Nice! Yeah that makes sense.
In theory if the back-testing is accurate then there should be no need to test further anyway right? I spent years working on algos, but for me it didnt matter how much testing I did they never seemed to work as I expected. I might give it another go in a few years time.
Right! Trying to get Back Testing and Live to "match" is a mistake, "fool's errand", imho. Likewise for forward testing and Live. It is not the goal. Instead, the goal is to get a workable system you can take to the next steps.
I.e. I realized like any production software, you have to "release it" and make it work in real world conditions. There is no other test. At least when something goes wrong, I have enough information to debug* quickly.
There are still 4-5 more things that I would have liked to do, but could not get them to work in back testing because it was not live. However, I am putting in all the conditions, up to, the order phase, to test them under Live conditions. Essentially creating-generating phantom orders under live conditions. IMHO that is how one can develop: Back Test, Sim, Live with Phantom orders, then just standalone Live. Each step has its own test suite and goals.
I might pause at some point and put them in and go sim (skip Back Test), but until the end of February, I will be live working out the system.
Cheers
*Testing is for finding the bugs that are unknown, Debugging is for when you know you have a problem, from testing, and need to track it down.
Anyways Back to ES Trading. This year should be pretty good action, if January is any indicator.
