HFT at the CME - rigged to those that can pay

At CME it's all equidistant. They have a giant room full of rolled up fiber to make it so.

 

As I stated in the post, it is probably at this time impractical at best to try to equalize multi-exchange at this point unless you slow down everyone to a crawl the way Katsuyama does by adding 30 miles of coil from the switch to the matching engine. Even that is possible, but ugly.

But within a single building, there is absolutely no reason that distances can't be equalized.

 

Link?

 

I understand the premise, but what does it accomplish?

 

You understand the premise but you don't know what it accomplishes? What? Seriously if I have to answer that question you don't belong in this thread.

 

I don't think you know the answer to the question...If you don't realize someone will always be faster than you, even with the changes you suggest, you don't belong in the markets.

 

Nah, I'm finding this to be a useful filter. A whole bunch of people who don't host there and don't know how it works just shooting their mouth off without bothering to even call the CME to ask.

This is all well known and was a main feature of the Aurora build out.

It's a level playing field. For about $1200 a month you can put in a sever and get your data and send orders just as fast as virtu or anyone else if you are handling your hardware correctly. So if you are not competitive it's all on your end and has nothing to do with the CME.

But this is great. I can't wait to sell my ignore list to the highest bidder.

 

I am finding the word "equidistant" in the docs:

 

This is treating the patient by treating the symptoms. Make it less about a latency arms race and more about committing traders to their orders. Cancellation fees (that HFT can't weasel out of) would put a significant dent in the "who's the fastest" angle. Obviously for cases where they're going to put the trade on regardless it won't make a difference - but then neither should 10ft of fiber.

 

This only handles things up to the point of demarcation where they hand it off to the server itself. Then it becomes about who can use the fastest NICs, who can write the fastest software that's residing at the kernel layer or even no kernel at all, just full blown custom hardware.

Are you saying CME is utilizing some kind of complex per-switch QoS to guarantee a level playing field? I seriously doubt that. Sure they might be providing equidistant layer 1, but that's *just* layer 1.