Just curious Karl/Gary/David?
SPXL is a sell currently in Income system. Since 10-10-25.
UPRO is an Open trade in the same system. Since 26-11-25.
Given both products are essentially the same 3X geared ETFs on the S&P 500.
UPRO Has a slightly lower expense ratio & has a higher daily volume but in total return over 10yrs they are basically the same result.
Just curious how one can be an Open trade & one a sell closed trade in the Income system considering they are both doing the same thing.
SSO is 2X geared S&P 500 it’s a sell closed trade.
SPUU is a 2X geared S&P 500 it’s also a closed trade.
With 4 S&P 500 ETFs in the Income system only 1 is an Open trade UPRO apart from the daily volume IM curious why that one is the only Open trade on the system out of 4 products.
Great question Paul. And one that goes to the heart of conducting data-driven research.
Even tho the underlying index is the same, the ETF providers evidently have their way of applying the leverage. Even if their processes are identical, if you think about it, if the volume is different between ETFs then the ETF provider will source ‘leverage’ from a combination of index constituents and a live trading futures market, aligned with buy and selling demand.
This results in slightly different daily ranges, which feeds into slightly different ATRVE values. Which, in turn, feeds into slightly different ATR multipliers that comes out of a data-driven parameter seeking process. And a fairly crude one at that to ensure not to curve fit. And one that avoids the temptation to manually override the data-driven process with human biases.
The net result is what we have-different signals.
For example, 19 Feb 2025. UPRO signals an entry which results in a loss trade. SPXL doesn’t. An amateur system designer would typically manually alter the UPRO entry signal. We don’t. Because that is a slippery slope step towards curve fitting.
Karl & I didn’t count the hours spent resolving seemingly tiny research issues such as rounding errors. Or Volatility Ranges on which to apply ATR multipliers. Applying different rounding limits even between the 5th to 9th decimal points can lead to different signals, such is the dynamics or computing with floating point numbers when dealing with datasets on which price splits are applied quite regularly.
It could even be seen as good thing that the 3X ETFs of the same underlying index are not exactly the same.