System development and trading

R

RockMachine

This thread is about strategy development with the most liquid markets (ETF's and Stocks), for now. I have not posted here in a decade. Now in my 50's iv'e decided to go full "Grey-box" with a mix of uncorrelated strategies. The strategies outlined have been thouroughly backtested and forwrad tested to my comfort level. To date there are 10 strategies, uncorrelated, that have an average "Modified Sharpe" of 2.30. If you don't know what this means, please read Acrary's threads.....

The Legend....
https://www.elitetrader.com/et/threads/system-development-with-acrary.33654/

To add complexity, I have and will be trading options (Not tested) instead of the underlying (Tested). I chose options because they are
easier to take directional bets and require a tiny capital commitment. This gentleman had posted 2 informative threads.....

The Pioneer....
https://www.elitetrader.com/et/threads/swing-trading-with-butterflies.360566/

I don't plan to divulge the strategies as I am currently trading them. However, I can certainly help others that are interested in going down this path. The 10 strategy mix is mostly swing (1 - 5) holding periods. The others can be a daytrade and 1 that is long term that is hedged, that I will explain later. I will not give out my positions as doing so would be a disservice as I could be exiting some or all by the end of that day.

Picture below is of 9 strategies, #10 is discretionary with SPY 0dte that I cannot backtest. Notice how the independent "bumpy" equity curves come together as one. Oh yeah, the ONLY data points to trigger trades and exits are the daily open and close, more on that later.

independent 2024.png


combined.png
 
I chose the period of 2019- present to test my strategies using Amibroker. The period was sufficient enough for me with trades mostly exceeding 100. Some think that more data is better, some think that more weight/data should be inflationary years (ie 1970's). It could be argued that the 2008 to present day is a better gauge. One thing I noticed is that in 2000 crash, the Fed let anyone fail. In 2008 the govt. backstopped the big companies "Too big to fail" (also 1998 LTCM). So 2000 crash took 7 years to recover, 2008 took 6 years and more recently the 2020 covid crash took only 6 months to recover. The covid "V" recovery was the govt. backstopping everything. This explains my recency bias, but only time will tell.

vti.png
 
RockMachine said:
......""One thing I noticed is that in 2000 crash, the Fed let anyone fail. In 2008 the govt. backstopped the big companies "Too big to fail" (also 1998 LTCM). So 2000 crash took 7 years to recover, 2008 took 6 years and more recently the 2020 covid crash took only 6 months to recover. The covid "V" recovery was the govt. backstopping everything.""
More...

Interesting Summary... Thanks...
 
RockMachine said:
Current real time results of 10 strategies.

View attachment 343749
More...

Yes, everything presented on this thread is directional "Long Only" option trades. 1 strategy is longer term, trend based on "QQQ". Here, the strategy entered a long in January 2023. I manually entered 2 weeks ago with a call position. After a few days the call was 169% in profit, so I added and ATM put. I did this spread, strangle or protective put to where the math made it impossible to lose money and would make a considerable amount if the market crashed or went parabolic. The initial call delta was 0.5, and I am experimenting with ITM options where TV is less than 2% of premium.

qqq strategy.png
 
RockMachine said:
there are 10 strategies, uncorrelated, that have an average "Modified Sharpe" of 2.30
More...

Sharpe ratio of 2.3 sounds too good to be true (for annual Sharpe). For monthly, implying annual Sharpe ratio of about 8, it's a fantasy. And having 10 of those, uncorrelated, makes it even less believable.
 
You must log in or register to reply here.
Back
Top