TradersEdgeSystems.com

Consulting | Custom Trading Systems
Built-In Indicators | Custom Indicators

October 2008 Stocks and Commodities Traders Tips


ARSI, The Asymmetrical RSI

Original article by Sylvain Vervoort
AIQ Code by Richard Denning
Traders Studio Code by Richard Denning

AIQ Version:

The AIQ code for Sylvain Vervoort’s article, “ARSI, The Asymmetrical RSI” is shown below. The author’s version of the indicator, which is a variation of Wilder’s Relative Strength Index or RSI, is illustrated using divergence of price with the ARSI indicator. A system to test the indicator was suggested but not specifically mentioned by the author. To test the indicator, I devised a simple trading system based on the author’s suggested divergence. The rules of the system are:
Enter a long position when:
1) Today’s low is less than the last pivot low of 10 bar strength and
2) The today’s ARSI value is greater than its value on the date of the price pivot low of 10 bar strength and
3) The ARSI has been less than 30 in the last two bars and
4) This is the first such signal in the last two bars

I ran an EDS test, which shows all signals, using the NASDAQ 100 list of stocks. The exit rules for long positions:
1) If the ARSI is greater than 70 or
2) If the bars in the position are greater than or equal to 3 and the ARSI is less than 50 or
3) If the bars in the position are greater than or equal to 10

The short sale rules are the inverse of the long rules and my testing was symmetrical.

I was curious as to whether the new indicator would outperform the original Wilder RSI. The author indicates that there should be more signals using the ARSI versus the RSI due to the modified indicator’s higher volatility. I used the same system and parameters as outlined above but substituted the RSI where the ARSI indicator was used. I ran comparative tests over the period 10/15/2002 to 7/1/2008 which was a mostly bullish period. On the long only test-comparison, shown in Figure 1, the ARSI system is shown in the left two columns compared to the RSI system in the right two columns. There are almost 6 times more signals on the ARSI system, confirming the author’s view of the modified indicator. Also the performance of the modified indicator is better than the original version with the average profit per trade improving by 2.5 times and the reward to risk ratio improving by 1.2 times. The short side only test-comparison was run over a mostly bearish time period of 3/1/2000 to 10/15/2002, shown in Figure 2. The other metrics for the ARSI on the short side test are worse than the RSI but the RSI had only 17 trades in the test and this sample is not large enough to be reliable. A trading system could more easily be built around the ARSI indicator with divergences, than the RSI indicator.

EDS Code for ARSI, the Asymmetrical RSI
ARSI Sys1.EDS



Traders Studio Version:

The TradersStudio code for Sylvain Vervoort’s article, “ARSI, The Asymmetrical RSI” is shown below. The author’s version of the indicator, which is a variation of Wilder’s Relative Strength Index (RSI), is shown by the author to be an improvement over the RSI when used to spot divergences. A system using divergence to test the indicator was suggested but not specifically mentioned by the author. To test the indicator, I devised a simple trading system based on the author’s suggested divergence. The rules of the system are:
Enter a long position when:
1) Today’s low is less than the last pivot low of 20 bar strength and
2) The today’s ARSI value is greater than its value on the date of the price pivot low of 20 bar strength and
3) The ARSI has been less than 30 in the last two bars and
4) This is the first such signal in the last two bars
5) Enter on a stop next bar at high of the set up bar
The exit rules for long positions:
1) If the ARSI is greater than 70 or
2) If a closing basis loss of 3 or more average true ranges (ATRs) occurs
3) Exit at market at next day’s opening price

The short sale rules are the inverse of the long rules and except the exit parameter for the short sales on the ARSI and RSI was set to 40 rather than the 30 that would be used if symmetrical testing had been used.

I ran both session level tests and trade plan level tests using three full-size futures contracts on the stock indexes, Russell 2000 (RL), S&P Midcap 400 (MD) and the NASDAQ 100 (ND). The trade plan used was one of the plans that ship with the product called, TS_PERCENTRISK, using 6% of the account balance to risk on each trade with starting capital of $1,000,000. The code for this trade plan is not shown because it comes with the software. The risk was set at 3 ATRs, the same as the stop loss level.

In order to determine whether the new indicator would outperform the original Wilder RSI, I used the same system and parameters, portfolio and trade plan as outlined above but substituted the RSI where the ARSI indicator was used. I ran comparative tests over the period 02/13/1992 to 8/12/2008. A summary of the key metrics that resulted from running the trade plan level tests on both systems using the 6% risk per trade is shown in Table 1. The ARSI system is shown in the left column compared to the RSI system in the right column. The author indicates that there should be more signals using the ARSI versus the RSI due to the modified indicator’s higher volatility and this was verified by my tests. There are almost 7 times more signals on the ARSI system. In addition almost all of the key metrics were significantly better using the ARSI over the RSI. A divergence trading system could more easily be built around the ARSI indicator, than with the RSI indicator as there are more and better signals using the ARSI.

Traders Studio code for ARSI, the Asymmetrical RSI:
ARSI Test System.zip

 

 

 

 

 

 

 

 

 

 

 

 

 

Disclaimer | Privacy Policy | Contact Us | Spam Blocker | ©2009 Traders Edge Systems