Autospreaders

[patrickrooney]

@dextor There are plenty of ways to avoid double fills.

You could quote only one leg.

You could submit sniper orders, which don't quote and only sumbit orders when the spread may be executed, in Autospreader https://library.tradingtechnologies.com/trade/as-submitting-a-sniper-order.html.

You could use a combination of backoff and payup
ticks.
http://library.tradingtechnologies.com/trade/asr-reference.html

You could create many different imaginable rules using the proprietary Autospreader Rules widget.

Please reach out to me if I can assist you with any of this.

Happy Thanksgiving-

Patrick

 

[Robert Morse]

@dextor There are plenty of ways to avoid double fills.

You could quote only one leg.

You could submit sniper orders, which don't quote and only sumbit orders when the spread may be executed, in Autospreader https://library.tradingtechnologies.com/trade/as-submitting-a-sniper-order.html.

You could use a combination of backoff and payup
ticks.
http://library.tradingtechnologies.com/trade/asr-reference.html

You could create many different imaginable rules using the proprietary Autospreader Rules widget.

Please reach out to me if I can assist you with any of this.

Happy Thanksgiving-

Patrick

I'd say given your description, this is a good choice. You should be able to calculate your single leg and spread margin with a simple spread sheet.

 

[sle]

As a voice of dissent, I have to say that I have never seen a good enough commercial spread engine. Either you end up with fill rates that are very low or you end up with mismatched leg executions. Most people I know ended up building their own.

 

[Robert Morse]

Most people I know ended up building their own

The Real Tick spreader is server side. I believe the TT one is cloud based. Unless I'm hosted in the CME data center with a FIX or direct iLink connection, I don't see how building anything that sits in my office can be faster.

Happy Turkey day sle...cheers

 

[sle]

The Real Tick spreader is server side. I believe the TT one is cloud based. Unless I'm hosted in the CME data center with a FIX or direct iLink connection, I don't see how building anything that sits in my office can be faster.

Happy Turkey day sle...cheers

For practical purposes, it’s an HFT strategy, except with more relaxed alpha requirements. It’s going to be latency sensitive (like most microstructure plays), so it has to be colocated, optimized for specific markets etc.

 

[Robert Morse]

For practical purposes, it’s an HFT strategy, except with more relaxed alpha requirements. It’s going to be latency sensitive (like most microstructure plays), so it has to be colocated, optimized for specific markets etc.

Not a retail solution for a small account.

 

[sle]

Not a retail solution for a small account.

I was going to say “it’s only going to cost 3-4 grand a month” but realized how ridiculous that sounds
Happy turkey day!

 

[i960]

As a voice of dissent, I have to say that I have never seen a good enough commercial spread engine. Either you end up with fill rates that are very low or you end up with mismatched leg executions. Most people I know ended up building their own.

Not to mention that they're all insanely overpriced for what they actually do technically. This isn't a rocket science problem but it's sold as one.

 

[dextor]

>> You could use a combination of backoff and payup ticks.

@patrickrooney , I have in fact used payup ticks while using TT. I know this may not be the fault of the paid up ticks but when I attempted to close the trade, it reversed my positions. TT support came in and attempted a fix - remove the quote on 1 leg instead of 2. My spreads since then closed properly but after a while the problem returned. Out of frustration, I changed to T4. Maybe TT has fixed the issue, we don't really know until I go back using TT.

Now with T4, everytime I put a spreads trade, the bid ask spread seems to be a lot wider than a trade done in TT. One could debate that volatility, timing could have played a part, but as a person who keeps seeing this; it's down to either I am not setting up my spreads correctly, or that TT has some advantage over T4 - yet to verified this.

I cannot help but feel that building an autospreader on my own could be a better option. The interesting thing is I want to earn a profit and not waste time troubleshooting autospreader issues. See the catch 22. If you need to build one, you will need to spend time troubleshooting. Until I trade bigger accounts would building your own autospreader be the way to go - lower cost and to resolve problems and not deal with software companies making changes creating new problems.

 

[sle]

Not to mention that they're all insanely overpriced for what they actually do technically. This isn't a rocket science problem but it's sold as one.

It’s not that hard to build something that beats commercial algos but a fair bit of work to build a good one. A good one would be product-specific, have varying degrees of agression, have a liquidity profile customization etc.

PS. I am surprised someone like IB does not offer a semi-lousy one as a part of their general software