not a tech even by the tiniest fraction but have a brother in sytems at cboe, and the exchanges are constantly adding and testing ,their biggest worry is an overload on a catastrophe day,and when do you think the folks that trade like you will be putting in the most orders ,when the system probably peaks at 3/4 of that..systems are finite
OK. I see where you are coming from. CBOE has issues because options are connected to stocks and the messaging rate is enormous. CME doesn't really have the same problem. I don't trade stocks. I find the practice of quote stuffing on stocks to break data feeds to be highly irresponsible and it just raises costs for everyone. Sensible solutions like proper messaging ratios (like CME, others) or minimum order life should be implemented. The CME is more of a monopoly than stock exchanges so it can implement rules to protect itself without fear that volume will go elsewhere. There are also bad things about it being a monopoly but this is one of the good ones.
Appreciate all the input on this topic. @ Gara It seems I lack the intelligence to formulate a winning strategy with a 100 to 1 ratio. My primary focus is on the sudden and apparently random bursts of volume + price spikes that occur on many/all(?) markets. Since I have no way of knowing when Mr Big will be in dire need of hitting the MARKET BUY button, how can I function without constantly re-positioning my limits in anticipation of an eventual (but I stress again, effectively random) volume-price-spike? If CME won't even let me re-position orders every couple of seconds during the overnight session, I'll most probably have to call it a day on trading.
First. From what I remember, message rates are per contract not per fill. so if you are quoting 100 lot you get to use 100 X as many messages as someone quoting a 1 lot. Next. I imagine you are just pegging your offer = ask price + x. Buy = bid price - x. X needs to be bigger than 1. In reality, given your latency and the profile of these spikes X needs to be bigger than probably 5. So then you change to offer = ask price + x +/- some padding. So then you only requote if price change > padding. That will bring your message ratio way down. Also helping is that you don't need to set it to requote every 100ms or so. Many things you can set to re-calc and decide on a slower time frame. Of course I wouldn't be explaining all this if that was all you need to do. As I described, some things will make money but not enough to offset the 'gotchas' inherent in this setup. (Imagine what happens when your power goes out, internet goes down, broker goes down, accountant charges you 10x what they used to... etc) To make stuff really work this is basically step 1 in a 20 step process.