Wednesday, September 1, 2010

Baeyens, RSI

Most people who use technical indicators have little understanding of what they were intended to do, let alone how they might be “reinvented.” Sometimes it's wise to re-read the literature. In RSI: Logic, Signals & Time Frame Correlation (Traders Press, 2007) Walter J. Baeyens explores alternatives to Welles Wilder’s interpretation of his Relative Strength Index, introduced in New Concepts in Technical Trading Systems (1978). Baeyens’ primary source of inspiration is the method taught by Andrew Cardwell.

What caught my fancy early on in the book was the analogy between watching RSI fluctuations on a chart and observing a skater going up and down in a half-pipe. “The half-pipe is the RSI chart, and its extreme values of 0 and 100 are the edges where the skater seems to be suspended before accelerating down the slope in the opposite direction. The skater’s speed is highest in the middle of the half-pipe. As he zooms upward on either slope, his speed slows dramatically, then drops to zero as his course reverses.” (p. 20) Translated into the language of price movement, “a relatively small change in price can cause a big move in the RSI value when it travels around the 50-level. However, large price changes are required to move the RSI value a little, once it approaches the extremes, as if an ever-increasing momentum were required to push it up an increasingly steep slope.” (p. 21)

For those who prefer arithmetic to imagery Baeyens offers a table based on the up/down ratio required to move RSI to a given value. He uses the standard 14 lookback period. Up = (((average up last 13 days) x 13) + value up today)/14; down uses the same formula, simply substituting “down” for “up.” Here are some of the critical 50+ RSI levels and the up/down ratios: RSI = 90: up/down = 10/1; RSI = 80: up/down = 4/1; RSI = 75: up/down = 3/1; RSI = 66.6: up/down = 2/1; RSI = 50; up/down = 1/1. It doesn’t take a mathematical genius to calculate the up/down ratios when RSI is less than 50.

Baeyens sets out to debunk some of the myths surrounding RSI—for instance, that price/RSI divergence signals a possible trend reversal. Divergence, he writes, “occurs frequently, as it is inevitable, and does not warrant any special treatment. It is just one way for the RSI to escape its mathematical constraints.” (p. 27)

One of Baeyens’ theses is that it is imperative to correlate different RSI time frames such as weekly-daily-hourly or daily-hourly-15-minute “in order to assess the market conditions and anticipate high probability scenarios.” (p. 53) He then does fairly complicated channel analysis on these charts.

Baeyens offers the reader a range of scenarios, showing in each case how RSI helps anticipate price movement. The book, some 350 pages long, is full of marked-up charts. I should caution that this is not a book for the trader who wants to automate an RSI system. I have no idea how a programmer could translate the myriad channel lines into code. But it’s an interesting, fresh look at a widely used indicator.

1 comment:

  1. Thanks for the brief review. Would you recommend the book. Also have you implemented any of the things that Baeyens has written in the book