Blog - Investing Notes

April 11, 2005 – Technical Trading Thoroughly Tested

In their March 2005 paper entitled "Re-Examining the Profitability of Technical Analysis with White’s Reality Check and Hansen’s SPA Test", Po-Hsuan Hsu and Chung-Ming Kuan examine the profitability of a very large universe of technical trading rules and strategies against the Dow Jones Industrial Average (DJIA), S&P 500, NASDAQ Composite and Russell 2000 stock indices. Their approach and findings, using data from 1989-2002, are as follows:

The authors consider a universe of 39,832 simple trading rules, simple contrarian rules and investor strategies, defined as follows:

• 18,326 simple rules in 12 classes: Filter Rules (FR), Moving Averages (MA), Support and Resistance (SR), Channel Break-Outs (CB), On Balance Volume Averages (OBV), Momentum Strategies in Price (MSP), Momentum Strategies in Volume (MSV), Head and Shoulders, Triangle, Rectangle, Double Tops and Bottoms and Broadening Tops and Bottoms.
• 18,326 simple contrarian rules that are the opposite positions of the preceding simple rules.
• 3,180 investor strategies based on the information generated from many simple rules in three classes: learning strategies (LS - following the best-performing rule within a rule class), voting strategies (across the trading rules in a rule class) and position changeable strategies (allowing weighted positions).

In applying all these rules and strategies against the four major stock indices identified above, the authors conclude that:

• The best simple rules are:
• For the DJIA, a momentum strategy in volume rule (NOT SIGNIFICANT)
• For the S&P 500, a contrarian on balance volume rule (NOT SIGNIFICANT)
• For the NASDAQ, a two-day moving average rule (HIGHLY SIGNIFICANT)
• For the Russell 2000, a two-day moving average rule (HIGHLY SIGNIFICANT)
• Most profitable rules and strategies for the NASDAQ and Russell 2000 are constructed from filter rules and moving averages rules. (See the table below from the paper.)
• No contrarian rule is significantly profitable.
• All of the profitable investor strategies are learning strategies, and there are many more profitable investor strategies than there are profitable simple rules. In fact, there are profitable investor strategies that are based on non-profitable simple rules. Technical investors are apparently able to extract and enhance information from simple rules.
• After accounting for transaction costs, the best rules and strategies generally beat a buy-and-hold strategy for both in-sample and out-of-sample periods.

In summary, technical analysis can be significantly profitable when applied to relatively immature stock indices (NASDAQ Composite and Russell 2000) but not when applied to mature stock indices (DJIA and S&P 500). Speed of information dissemination and liquidity for stocks within an index may be critical to the success of technical analysis on that index.

See Blog Synthesis: Some Trading Indicators for analyses of the usefulness of other technical indicators.