Mean Reversion Strategy Journal

Status
Not open for further replies.
Well that's a wrap for 2021, so far off to a good start, all 3 of my strategies started earlier this month on the Collective2 platform are in the green.

Happy New Year everybody I'm looking forward to getting stuck into 2022 and continuing to use my custom indicators to build a profitable track record that can easily be replicated. Cheers!

 
Last edited:
3 trades were closed today $C $SAIC and $VIAC for a total net profit of $67. There aren't many oversold stocks in the market right now so the portfolio is only at half capacity waiting in cash for opportunities to present themselves.

af8bb3_93ba1828de9b47429db28cf635ab20b0~mv2.png

af8bb3_4b88520ef46946ec8a920348d63e0891~mv2.png
 
Could you give a portfolio value as of Jan 1 2022, and then update that value as the year goes on? That way we can compare it to a buy and hold of SPY.
 
deaddog said:
Could you give a portfolio value as of Jan 1 2022, and then update that value as the year goes on? That way we can compare it to a buy and hold of SPY.
More...

Yes, you can view the live portfolio value and performance statistics at anytime with this link https://collective2.com/details/138464838 and can see it started the year at $51,648.

All the trades are sent automatically from my IB account into the Collective2 platform with no interference by me so it mirrors my live trading account. You can also have my trades automatically placed in your own trading account too via the Collective2 platform.
 
I commend you for having the strategy live on C2 which suggests you at least believe the signal works. If not for that, that I'd suspect this of being a scam... most signals services are scams, after all. The economics rarely make sense because people who actually have an edge can just trade or manage money.

I still doubt it works though. Unfortunately the backtests you show on the webpage are statistically meaningless by themselves. Indicators that fire infrequently like that will always be "profitable" on a lot of stocks just by random chance. Some of those profitable ones will also have really nice graphs just by chance. You need to either show that the indicator works on an entire universe like S&P 500 without cherrypicking, or be able to filter stocks using a walkforward learned function with a plausible form to guard against data snooping.

EDIT: do I understand correctly that the C2 strategy includes discretion on entries? If so, that would really hurt its ability to prove that the signal itself has edge.
 
Last edited:
cruisecontrol said:
I still doubt it works though. Unfortunately the backtests you show on the webpage are statistically meaningless by themselves. Indicators that fire infrequently like that will always be "profitable" on a lot of stocks just by random chance. Some of those profitable ones will also have really nice graphs just by chance. You need to either show that the indicator works on an entire universe like S&P 500 without cherrypicking, or be able to filter stocks using a walkforward learned function with a plausible form to guard against data snooping.

EDIT: do I understand correctly that the C2 strategy includes discretion on entries? If so, that would really hurt its ability to prove that the signal itself has edge.
More...

If you go back and read through all my initial posts in this thread you'll understand what I'm doing, I actually say not to follow any indicator blindly and I apply numerous filters and discretion on what trades I take. This is not a 100% systematic strategy as you're suggesting it should be. It's well known that long-only mean reversion strategies can work on stocks and I don't think the indicator has to work on every single stock unfiltered as you suggest for it to be part of a profitable strategy. It's an indicator to help find trades, not a black box ATM machine.
 
StockChartist said:
If you go back and read through all my initial posts in this thread you'll understand what I'm doing, I actually say not to follow any indicator blindly and I apply numerous filters and discretion on what trades I take. This is not a 100% systematic strategy as you're suggesting it should be. It's well known that long-only mean reversion strategies can work on stocks and I don't think the indicator has to work on every single stock unfiltered as you suggest for it to be part of a profitable strategy. It's an indicator to help find trades, not a black box ATM machine.
More...
What I'm saying is that you haven't actually presented any evidence that these signals have an edge, and I explained how you might go about doing that. If you can show that the signal works unfiltered that would be very good evidence. The more filtering that is used the weaker the evidence. (And if the signals only work on a certain set of stocks then that set should be part of the signal, otherwise how are people supposed to use it?) Throw in discretion and who knows what's going on anymore.
 
Status
Not open for further replies.
Back
Top