A View on Technical Indicators and Trading Systemstechnical analysis writer show how or whether those technical stock indicators have predictive value over a broad range of stocks over a long period of time (relative to the typical trade horizon at least). That is not for lack of well known statistical techniques, but rather because (in our view) the results of such analyses would be less than impressive in a wide array of cases and, as we will show below, would contradict many of the precepts of technical analysis. Similarly, maxims like "the trend is your friend" suggest that stocks that are currently in an identifiable uptrend are more likely to continue that uptrend rather than reverse. Although there are straightforward ways to formalize and test such statements, that is rarely done, at least in the publications that make such claims.
In these pages, we describe some common technical indicators and their traditional interpretations. We describe a scoring system based on the interpretations of those traditional technical indicators, and also a scoring system based on statistical analysis using the same indicators. We demonstrate the predictive ability of each of the two systems, and how the two are related. These two scoring systems are highlighted in the Weekly Technical Status Report. Specifically, we show how well each does in predicting price changes 20 days forward for 100 stocks over 1,000 days of market data. We then present, test, and discuss 15 different trading systems. We draw several conclusions about technical indicators and trading systems from these results.
Contents: A View on Technical Indicators and Trading Systems
- Common Technical Indicators and Interpretations
- Moving Averages of Closing Prices
- Wilder's Relative Strength Index (RSI)
- Moving Average Convergence/Divergence (MACD)
- Two Technical Indicator Scoring Techniques
- Predictive Value of Technical Scores
- Simple Trading System Performance
- Moving Average Crossover Trading Systems
- MACD Trading Systems
- RSI Trading Systems
- Systems Using the Traditional and Statistical Scores