Good Remote AND Automated Trading Prop Firms?

E

ezbentley

Can the experienced remote traders recommend some prop firms that have good support for "remote AND automated" trading? By support I mean some API-specific programming just to get the automation part going. I am a C++ programmer so I don't need basic hand holding, but in general it is much easier to start with some example code and make necessary modifications.

Would prop firms have a different expectation when they decide whether to accept a remote trader vs on-site trader? I get the impression that people generally assume automated trading to be high-frequency large-volume trading. However, I think automated trading can be just simple, mechanical, fully unattended trading, without any implication in trade frequency and volume.

Any advice is greatly appreciated.
 
I hate talking about what firm I trade with because I don't want to be a shill for anyone, but you described what I do pretty exactly so I will chime in. I trade remotely and am very close to full automation, about 90+% of what I do is automated these days. I don't really know how anyone who has not hired a programmer and built some sort of automation can keep up anymore, it is just the way the business is these days.

Having been around the block in this business, can say I don't know why anyone who does automation would be anywhere but Echo. You get the leverage of prop plus very good support for developing your code. I can't tell you how many times I have been stumped trying to figure out a problem and will call the guy there that supports this (won't say his name, starts with a J) and he will have the Echo programmers send over a bunch of sample code that I can pretty much cut and paste into my app that fixes the prob.

Some of the recent reviews of Echo have been negative, and I can pretty much guess they are from the small traders that are getting squeezed out by the automated guys. There are only 2 real negatives to Echo, the new Merrill sheets that show you your positions and balances pretty much stink. The info is there, but it is not very pleasurable on your eyes to read it. And a couple times a month I find myself having to quickly log out and back in because a quote server I am on clipped me, which logging in fixes but with my program takes me 5 mins or so to get it all back up and running. I have experimented with back up quotes, but scrapped it because they've been good lately.

As for everything else, the back office support is very good, and it seems they really have morphed the firm into a place that almost speciaizes in remote traders with black boxes. If you have light to moderate volume (would guesstimate under 1 mil shares a day) they will want you to use the Active X through Echopro (Sterling) - Sterling support is a total joke but Echo programmers can troubleshoot Sterling issues better than Sterling can I have found. Once I found my shares increasing past 1 mil per day they offered me some even better connectivity options, they have FIX connections and even let you house a server on the exchange if you are so advanced that you need that kind of speed.

Hope that helps, I know already I will regret saying as much as I did since it seems anytime you get too specific on these boards with what you do or where you do it you get PM'ed like crazy (please no one PM me - I am a hermit who hides remotely so I can be left alone), but since your post so exactly sounds like what I do I thought it would be my good deed to point you in the right direction. Good luck!
 
I won't pm you as you request, so I'll say thank you here. Was looking for input on Echo in light of the post in the brokers section recently. Just about finished studying for Series 7 and need to make the last decision soon.
 
Yeah, as much as I hate putting myself out there speaking out for or against any firm, am thinking about putting up a positive review for them in the brokers section because I know there are a ton of guys here that are happy, none of them ever look at the brokers section I think or more of them would chime in. Think with any of the firms here, a firm will have a guy blow out then blame the firm because taking the blame of their bad trading decisions is too much to bear.
 
Thank you NazSpaz. I won't ask any specifics.

A general question I have is: Do people generally start with small volume at first to make sure the automation is running as expected, and then scale up only if the system runs fine? To me that seems to be a logical sequence.

This is my first attempt to fully automate the mechanical strategies I designed. I have emailed a few prop firms regarding this but they don't seem to respond to me. Is that because they only want high volume automated traders?
 
If your trading is making money manually...and then you automate it, that is the best way to go (in my opinion)... trying to develop a "system" of some sort without actually trading it... tends to be counter productive...Just my 2 cents.


Don
 
Quote from ezbentley:

Thank you NazSpaz. I won't ask any specifics.

A general question I have is: Do people generally start with small volume at first to make sure the automation is running as expected, and then scale up only if the system runs fine? To me that seems to be a logical sequence.

This is my first attempt to fully automate the mechanical strategies I designed. I have emailed a few prop firms regarding this but they don't seem to respond to me. Is that because they only want high volume automated traders?
More...

Don't know why they wouldn't respond, every big trader starts as a small one. Would be very surprised if you said Echo was one of the ones that did not respond, they seem to go way above and beyond for automated guys, they even have test servers that you can run automated stuff against to get fills to make sure all works well before going live, although they don't let you test on those for long periods of time. They say they are for troubleshooting, not for time-testing P&L results, I was booted off one for using it too long. :)

With first question, YES! If you run anything scaled up the first time you will have a very short life in this business. Everything new I have tried or will try starts with 100 lots and runs smooth for weeks to months before I even consider stepping it up. And even when you do step it up, do it in small increments.

No matter what you do or how well you think you have backtested, no testing can account for real-world slippage and other factors. And even running 100s you will find so many things to tweak and change before ramping up.

Let's say your strategy goes well, everything works great, and you start stepping up, no matter what the strategy you will get to a point of maximum profits, say it is 2000 shs per trade for example, where 2000 makes more than 1900, but 2100 makes less, and 2200 makes even less. You need to step up slowly and carefully track your results to find that point where too much size begins to hurt you through slippage. Obviously it is all geared to how much each stock itself can handle, CSCO can handle a ton more size than say CAKE (not that I trade either one, just using an example from first stocks that popped into my head).

Test and test and test, and then when you are ready, test some more, before taking a strat live. The one time you screw something up deep down in the code you'll be glad you did. Hope that helps.
 
Quote from NazSpaz:

...a couple times a month I find myself having to quickly log out and back in because a quote server I am on clipped me, which logging in fixes
More...

NazSpaz, how are you getting your quotes?

Through the Sterling API (activeX sti)? Or are you accessing the Activ feed directly through the Activ API or reading the Activ UDP multicast packets directly?
 
Hi NazSpaz,

Thanks for your insight and I agree with everything you said. I emailed Echo on Monday and they just got back to me now(Thursday). I don't know if it's normal for them to respond couple days later.

I know that real-world slippage can cause divergence from backtested results. That's why I want to start with small volume to see how the strategies perform even though I have confidence in the backtested performance.

The main reason I want to automate my my strategies is not just to trade fully unattended. Rather, some strategies require monitoring hundreds of stocks in real-time. So it is not really practical for manual verification of a strategy before going fully automated.
 
You must log in or register to reply here.
Back
Top