One of the problems associated with many technical indicators, such as moving averages, is the sudden shift in their value as a data point from the beginning of the series drops off. For instance, a moving average indicator may suddenly give a trade signal even though the market has remained inactive. This is due to the fact that a large value from x days ago dropped out of the average calculation as a new value was added. One indicator that remedies this situation is Wells Wilder’s Relative Strength Index or RSI. This indicator mitigates the impact of the data’s volatility by working with percentages.
The RSI indicator in this instance is considerably different from the traditional RSI indicator used in the equity markets. The RSI used in the equity markets compares the performance of a security to the performance of its peers in the same market segment. The RSI used in futures trading compares current price data to previous price data in the same time series. The RSI discussed here compares the magnitude of higher closes to lower closes over last x days in the following manner:
RS = (Sum of positive net changes over the last x days) / x
(Sum of negative net changes over the last x days) / x
RSI = 100 – (100 / 1 + RS)
As with many technical indicators, a longer analysis period will reduce the indicator’s sensitivity and vice versa. Traders generally use either a 9 or 14 day analysis period. The percentage value generated is generally considered to be overbought at 70 and oversold at 30. The caveat here is to not blindly sell an overbought or buy an oversold market. A strongly trending bull market will remain overbought for a considerable amount of time and money. It is best to look for the RSI cross back below the 70 level before selling and/or wait for the RSI turn lower as the market continues it’s rally, allowing you to capitalize on a bearish divergence. The 50% level will often act as potential support for a declining market and resistance for a rising market. Another technique for trading with RSI is to join the trend in trendline breakouts of the indicator itself.