REALLY bad fill

Pangolin Z had an open order of 8 shares of GHC, purchased at 703.58. This morning the stock opened at 717.97 and the LOD is 716.29. It is currently trading at 725.96, up 7.16 on the day.

Sounds good, right? Well, C2 closed the order as it thought the stock price had mysteriously dropped to 71.88 (a loss of 90%), thereby tripping the stop loss and costing our system slightly over $5000. Please fix this immediately.

Hang on, Kevin. We’re looking into this. We need to investigate before fixing.

By all means - my first thought was that there must have been a 10:1 split but no dice. I am confident that you will correct this, Matt. Thanks for the quick reply.

Matt - I received your email. Can’t say that I am happy (I am not) but since there is nothing I can do to get this properly fixed, and your hands are tied, complaining any more is a waste of everyone’s time. In actual trading I’m sure this order would have been rescinded and the trade would still be active.

I dont understand what you want MK to do , your live auto trading clients closed their positions at 716… already , MK cant add virtual profits to your account it doesn’t make sense .
In real life problems happens systems go down , electricity cut … etc , MK cant give system developers immunity form such events . The one thing MK can do is if your fill price is wrong and doesn’t reflect reality then he can correct that for you .
Cheers .

I agree that nothing can be done to fix bad trades in the past. However, I’ve come across many occasions as a subscriber where an autotrader’s stop or limit got filled prematurely and it triggered market orders for all autotraders. It would seem that some simple logic could be added such that if an order’s fill price does not coincide with the stop or limit price that the fill is ignored for the purposes of keeping the subscribers’ positions in sync. Perhaps that is too simplistic, but it seems like it should be a solvable problem.

Just my two cents…


Just to show this is not an isolated incident, take a look at the most recent closed trade of Pangolin Z ( MAC was supposed to close out with a limit at 65.74. If you look at all of the autotrader fills, you can see they all closed at 65.69. If it were a case of one lucky autotrader getting filled at the limit price then I would expect to see that trade listed, in addition to the unlucky autotraders who had to exit at the market a few seconds later to stay in sync.

In this case the difference was only a nickel (and it could prove to have been better than waiting for the limit), but enough of these premature trades over time would seem to cut into the subscribers’ profits (not to mention the loss of confidence in the autotrading engine).


This morning there was another premature trade in the same system. SBUX was supposed to be purchased at a limit price of 76.79, but the closest autotrade fill was 77.04 – a full 25 cents away from the limit price!

As the system developer, I can tell you that this happens a lot - C2’s autotrading system seems to ignore the limit orders as often as it hits them. Despite these constant erosions in trade profitability, the system stats are still holding up fairly well.

And the entry price for SBUX was even worse than what you reported - it was 77.08, almost a full half percent worse than the limit entry target. For short term trades like these, this can significantly impact performance.

Hi, Guys:

SBUX traded in real life at 76.76 at the market open. When your limit price fills anywhere in the world, C2 will convert all non-lucky-people-who-didn’t-get-the-limit-price to a market order.

Hi Matthew,

Thanks for your response. This logic makes sense for simulated results when there is no autotrading data, but wouldn’t it seem to make sense to use data from live trades when available? I realize this would add some complexity your software, but it seems like it would eliminate most of these cases of premature fills based on bad ticks.