Good luck first of all - keep this thread going
Same position size for all 23 trades ?
Across the 23 - what’s your average holding period and how many do you put on concurrently ?
System development and trading
- Thread starter RockMachine
- Start date
Same position size for all 23 trades ?
Across the 23 - what’s your average holding period and how many do you put on concurrently ?
Thanks for your support.
Position size: When I find a strategy that is uncorrelated, I look at the monthly avg max and min pnl and determine how much it fluctuates from the existing portfolio. So if existing strategies A-E returns $100 (1 unit) per mo avg and new strategy F returns $200 (1 unit), I will compare mod sharpe under each and adjust. So in this case strategy F I will test at ( 0.5 unit).
This is not optimized, as the combination of 10 strategies under multiple unit sizes would be in the millions. I do recall Acrary built his own software to optimize unit sizes. Seeing how I am trading an edge (with underlyings tested) but with options (not tested), any optimization of units would be misleading. However, in realtime with options I can collect live data and adjust after an adequate sample.
Avg hold time: It fluctuates from same day (directional) to 70 days (hedged).
To avoid clustering of positions I am relying on trading different sectors (tech, energy, foreign etc.) that have historical uncorrelated returns, and have different holding periods. I know first hand that this works until it doesn't. In 2008 I had a commodity portfolio that was diversified in a trend model I developed. I was an equities director/MM with an Intl. banks and was restricted from trading equities in pa. Barrons cover came out about a commodities bubble and wham, all of a sudden my pre planned diversified portfolio had turned to a correlation of 1, ie risk off. This will happen again, and the only way I know to cushion the inevitable is to keep bet sizes small (options) and hedge lengthy exposure (options).
Last edited:
I have a bunch of strategies in the "Bullpen" file and could trade 5 or so uncorrelated SPY strategies, mean reversion, trend, daytrade, swing, long only, short only etc. but there would be periods where these clusters would occur. For now my plan would be to get smaller if clustering occurs. Defense wins championships.
Last edited:
Just saw this headline today. Generally I don't believe anything that the Vampire Squid says to the public, however I am now seeing my trend based strategy QQQ (Long) now outperforming on Put side (hedge) vs. the current Call position.
I haven't been active here as my goal is not to win a popularity/click's contest. Hopefully some have interest in starting to developed/trade strategies that have a proven track record. I will update soon the stats that I posted earlier.
I haven't been active here as my goal is not to win a popularity/click's contest. Hopefully some have interest in starting to developed/trade strategies that have a proven track record. I will update soon the stats that I posted earlier.
Midweek update.. I added new column to reflect "Option Payoff" vs. what was tested in the underlyings.
End of week update.
I added a few strategies for a total of 12 as of this writing. Trades for the week were in the following, AAPL,TSLA,SPY,FXI,GDX,XLE,QQQ. All are closed out except for 2. QQQ is the longer term trend strategy long via calls and added puts to capitalize if I'm wrong.
As a reminder, Yellow and red arrows are tested entries and exits long/short. Since I am late to enter "Green arrow", my long bias has reversed to green entry. So I gave up 38% of PnL but remain profitable thanks to the Put side of this position. Overall it was nice to see that the strategies did survive the largest down week of the year.
I added a few strategies for a total of 12 as of this writing. Trades for the week were in the following, AAPL,TSLA,SPY,FXI,GDX,XLE,QQQ. All are closed out except for 2. QQQ is the longer term trend strategy long via calls and added puts to capitalize if I'm wrong.
As a reminder, Yellow and red arrows are tested entries and exits long/short. Since I am late to enter "Green arrow", my long bias has reversed to green entry. So I gave up 38% of PnL but remain profitable thanks to the Put side of this position. Overall it was nice to see that the strategies did survive the largest down week of the year.
How do you choose which symbols to trade?
I keep a watch list of liquid etf's in different sectors. If sector doesn't have a liquid etf then I will look a individual stocks. I do this with a general guess that they aren't too correlated. Next I will take a trend strategy with limited variable/s and see how it performed. This accomplishes 2 things. First it determines if strategy works in other markets and second, if their is an exploitable edge in other market. Lastly if 1 and 2 checkout, I will see if it's correlated to what I am already trading. If yes, then I will add to bullpen and monitor it's performance every so often.
Sub sectors are also considered depending on what is moving. I wouldn't consider trading USO, XLE, CVX, XOM at this point in time. However, I do have strategies that will trade QQQ, TSLA, NVDA and AAPL.
Sub sectors are also considered depending on what is moving. I wouldn't consider trading USO, XLE, CVX, XOM at this point in time. However, I do have strategies that will trade QQQ, TSLA, NVDA and AAPL.
Midweek update. Putting this thread on hold due to lack of interest, I will report back from time to time. GLTA
End of week update.
Been having good success with a 0dte strategy that I developed. I haven't modeled it as it's purely discretion on a single setup that I have been noticing. Admittedly, 0dte is not for the faint at heart but the R/R is mind blowing. Swing trading is great when you wake up with a gap move in your favor, but it cuts both ways. There are hundreds of video's on youtube about daytrading where you exit trades at 2/1 RR, but that's not my style. In the past week iv'e averaged 6/1 RR. Yesterday I had my best for this at 7.7/1 RR.
Been having good success with a 0dte strategy that I developed. I haven't modeled it as it's purely discretion on a single setup that I have been noticing. Admittedly, 0dte is not for the faint at heart but the R/R is mind blowing. Swing trading is great when you wake up with a gap move in your favor, but it cuts both ways. There are hundreds of video's on youtube about daytrading where you exit trades at 2/1 RR, but that's not my style. In the past week iv'e averaged 6/1 RR. Yesterday I had my best for this at 7.7/1 RR.