Uncovering the Truth About RSI’s Effectiveness in Modern Markets

[@SzalonyTygrys69]

Just by looking at this strategy I know that your secret of success is in making asymetry that help you win more than lose ;-) When you have this kind of algo it doesn't matter if you use rsi, macd or even throw the dice.

[@AlgoTradingWithKevinDavey]

Thanks for the comment! I don't look at win %, by itself it is mostly meaningless. I look more at expectancy.

[@SzalonyTygrys69]

@@AlgoTradingWithKevinDavey I didn't mean win %, it is just history. Every trade is like throwing a dice. I mean pyramiding, diversification, holding profits/cutting losses. It all combined makes you a cassino just like Mark Douglas wrote in his books or Jessy Livermore memories. Nevertheless thanks for sharing your experience. Maybe you can make a video how do you estimate correlation between assets. It is about time correlation or behaviour?

[@SzalonyTygrys69]

@@AlgoTradingWithKevinDavey Oh, I found extensive article at your blog. If covers my question, thank you for your work.

[@AlgoTradingWithKevinDavey]

@@SzalonyTygrys69 Glad it helped!

[@martyfrancom7148]

My experience tells me that Gap Up and Gap Down will often turn even the best strategy into a looser. How do you deal with Pre and Post Market Gap Up/Downs?

[@shaunlawton2227]

Great! I like the RSI and use it a lot, but haven’t used it with a MA before. I enjoy these, please keep going.

[@AlgoTradingWithKevinDavey]

Thanks, will do!

[@jakesvandermescht694]

Great vid Kevin! :) Thank you. I read some comments below, and I think like Kevin says, test, test, test. Once you find a robust algo on a specific asset, going through Kevin's motions of testing, it still has a chance of failing. Much much less probability, you given yourself a very good chance of success, but... market conditions can change, and you might find yourself in an extended drawdown period, something this asset never had over your tests, extended ranges. All that said, my best performance algos had some kind of RSI code in it. Like Kevin said, there are good fundementals of trading in RSI formula. Good luck and my best advice is, follow and listen carefully to Kevin's advice! :) Good work Kevin.

[@AlgoTradingWithKevinDavey]

Thanks for the comment! Glad you have had success with RSI!!

[@lowcorrelation]

Hello Davey. I have a question that has been bugging me for a few days now. My entry is RSI based. I have been testing some of your exits from your entries and exit confessions book (love the book btw) Have has some promising results from the exits you present in your book. I cant remember for the life of me who said that having the same entry but different exits can be considered to be a portfolio. I cant find it in my notes or any of my books. You kind of touched on it towards the end of entries and exit in the Some Odds and Ends, where you talked about pairing different exits. Maybe I heard it from one of your podcast episode from better system trader. or I could be dead wrong, haha. What do you think? Can you have a portfolio of systems that all have the same entries but different exits? Part of me worries about correlation. I will be adding a stop and reverse exits to my list of exits so thanks for that! ..so many entries and exits, so little time!

[@AlgoTradingWithKevinDavey]

Different exits for the same entry will create different strategies, but their performance will likely be corrleated, since they use the same entry signals. I'd be careful about too much correlation in these cases.

[@covingtoncreek]

Hi I'm not sure if you've taken the Strategy Factory but one of the things Kevin teaches is how to determine if strategies in a portfolio are correlated. There are several tools provided for this. Highly recommended. Just like you mentioned, if you have two fantastic strategies, but they are highly correlated, then you've just doubled your risk.

[@lowcorrelation]

@@covingtoncreek Hi. I have not taken Kevins Strategy Factory yet. I also use the Pearson correlation formula to determine my systems correlation in my portfolio. > .1 is what im shooting for right now. Strategy Factory has always been my plan B, if things do not work out going solo. Thanks for the recommendation!

[@springbok68]

Thanks Kevin. I find your videos very useful. Keep up!

[@AlgoTradingWithKevinDavey]

Glad you like them! I appreciate the comment!!

[@jasondesmul]

Another great contribution, thank you!

[@AlgoTradingWithKevinDavey]

Glad you enjoyed it!

[@ruthgresham949]

question with regards to optimization / curve fitting: if only a fixed percentage stoploss and fixed percentage TP are optimized, and no other parameters, would you consider that over-optimizing or is that acceptable practice?

[@AlgoTradingWithKevinDavey]

That is good to only have 2 inputs optimized, but it also depends on the number of trades taken, and also how many tests you run (how you treat your out of sample data). For example, if you only have 4 trades over your test period, then 2 inputs could easily be used to curve fit.

[@JohnCatoUK]

like your comment, and I am paraphrasing creatively, "ruthlessly throw out the ones that don't work and trade with the best of the best that do"

[@AlgoTradingWithKevinDavey]

More or less, IF you've developed the strategy properly. Keeping good performers built with a faulty development process is NOT good, for example.

[@stephenkelly6168]

Yes, please make more!

[@AlgoTradingWithKevinDavey]

Thanks for subscribing, I will try!

[@philchapu2421]

Love the videos , "please make more!"

[@AlgoTradingWithKevinDavey]

I will try to, glad it helps you!

[@covingtoncreek]

Kevin I like these slightly deeper dives into the Confessions book. It gives me better insight. It's also interesting to see you carry the ideas just a bit further than the ideas in the book. I'm curious if you've experimented with RSI2? I'm sure you've looked at it but do you trade any strategies that incorporate use of RSI2? Seems wildly popular these days.

[@lowcorrelation]

If you havent already, check out the book, Short Term Trading Strategies That Work by Larry Connors and Cesar Alvarez. Larry Connors is the that developed it and Alvarez goes deep on the strat. In my testing, the RSI2 is by far the best entry I have come across and pretty sure Kevin portfolio has the RSI2 in the top 3...maybe 5 performers, lol!

[@AlgoTradingWithKevinDavey]

Thanks for sharing!

[@rickliao1975]

Please make more something like this video!

[@AlgoTradingWithKevinDavey]

Hopefully it helped you!!!!!

[@deveshdembla5252]

Hi, I tried out this system for a few products and the timeframes shown in the video. It results in a loss for me, I'm probably doing something wrong / missing something but shouldn't the number of trades be very less? Shouldn't be a lot of cases when the RSI shows an oversold market and the close is above a certain moving average right?

[@deveshdembla5252]

Or are we not checking for an oversold condition but rather the threshold range between 50-80 as mentioned in the vid

[@AlgoTradingWithKevinDavey]

I'm not sure what you may be doing correctly or incorrectly. The code is in the video though.

[@martyfrancom7148]

You didn't mention how this strategy determines when to exit. What are your take profit lines of code? Seams to me to be just half of a strategy. Also, does this strategy deal with Pre and post Regular Market time? Why not post a complete strategy?

[@AlgoTradingWithKevinDavey]

This is a complete strategy as is. It is a stop and reverse system, always in the market, either long or short. No stop loss or profit target. It is a complete strategy by itself. Not sure what you are asking regarding "pre and post market" - this strategy runs on futures markets 23 hours a day.

[@florianpetre151]

Please make more videos like this

[@AlgoTradingWithKevinDavey]

Florian, I will definitely try. Glad you like the RSi video!

[@HateDietPepsi]

This strategy seems counterintuitive. Seems like it would be better if the RSI used close[1] or even a few bars recent occurrence.

[@AlgoTradingWithKevinDavey]

Your ideas are certainly worth trying, I can't say this approach is any grand theory I have on the markets. I just test a lot of different things...

[@Jack-uc5ly]

You show a graph of BRN 725, so I'm assuming you didn't do a walk forward test using time frame as a parameter, why not?

[@AlgoTradingWithKevinDavey]

I assume you mean changing the bar size during the historical test, such as using 725 in 2019, 365 in 2020, 453 in 2021, etc. very tough to do walkforward testing like that.

[@Jack-uc5ly]

@@AlgoTradingWithKevinDavey do you mind sharing any information on why? Or any articles books etc, I have one of yours. I read there's a few ways to do walk forward testing, unless I misunderstood it. As far as I am aware, you'll train and optimise parameters for the past 3 years (for example), and then use those for the next year, and keep going, almost as if you're using this strategy at that point in time. So wouldn't it just be an extra variable in the optimisation/walk forward? Or is this just the case with the software you're using?

[@AlgoTradingWithKevinDavey]

@@Jack-uc5ly Software issue. It could be done, just not very easily with most testing software.