Changes to Support and Resistance algorithms

Published 20 September 2019

Investtech’s research into Support and Resistance conducted in the autumn of 2019 has revealed important findings. The research results are consistent with technical analysis theory in some cases, while in others the results indicate that the theory is weak or even wrong. We have found it appropriate to change the algorithms used in Investtech’s automatic technical analyses.

The three main conclusions from this research are:

Conclusion 1:
Stocks that have had a negative reaction and are testing support (buy signal) or that have had a positive reaction and are testing resistance (sell signal) overall developed in line with benchmark in the following period. These signals give little information about future price development and can be dropped from a technical analysis of a stock without any significant loss. These research results call into question an important element of classic technical analysis theory.

Conclusion 2:
Stocks that have broken upwards through resistance (buy signal), have continued to rise, and they have risen more than general market development. Stocks that have broken downwards through support (sell signal) have continued to underperform. These signals can be valuable in a decision of whether to buy or sell a stock. The period after a break upwards through resistance will statistically be a good time to buy. This aligns with classic technical analysis theory.

Conclusion 3:
Stocks that are far above support and without any resistance, have risen well in the coming period. Similarly, stocks far below resistance and without any support, have underperformed. Despite these being considered neutral situations in theory, the research results show that they are statistically very strong signals.

The results appear consistent, strong and robust. Starting from closing price on September 19, 2019, we have implemented these changes to our automatic analyses:

• Reaction back and test of support is considered neutral. This was previously a positive signal.
• Reaction up and test of resistance is considered neutral. This was previously a negative signal.
• Price far above support and lack of resistance is considered a positive signal. This was previously a neutral condition.
• Price far below resistance and lack of support is considered a negative signal. This was previously considered a neutral condition.
• Price between support and resistance, while not very close to either, is considered a neutral condition. This was previously a positive signal if the price was much closer to support than resistance, and a negative signal if the price was much closer to resistance than support.

In the following cases, the Support and Resistance algorithms have not been changed:

• Break upwards through resistance is considered a positive signal.
• Break downwards through support is considered a negative signal.