Home › Forums › Trading System Mentor Course Community › Trading System Brainstorming › A Question For Stretch Re-optimisers
- This topic is empty.
-
AuthorPosts
-
December 28, 2022 at 6:37 pm #102250TerryDunneParticipant
I was wondering how people are going to treat this year when doing their regular(?) re-optmising. If you have any trading systems that made a loss in 2022, will the re-optimisation favour extreme (as large as tested) stretches, so that there are fewer trades?
I do a similar process for portfolio allocation and have concerns that allocation will be too low for any systems that performed poorly in 2022.
December 29, 2022 at 7:58 am #115323JulianCohenParticipantThat’s an interesting question.
My initial thoughts are that if the drawdowns for each strategy are generally within expectations, then I will just treat it as any normal year. Good with the bad and all that
December 30, 2022 at 12:37 am #115325TerryDunneParticipantHi Julian,
I agree with your comments. However, if a system had a losing year and you re-optimise stretch, then the ‘best’ result will be the stretch that has the fewest trades to minimise the loss (i.e. the biggest stretch)?
Am I missing something?
December 30, 2022 at 2:53 am #115324LEONARDZIRParticipantTerry,
Interested in your comment about not allocating enough of your capital in 2023 to systems with poor results in 2022.
Do you have a system for allocating your capital to your various stock systems?
I am running multiple systems and allocate equal amounts of cash to each system. at the beginning of each year. The one exception is my mean reversion system which I trade for income and which has the smallest allocation. I am really not sure that is the effective portfolio method. Curious what you think or what other members are doing.December 30, 2022 at 4:40 am #115326JulianCohenParticipantQuote:No I agree with that. That’s where the conundrum comes in. It could be that by reoptimising you restrict the trades and that prevents further losing trades until the strategy starts to recover and then the next reoptimisation might improve things, or the opposite could happen.I don’t often reoptimise my strategies. I test on 2014- present and use whatever that may bring up.
My out of sample is 2003-2013 as that has two other major crashes in it, the GFC and the flash crash of 2011.
I tend to use whatever is best over all those periods and leave it after that.
For example, I am experimenting with volatility based position sizing. It can make a huge difference to the bottom line from 2003-2013, cutting the MDD in half, but from 2014 to present it actually increases the MDD a couple of %points and affects the CAGR….do I use it? Do I keep fiddling? conundrums
December 30, 2022 at 4:45 am #115327JulianCohenParticipantLen, for what it’s worth, this year, as I assume all trend following strategies will go to cash on the weekend, I will rebalance cash amongst my trend following strategies, but if they were all staying in stocks, then I probably wouldn’t.
I think being in cash is a good opportunity to rebalance, but if they were all in the middle of trends, then I would probably leave them to run their course before rebalancing.
December 30, 2022 at 10:39 am #115328TerryDunneParticipantLen, I use the results of my MCSs to allocate percentages of my portfolio. I use the average Ulcer Performance Index as my measure, but I reckon any number of other choices (e.g. MDD; Sharpe Ratio; CAGR/MDD etc) could be used without there being a lot of difference. I’m also interested in what others are doing in this regard, if people wish to share…
Julian, I guess if you re-optimise over eight years, 2022 will have some, but not too much, effect. It might be time to get on with doing it and see what the results look like compared to now. I don’t plan to re-allocate until the middle of 2023, so at least I have time.
December 30, 2022 at 2:21 pm #115329LEONARDZIRParticipantJulian,
Thanks for the reply. I decided to add a daily trend filter to my portfolio. So if the daily trend is down which is the current state of SPX I cut my allotment to my stock systems that are still long by 1/2. I haven’t backtested this idea but I note my current drawdown is just under 10% which I consider an accomplishment this year.December 30, 2022 at 4:16 pm #115334LEONARDZIRParticipantTerry,
Interesting approach. Thanks for letting us know. -
AuthorPosts
- You must be logged in to reply to this topic.