Server Side OCO Orders and Connection Failure

[jpennybags 11]

My brokerage is getting out of the retail business, so once again I'm faced with choosing another futures broker (I like self directed change… not so much when brought by circumstance). From what I've read recently, consolidation in the business may be a trend that will continue. Some of this has been prompted by Dodd-Frank regulations, and some balance of this trend is due to the economy of scale. I don't really have an opinion one-way-or-another if this is a good thing. I do think that it may eventually limit choices for small retail traders. That's not what this thread is about, and I only offer it up as an opinion of what may happen going forward (part of my thought process).

 

I have experienced power failures, and internet connection disruptions in the past. I haven't kept records on these occurrences, but in a general sense, my recollection would be 3 - 5 times per year (If a tree falls in the forest, does it make a noise?). A little backfill is necessary… hang with me.

 

I use Sierra Chart, and really, really like placing (entering) trades from the chart. When I first discovered this functionality, I wanted to throw rocks at my old trading platform. I also like using attached OCO orders for stop and target. My trading style is such that my average "time in trade" is < 120 seconds. In a typical trading day < 35 minutes with money at risk. For anyone who places these type of bracket orders, please understand that if the server that you route through does not support OCO orders, then these orders will be managed locally at your trading platform. So, if you lose connection for any reason you have no stop loss or target in the order que.

 

To date, I have never had a connection disruption of any kind while in a trade. I'm not bright enough to come up with a risk profile that would actually put a number on the odds of my eventually getting pinched in a trade, but I think the odds are that it will happen eventually. How scary is that number? From a business perspective, I could get pinched once a year for $5K and it wouldn't hurt. It would be a pisser, but I could write it off (mentally) as the price of doing business. Trading the TF with (3) contracts per trade, price would have to move against me 150 ticks before I would approach that number. Could that happen? Yes… yes it could… possibly worse. Hopefully, I could get through to the order desk and get it shut down before too much damage is done.

 

There are brokerages that route through servers that accept server side OCO orders, but your choices become limited… I like choice.

 

I would be curious to hear thoughts or possible solutions to the problem. Honestly, I could go back to doing business without attached OCO orders, but I like them. I don't know that I would want to give up trade entry from the chart though.

 

Thanks.

[Patuca 13]

Last i heard sierra supports the OEC (0pen ecry) platform and oec also has OCO capability. OEC was recently obtained by Gain capitol. You might wantto check them out since you like sierra and oco orders. If you do so let us know what you find out about them.

[jpennybags 11]

  Patuca said:

Last i heard sierra supports the OEC (0pen ecry) platform and oec also has OCO capability. OEC was recently obtained by Gain capitol. You might wantto check them out since you like sierra and oco orders. If you do so let us know what you find out about them.

 

Certainly, if I find a solution that works for me, I would be more than happy to share my findings... and explain why I made the move.

 

Certainly, there are brokers that are able to offer server side OCO's... you are limited though.

 

In part, the problem that I've seen (gathered from my research... incomplete as it is) is that if you have a Rhithmic feed, Rhithmic has a solution, but it doesn't support third party software. R/Trader will support server side OCO orders, and R/Trader Pro will support trading from the chart, but it only supports time based charts. I place trades from a 40T chart on SC. If R/Trader Pro supported a larger variety of charting types, I could see myself making the move. As it is, it doesn't work for me.

 

From what I've gathered from viewing the Sierra Chart forum (also to include Ninja Trader forums) there has been a plethora of requests for server side OCO capability. Both parties lay this off on Rhithmic and others. It seems to me that if Rhithmic is trying to sell their own software, server side OCO's would be a selling point (to some extent) and why give that up to third parties. I don't have a problem with that from a business perspective, but it's no help to me and what I want.

[jpennybags 11]

As an update:

 

I communicated through the support board at Sierra Chart about a work around for OCO orders. They offered up an idea, that they may eventually implement as an option. Instead of server side OCO's: If the parent order fills, a regular stop order would be sent to the server, but the target order would reside locally. If the target order is triggered, the platform would then send a cancel order for the stop residing on the server side. In this way, you would still have the stop order on the server, in case a connectivity issue arises. The target order is lost but the stop is in the queue on the server side. No word on a time line.

[jpennybags 11]

  jpennybags said:

As an update:

 

I communicated through the support board at Sierra Chart about a work around for OCO orders. They offered up an idea, that they may eventually implement as an option. Instead of server side OCO's: If the parent order fills, a regular stop order would be sent to the server, but the target order would reside locally. If the target order is triggered, the platform would then send a cancel order for the stop residing on the server side. In this way, you would still have the stop order on the server, in case a connectivity issue arises. The target order is lost but the stop is in the queue on the server side. No word on a time line.

 

As an update: Sierra Chart has begun work on the project... should roll out shortly.