Profitable ETF trading strategies: a thought experiment with relative strength
Typically, traders will apply various technical analysis indicators to the behavior of price of an instrument in order to find moments of advantage. Many times these are combined in systems of ever-increasing complexity and apparent sophistication. All of these concepts however are based off of direct behavior of price. That’s not the only way that technical analysis can be used however.
Here is an example of using technical analysis in a nontraditional way to quickly and consistently assess the relationship between a broad index and one of its underlying components.
Begin by calculating the relative strength of an instrument against its index. Perform this calculation over a look back. That makes sense based on the timeframe you intend to hold the instrument. As an example, you may have the belief that swing trading large-cap stocks in the Dow 30 industrials offers an advantage to the agile trader who cannot afford to watch the market throughout the day.
You might take the Dow 30 industrials is a reasonable set of ultra-large-cap stocks which are not likely to go bankrupt overnight and which behave in a conservative fashion compared to some small-cap and overseas stocks based on the wealth of analysis and broad ownership by institutional money. It is rare for these stocks to be wildly mispriced and their stock performance is characterized by relatively smooth changes in price from day to day.
At any given time, one or more of these stocks will exhibit leadership qualities in that it will be outperforming the broad index and its peers for one to four weeks. At the same time there will be a set of these stocks that are underperforming for the same time.
There comes a moment when the individual leaders and laggards can no longer maintain their extreme behavior and they begin to revert to the mean of the index itself. By calculating the relative strength of each of the stocks on a daily basis and plotting that performance on a standardized scale of 0 to 100 for each stock, you may be able to find moments when the extreme condition begins to stabilize and reverse.
In the case of the laggards, this may alert you to a possibility to buy value just as it is beginning to revert to the mean.
In the case of the leaders, this may alert you to the possibility of the outperformance period coming to an end.
By applying these relative strength performance curves on a channel-based technical indicator like Williams %R, you might treat the upper 20 and lower 20 percentiles as the regions of extreme behavior and thus ready yourself for a reversion of the previous trend.
I have not subjected to this idea to rigorous back testing yet, but the concept is intuitively appealing. The insight comes from inductive reasoning based on years of observation and may be worthy of closer study.