READ-ONLY

The forum is now a read-only archive.

For bug reports & platform questions → [email protected]

Our community lives on Discord and YouTube — come join us!

The importance of swap

3 replies

Ben

Subscriber, bbp_participant, customer, community, 4 replies.

Visit profile

4 years ago #276181

Let’s imagine that a random half of your strategies in your databank, the first half, are not charged commission, and the other half, the second half, are charged commission normally. What would happen when you apply a filter such as profit factor? Some strategies from the first half are too easily passing the filter causing lots of garbage from the first half to get through. So we make the filter tougher or maybe we add some cost to both groups via the spread, right? Well, then the good strategies from the second half won’t make it through and are wasted…

Not including swap is similar to this situation. The long holding times strategies are getting a really great deal, they are missing a lot of the trading cost of real life, they are in the first half from the example above. The shorter holding times are priced properly as there is little or no swap that needs to be added. The first half has an unfair advantage and passes all the filters too easily. If we add cost via spread or commission to make up for this, then many of the good strategies from the second half (the short-term holding strategies) are filtered out.

This is what I mean when I say not including swap causes the stats to be skewed towards longer holding times.

0

Tomas Matejka

Administrator, sq-ultimate, 2 replies.

Visit profile

4 years ago #276219

Hello,

the request for adding the swap into the SQ builder is the top voted task in our system I we consider adding this into next update

0

Ben

Subscriber, bbp_participant, customer, community, 4 replies.

Visit profile

4 years ago #276234

If the intended function of SQX is to have a reasonably accurate backtest or a backtest that reasonably matches the native engine it is trying to reproduce then missing a trading cost which is sometimes in magnitudes of the commission, which can cause a PF 1.7 backtest to become 1.4, which is sometimes making up 40% of the profits of a backtest could be considered a bug so thank you for considering including this top voted task into the next build so we can include swap in the SPP studies, optimizations and monte-Carlo tests.

0

Ben

Subscriber, bbp_participant, customer, community, 4 replies.

Visit profile

4 years ago #276309

Viewing 3 replies - 1 through 3 (of 3 total)