Hi Khaled, might it help you, in a list of 182 closed trades on 24 trading days, I found 5 of those myself. They are all from the same date (Dec. 16).
I looked into a couple of these trades and I don’t see anything special about them.
The first one going odd is the first started trade of that day. Maybe a glitch ?
Don’t laugh … (or do, why not actually) …
Attachment 1 … been working on that the whole weekend. 1M bars, and the result I see is consistent with what I am used to. Could have been somewhat higher, but all in line with shorter periods. Ready to put to Live, as I though 2 hours ago.
I noticed the for me very short time InMarket at least today the whole day, and the low number of trades disappoint me (less fun and it should be twice as high at least). … then I finally saw the time period in the top part of that screenshot …
Stops at Dec. 16 ??
I adjusted the start of the Backtest period to Dec. 16 and observed 9 trades. Only that day, and no further again (should be till last Friday Jan. 7).
Skipped another more day to Dec. 17 and the result is in attachment #2.
Huh ?
Of course, markets change, but this ??
Dear people, this *is* for a bit of fun, but it might be to warn you that if you run into similar unexplainable issues, it could be the historical data. In my case this is EUR/USD S1 data, 1M bars.
Obviously I thought of my own post quoted above, because that was already strange and from the very same date.
Most probably it is me and my code which did not want to trade any more from of Dec. 16. Debugging that is quite complex (my code allows postponing trading) but it is odd (for me) that if I jump over that date that a. all starts to work again (bugs can be nasty, so still feasible) and b. suddenly all starts to lose all over (can still be because it is me who writes the code).
The real fun message is : I have been working the whole weekend on this (with an explicit target, resulting in the 1st attachment – so as expected), and at 5 to 12 I just in time see this issue. Moral : probably a complete weekend wasted (think 20 hours).
In the end it is a warning for all, might it be a data issue indeed.
PS: In attachment #3 the 150K bar data of the same (last two days of what’s in attachment 2). Of course these matters don’t compare easily because of different trade-time-starts, but the dive of the 2nd attachment comprising the same period compared to this #3, is just too much of it.
.
.
