I recently wrote an algorithm that performs wonderfully (low beta, high alpha) to try to qualify it for Q's monthly contest. Well then I noticed that I was violating rule #1, that the algo must always hold a short position. ( My original app determines if its time to get out of the market, and if so allocates funds to a careful pick of what IS going up, if anything.)
Anybody know if non-leveraged short ETF'S are excluded from the contest? I read the July rules, and it's not clear to me. Let's say I was going to buy 50% SPY, 50% TLT on a given day. Would it be in violation of the rules of the contest (or at least its spirit) to allocate as follows:
=50% SHY , +50% TLT.
Anyways for what its worth, the following is a simple test to see if shorting a reverse ETF would be better/worse than buying its reverse emulated source ETF. So this compares SPY to SHY. I expected shorting SHY to underperform SPY, because you face borrowing costs. Does Q's algo not account for those?
FWIW, the plot thickens if you short the SH one time at the start, and never touch it thereafter. That produces only half of the gains!
I'm really not sure why.