Alpha on the Edge
@alphaontheedge.bsky.social
Research and insights from the world of AI and algorithmic trading | www.alphaontheedge.com
So it turns out that the RSX is very promising! Props to Jurik Research!
We started with the basic RSI reversal strategy and went on to test other variants with SPY. The R2T Backtests of 8 strategy variants on SPY are posted here:
www.alphaontheedge.com/p/an-upgrade...
We started with the basic RSI reversal strategy and went on to test other variants with SPY. The R2T Backtests of 8 strategy variants on SPY are posted here:
www.alphaontheedge.com/p/an-upgrade...
An Upgraded Model?: Testing the RSI vs the RSX
Evaluating the claim "new and improved"
www.alphaontheedge.com
March 15, 2024 at 9:33 PM
So it turns out that the RSX is very promising! Props to Jurik Research!
We started with the basic RSI reversal strategy and went on to test other variants with SPY. The R2T Backtests of 8 strategy variants on SPY are posted here:
www.alphaontheedge.com/p/an-upgrade...
We started with the basic RSI reversal strategy and went on to test other variants with SPY. The R2T Backtests of 8 strategy variants on SPY are posted here:
www.alphaontheedge.com/p/an-upgrade...
I tend to be immediately skeptical of “improved” versions of time-honored strategies. So I wanted to put the RSX, which is Jurik Research’s version of the RSI, through the wringer. Finishing thorough backtesting and will post results on Monday.
March 9, 2024 at 8:41 PM
I tend to be immediately skeptical of “improved” versions of time-honored strategies. So I wanted to put the RSX, which is Jurik Research’s version of the RSI, through the wringer. Finishing thorough backtesting and will post results on Monday.
More on trend following with deep neural networks - we backtested models on FX pairs with good results!
Full post here:
www.alphaontheedge.com/p/trend-foll...
Full post here:
www.alphaontheedge.com/p/trend-foll...
March 7, 2024 at 5:05 PM
More on trend following with deep neural networks - we backtested models on FX pairs with good results!
Full post here:
www.alphaontheedge.com/p/trend-foll...
Full post here:
www.alphaontheedge.com/p/trend-foll...
Poking around for improvements on the RSI for $TSLA - using the RSX instead tends to results in a slightly higher in-sample win rate:
March 6, 2024 at 10:14 PM
Poking around for improvements on the RSI for $TSLA - using the RSX instead tends to results in a slightly higher in-sample win rate:
If you're trading $TSLA, we tested 25k configurations of a long/short RSI trading algo.
The strategy with the highest win rate used an RSI window of 46 bars, an upper th of 68 and lower th of 38.
Be careful - results vary wildly depending on config.
The strategy with the highest win rate used an RSI window of 46 bars, an upper th of 68 and lower th of 38.
Be careful - results vary wildly depending on config.
March 6, 2024 at 1:10 AM
If you're trading $TSLA, we tested 25k configurations of a long/short RSI trading algo.
The strategy with the highest win rate used an RSI window of 46 bars, an upper th of 68 and lower th of 38.
Be careful - results vary wildly depending on config.
The strategy with the highest win rate used an RSI window of 46 bars, an upper th of 68 and lower th of 38.
Be careful - results vary wildly depending on config.
Backtesting the entire S&P 500 with over 86 million simulations. The final word on this strategy.
www.alphaontheedge.com/p/backtestin...
www.alphaontheedge.com/p/backtestin...
March 1, 2024 at 4:22 AM
Backtesting the entire S&P 500 with over 86 million simulations. The final word on this strategy.
www.alphaontheedge.com/p/backtestin...
www.alphaontheedge.com/p/backtestin...
Which moving average crossover works for which FX pair? Partial backtesting results attached. More at www.alphaontheedge.com
February 27, 2024 at 8:46 PM
Which moving average crossover works for which FX pair? Partial backtesting results attached. More at www.alphaontheedge.com
I thought that an MA based on Pascal's triangle was just a gimmick, but historically, it has a higher median win rate than the SMA. Super interesting.
February 26, 2024 at 4:49 PM
I thought that an MA based on Pascal's triangle was just a gimmick, but historically, it has a higher median win rate than the SMA. Super interesting.
Really, really thorough backtests on 18 different MAs - let's see if any stand out!
www.alphaontheedge.com/p/lets-test-...
www.alphaontheedge.com/p/lets-test-...
Let's test every moving average crossover algorithm
The final word on the MA crossover?
www.alphaontheedge.com
February 26, 2024 at 4:43 PM
Really, really thorough backtests on 18 different MAs - let's see if any stand out!
www.alphaontheedge.com/p/lets-test-...
www.alphaontheedge.com/p/lets-test-...
We looked at using the SMA vs. using the EMA as a foundation for an MA crossover strategy. Depending on the FX pair, the benefits are mixed but generally not significant.
February 26, 2024 at 4:34 PM
We looked at using the SMA vs. using the EMA as a foundation for an MA crossover strategy. Depending on the FX pair, the benefits are mixed but generally not significant.