In general, I don't like seasonality. To me, it sounds like investing based on mostly coincidental data, akin to buying when skirts are short, or staying invested after a positive January. That said, Ryan Detrick's regular glance at seasonality had a better track record for the past three years than most investors.
As to your algorithm, perhaps you'd consider trying out a quarterly approach? The pick-of-the-month approach feels a bit over-optimised to me.
Thanks, Martin! Yeah, that's probably a prudent take. I've not known about Ryan Detrick. I'll check his work. I'll also look into a quarterly approach to see if it can get a comparable performance or not.
Not a bug. You need to keep in mind that trades can happen only on market days. So for this strategy, PALL will be entered only from the market close of 12/2, not from 11/30 or 12/1. That's why you see the discrepancy in monthly return numbers. And what makes you think the MDD has never been over 30%?
As of the end of February, it is -3.71%, and as of the end of March, it is -6.16%. Because 2 months have accumulated.
And the current low in April is -22.64% lower than here. Then MDD is around -27.4%. This program is currently (if you change December to PALL) -42.2%.
Did I do my math wrong?
* I have been very interested in seasonality for several years, so I looked into various seasonality and momentum investments. I really respect Larry Williams. Haha.
So I am going to try this with some of my money. :)
That can be an interesting approach. It'll be pretty tedious with QuantMage, though, since it doesn't report median returns per month ^^; Unless there is a site or tool that conveniently reports them.
Intriguing! Thank you.
In general, I don't like seasonality. To me, it sounds like investing based on mostly coincidental data, akin to buying when skirts are short, or staying invested after a positive January. That said, Ryan Detrick's regular glance at seasonality had a better track record for the past three years than most investors.
As to your algorithm, perhaps you'd consider trying out a quarterly approach? The pick-of-the-month approach feels a bit over-optimised to me.
Thanks, Martin! Yeah, that's probably a prudent take. I've not known about Ryan Detrick. I'll check his work. I'll also look into a quarterly approach to see if it can get a comparable performance or not.
Detrick is quite active on the Xitter.
As of 2025.04.16, MDD is over -35%!
But it has never actually been over -30%! The program seems weird.
Am I not understanding it?
Yes, it's performing pretty bad lately. What do you mean by "seems weird"?
I switched to PALL in December. December 2024 is actually -7.45%, but the data from this program is -5.19%.
I don't think it's right at all.
The strategy doesn't have the PALL position...
I tried changing December to "PALL".
In reality, PALL in December 2024 is -7.45%, but in this program, it is -5.19%. Is this a bug in this program?
And look at the MDD graph carefully. As of April 16, 2025, MDD was over -35%! However, in reality, it has never been over -30%!
Is this also a bug?
Not a bug. You need to keep in mind that trades can happen only on market days. So for this strategy, PALL will be entered only from the market close of 12/2, not from 11/30 or 12/1. That's why you see the discrepancy in monthly return numbers. And what makes you think the MDD has never been over 30%?
As of the end of January 2025, MDD is 0%.
As of the end of February, it is -3.71%, and as of the end of March, it is -6.16%. Because 2 months have accumulated.
And the current low in April is -22.64% lower than here. Then MDD is around -27.4%. This program is currently (if you change December to PALL) -42.2%.
Did I do my math wrong?
* I have been very interested in seasonality for several years, so I looked into various seasonality and momentum investments. I really respect Larry Williams. Haha.
So I am going to try this with some of my money. :)
Drawdown means peak-to-current. You can see the latest drawdown started from early August last year in the underwater chart. Not from Feb 2025.
Maybe consider median return instead of average?
That can be an interesting approach. It'll be pretty tedious with QuantMage, though, since it doesn't report median returns per month ^^; Unless there is a site or tool that conveniently reports them.
Would you mind saying more about where the data is for which etf's performed best each month, the data used for your selection?
I just manually checked sector / country ETFs and other asset ETFs (like VIXM and bond ETFs) for their monthly returns in QuantMage :)