Comments on: Evaluation of Momentum Oscillators http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/ A quantitative approach to profit in the US equity and futures markets, trading the markets like professional card counters are playing Blackjack or expert poker players are playing Poker. The key is to have the odds on your side and bet accordingly, knowing what, when, where, why and how much to bet on each trade or wager. Mon, 06 Feb 2012 13:40:20 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: ETF FOOL http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4210 ETF FOOL Sun, 07 Mar 2010 02:02:09 +0000 http://www.tradingtheodds.com/?p=32391#comment-4210 [...] Evaluation of momentum oscillators (TradingTheOdds) [...] [...] Evaluation of momentum oscillators (TradingTheOdds) [...]

]]> By: James http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4209 James Wed, 17 Feb 2010 01:01:33 +0000 http://www.tradingtheodds.com/?p=32391#comment-4209 Hi Frank, Got a question on your trading strategy I. For today 2/16, the 2 day RSI for SPY closed at 94.3. According to your trading criteria, for your extreme market condition 2.3: "the 2-day RSI did NOT close above 94, and the >indicator 1< did NOT close above xxx." Since today's RSI was above 94, doesn't this automatically make it a sell signal. Please explain, I'm confused. Thanks for your great work. James Hi Frank,
Got a question on your trading strategy I. For today 2/16, the 2 day RSI for SPY closed at 94.3. According to your trading criteria, for your extreme market condition 2.3: “the 2-day RSI did NOT close above 94, and the >indicator 1< did NOT close above xxx." Since today's RSI was above 94, doesn't this automatically make it a sell signal. Please explain, I'm confused. Thanks for your great work.

James

]]> By: TradingTheOdds http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4208 TradingTheOdds Tue, 16 Feb 2010 20:10:49 +0000 http://www.tradingtheodds.com/?p=32391#comment-4208 TopTick, this would've been a completey different approach. I was looking for the extrem ends in indicator values for the then current history since 01/01/1990, not for relatively high/low indicator values in the then current recent past. But it's an excellent idea and will probably subject to one of my next posting. In additin, I did not check for the robustness of those momentum indicators, means they were not evaluated for different underling (major market indices) and/or different time frames. best, Frank TopTick,

this would’ve been a completey different approach.

I was looking for the extrem ends in indicator values for the then current history since 01/01/1990, not for relatively high/low indicator values in the then current recent past. But it’s an excellent idea and will probably subject to one of my next posting.

In additin, I did not check for the robustness of those momentum indicators, means they were not evaluated for different underling (major market indices) and/or different time frames.

best,
Frank

]]> By: Toptick http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4207 Toptick Tue, 16 Feb 2010 20:02:33 +0000 http://www.tradingtheodds.com/?p=32391#comment-4207 Hi Frank, What was your motivation for ranking the indicators over the entire past history, versus, perhaps, a window of the trailing 252 days? Does it give better results, or were you thinking that more data is better than less? Hi Frank,

What was your motivation for ranking the indicators over the entire past history, versus, perhaps, a window of the trailing 252 days? Does it give better results, or were you thinking that more data is better than less?

]]> By: Toptick http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4206 Toptick Tue, 16 Feb 2010 14:28:21 +0000 http://www.tradingtheodds.com/?p=32391#comment-4206 Excellent work. Thanks! Excellent work. Thanks!

]]> By: CarlosR http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4205 CarlosR Mon, 15 Feb 2010 03:29:02 +0000 http://www.tradingtheodds.com/?p=32391#comment-4205 Frank, Very useful info, thanks much for doing this. Somewhat supporting the post made by the previous commenter, you might look at this: http://epchan.blogspot.com/2010/01/method-for-optimizing-parameters.html In that post, Ernie Chan references an article that supposedly shows that an RSI of 3 or 4 bars is significantly more profitable than RSI(2). So I do think that having the best parameter values for each indicator is important, and Chan's column discusses a method for doing that. On the other hand, coming up with those values may be so much work that it never gets done. I could understand that also! Frank,

Very useful info, thanks much for doing this.

Somewhat supporting the post made by the previous commenter, you might look at this:

http://epchan.blogspot.com/2010/01/method-for-optimizing-parameters.html

In that post, Ernie Chan references an article that supposedly shows that an RSI of 3 or 4 bars is significantly more profitable than RSI(2). So I do think that having the best parameter values for each indicator is important, and Chan’s column discusses a method for doing that. On the other hand, coming up with those values may be so much work that it never gets done. I could understand that also!

]]> By: Bob http://www.tradingtheodds.com/2010/02/evaluation-of-momentum-oscillators/#comment-4204 Bob Sun, 14 Feb 2010 19:00:46 +0000 http://www.tradingtheodds.com/?p=32391#comment-4204 interesting study. I think fairer comparison would be to optimize the parameters of each indicator in a cross-validation loop and rank the indicators based on their out-of-sample metrics. This way you will be comparing the ideal setting of each indicator rather then the arbitrary parameter settings and entry rules that you used here. interesting study. I think fairer comparison would be to optimize the parameters of each indicator in a cross-validation loop and rank the indicators based on their out-of-sample metrics. This way you will be comparing the ideal setting of each indicator rather then the arbitrary parameter settings and entry rules that you used here.

]]>