Why the difference? Sector rotation

General
1

I tried to compare PV and the same strategy at P123.
It gives huge discrepancies in returns, while the only difference I see is a few different purchase dates, but all are concentrated towards the end of the month.

https://www.portfolio123.com/port_summary.jsp?portid=1709000

PV

Does anyone see why PV gives a result for the same period, with the same ticker, and the same seasonality of 11%, while P123 is around 8%?

image
image810×426 47.4 KB

image
image1126×304 20.4 KB

Here the performance over a even longer period is 15% https://web.archive.org/web/20220814145014/https://dualmomentum.net/2020/07/27/seasonality-factor/ (There is a difference because of the number of position in may-oct)

2

You have specified 25% weight and Sizing Method % Portfolio Weight.
The buy rules provide for 3 ETFs or 1 ETF.

So your model always holds cash, 25% with 3 ETFs and 75%with 1 ETF.

3

Yes, you are right. I didn’t see that, but is there any way of forcing the simulator to be fully invested in the 3 ETF`s nov-april and the 1 in may-oct?

I did some of the same test in Screen: https://www.portfolio123.com/app/screen/summary/264601?st=0&mt=9 But that did not see to yield anything near what I get in PV.

4

Yes. You must use Sizing Method: Formula Weight and put 1 in the formula box.

5

image
image1439×1005 128 KB

That gives you the same performance as PV.