that the CCI Correction strategy is not meant as a stand-alone system. Chartists need to consider how to implement stop-losses, when to take profits and how to tailor the strategy to their own goals and trading style. More aggressive traders could prefer a shorter look-back period to generate quicker signals, while less aggressive traders might prefer a CCI surge above +100 to generate signals on the short timeframe. Keep in mind that this article is designed as a starting point for trading system development. Use these ideas to augment your trading style, risk-reward preferences and personal judgments. Click here for a chart of the S&P 500 with the CCI indicators.