Why isn't TQQQ opposite of SQQQ?

Hi Sean,

I don't understand why the the win rate doesn't become the loss rate when I switch from ticker TQQQ to SQQQ.  Shouldn't these two be opposites of each other?

The win rate of the algorithm trading TQQQ is very similar to the win rate of the algorithm trading SQQQ because the algorithm is essentially placing the same trades in both cases. With the current parameters, the algorithm compares the price at 11 AM to the closing price of the previous day. If the price at 11 AM is less than the previous close, the algorithm takes a long position in the security. Otherwise, the algorithm takes a short position in the security.

TQQQ and SQQQ have opposite daily returns. Therefore, if the algorithm takes a long position in TQQQ on day x, it takes a short position in SQQQ on day x. Since the securities have opposite daily returns and the algorithm places opposite trades for each security, the win rate of the algorithm doesn't change much when we change TQQQ for SQQQ.

The fees also seem to be significantly lower on the SQQQ example even though it places the same exact number of trades as the TQQQ example?

The fees are lower for SQQQ because the share price of SQQQ is higher than the share price of TQQQ, so the algorithm buys fewer shares of SQQQ. The algorithm is using the default fee model, which charges $0.005/share. The algorithm can buy more TQQQ shares per trade, so it ends up costing more fees to trade.

Best,
Derek Melchin

Want to invest in QuantConnect as we build the Linux of quant finance? Checkout our Wefunder campaign to join the revolution.

Hi Derek,

Thanks so much!  I don't know why I couldn't wrap my head around those two points.  Thank you for making those clear and easy for me now.

So maybe there is no way to have the loss rate become the win rate on this attached example?  Can you think of a way?  I thought going long vs going short on the same security would have opposite results.  When I use Short for the “GreaterThanOpen” parameter in the attached example, I see a 40% win rate.  When I switch the “GreaterThanOpen” parameter to be Long instead of Short, I see a 45% win rate.

Thanks!
Sean

Hi Sean,

In this algorithm, flipping the rules would mean changing each -1 to 1 and changing each 1 to -1 while keeping the security constant. However, if we flip a strategy's rules so that it places the trades in the opposite direction, the win rate doesn't become the loss rate because of commissions and spread fees.

Best,
Derek Melchin

Want to invest in QuantConnect as we build the Linux of quant finance? Checkout our Wefunder campaign to join the revolution.

Hi Derek,

OK thank you!  So the difference between going long and going short will have that few cent difference between the bid and the ask?  That makes sense.  And then the win rate only counts a trade as a win if the trade is positive after commissions are taken into account?

Thanks!
Sean

Hi Sean,

the win rate only counts a trade as a win if the trade is positive after commissions are taken into account

Yes, to view our definition of win rate, see the glossary.

Best,
Derek Melchin

Ah ok perfect thank you again Derek!