StingRe wrote:

what happens when you create something brilliant

Thanks for the praise but I think a quick reality check may be in order: we don’t know yet if it’s brilliant or not. Based on the data we’ve got, it looks like it could have potential — and that’s about it. 32 months is still a very short back test. PRT v11 will hopefully be available later this year and then we’ll see what it looks like with a proper 10 year back test. Could be a bust.

Some of the best trading advice I ever received (and often v hard to follow) is, always look out for reasons why a trade might fail. It’s all too easy to talk yourself into a false optimism — and I am the world’s worst offender!

If you look at the equity curve for example, there is a very flat bit from April to Oct of 2018. It does virtually nothing for 6 months. That’s a long time to wait, not knowing if it’s ever going to fly. That could happen again in 2020, in which case we’ll all be disappointed.

Of course I hope that it makes us all stupidly rich. But as they say, don’t count your yachts until they’re on the water.

(I think that’s what they say, don’t they? something like that?)

fifi743 wrote:

things missing so that the algorithm is not rejected

Oh no — help! Like what, par exemple?

fifi743 wrote:

If it is Friday and it is more than 11 PM and algo want to make a change.

Okay, I double checked and results are almost identical with Ctime = time >= 0 and time <= 230000

That solves the Friday problem, no? Can’t open a trade after 11pm.

But why Monday? If market is open from 11pm Sunday.