New Relative Strength Backtest 'App' Added To Site. Check it out.

Apr 26, 2010 in Backtest | Relative Strength

We have added an innovative new relative strength backtest application to the site.

Click Here: ETF Relative Strength Backtest App



For a video tutorial on this 'app' click here: Relative Strength Tutorial"

Follow us on Twitter:


Comments (9) -

Apr 25, 2010 21:37 #

This is a very nice tool, and will only get better as these ETFs develop a longer trading history.  It would be great to eventually see slots for a few more ETFs, if it doesn't bog down your server.  Thanks!

Erik United States

Apr 27, 2010 00:14 #

Keep in mind that the 'ETF Screener' page has many more securities in it right now that can be viewed for performance and can be 'backed-up' through the calendar control in the upper right corner.  

The backtest function for portfolios will be offered at some point but we wanted to start with this to get a sense of how users like the interface and work out any details that users find difficult.

Stay tuned -- this site will get significantly better as projects we've been working on for a while get their final issues worked out.   The server is not a problem so long as you set some kind of reasonable limits in terms of combinations.  The server is quite powerful and this RS 'App' does not come close to testing the upper limit.


ETFreplay United States

Apr 28, 2010 00:35 #


I have been trying a similar approach using a set of highly liquid ETFs and tracking both relative strength and Sharpe/treynor ratios, swapping in and out of the top 10 as they rotate into and out of those positions.

Your approach is cleaner, much simpler to manage and implement in a trading account or 401k, and results in superior performance with reduced risk.

Two thoughts:  first, it is common wisdom that a diversified portfolio containing both stocks and bonds reduces risk - any way to assign proportions to each of the three ETFs used here (example, own 3 units of #1, 2 units of #2, and 1 unit of #3)?

Second, exactly how is volatility used in the overall calculation?  I am not sure if it is subtracted, divided by or what?  In essence, what is the equation you are using that determines overall relative strength?

Thanks for adding this capability - it is a real jewel.


Kevin_in_GA United States

Apr 28, 2010 03:59 #

Overall relative strength is computed as the weighted average of the RANKS of the 3 categories.  For this reason, cash will always be 'in the hunt' because it will always rank #1 for volatility.  Notice that this is the exact same thing we do with the ETF screener page for portfolios.   This is a miniature version of that page with backtesting.   We can do portfolios of ETFs in an application as well.   But this was a more simple way to do this --- and the backtest results are quite impressive.

We haven't thought about the proportion issue you brought up.   Good thought and let us get our current projects completed and then we will consider that.


ETFreplay United States

May 03, 2010 14:18 #

This is an incredible tool, guys.   Great work!

My wishlist is only to be able to add more than 2 ETF's to the rotation,  and to not be locked in to SHY as your safe haven.


Sbuck143 United States

May 04, 2010 05:35 #

Excellent!   One other item just occured to me.........the ability to change the day of the month you do your rebalancing.       Surely the results will be different (in some cases substantially different) depending on which day of the month the RS comparisons are made, especially if your look back timeframes are shorter.

Thanks for your work,  this has really opened my eyes to a new way of investing!

sbuck143 United States

May 14, 2010 01:51 #

I agree with Scott, this is really incredible tool!

CFD Trading United Kingdom

Sep 18, 2010 09:12 #

Great tool - two requests/thoughts:

1. some sort of Notification tool so that you can get that sneak peak at rebalancing

2.  Ability to save the parameters for each portfolio, so these are applied automatically (as part of 1) above?

Dean United States

Nov 28, 2010 14:14 #

I am also for "the ability to change the day of the month you do your rebalancing". that might give the backtesting more credibility and may show the days most favorable for rebalancing.

Furthermore I suggest an alternative semi-log plot of the total return graph. then a drawdown (of e.g. 10%) would show the same change everywhere on the graph.  

hugos Germany

Comments are closed