Quote from: ustone on March 31, 2014, 09:22:17 PM

Quote from: ustone on March 31, 2014, 08:28:52 PM

Quote from: CambioBTC on March 31, 2014, 07:05:45 PM

Quote from: ustone on March 31, 2014, 06:14:52 PM

can somebody here confirm that stops sometimes would not executed? i had same experience two times in the last couple of days....i assume if the stop price is triggered a market order should be created..today i waited ca 1 minute after stop price was triggered but no execution was done...so i slipped 8 points into negative....have i missunderstand something?

Stops not executing,
that is something new,
I've never had that problem,
and would not want to.

Are you sure you are placing
your stops correctly ?

Not to talk down on you,
but if buying then a sell stop is chosen from the the sell side.
if selling then a buy stop is chosen from the buy side.

here you see a sacreenshot from last friday..even its a crap resolution you can see following: time upper left CET. Stop 507, 3 hours old..look at bitcoin wisdom and you'll see that in these 3 hours the price moved over the stop...490 to 510 22:39-22:42 CET...price was not long above 507.. that's why i thought that the stop doesn't trigger a MKT order....am i wrong?

ok..here's what i got from the support...

Stop orders are executed based on ticker price value. Sometimes, for big
orders, the execution price moves much higher or lower than the ticker
price, because by the time the order is executed new asks or bids are
posted.

In the case of this quick spike to 510 (assuming we are talking about
the same one), that is what happened.

I guess that could happen
the way support explained it,
basically a gap, like when
the market closes and open
back up on the nyse / nasdaq
after after-hours trading.

but they also referred to big
orders, was you order a big order ?

Scary though,
thanks for the heads up.

Good Luck,
with it.

P.S.: You can also try using Qt-Trader Rules to exercise your stops from client-side
i always thought having one's stop server-side at BitFinex was safer than client-side,
apparently not.