I'm assuming latency is irrelevant for regular non marketable limit orders. But what about stop limit orders? When price hits my stop, I need that limit order to hit the books ASAP. What are the controlling factors in executing speed for these orders? Is it the brokerage speed? Is it my internet connection speed? Is it my software program speed? Asset is trading at 19.80. I'm short and don't want to buy back unless I have to, so I set the stop at 20.10 with a limit order to buy. I'm guessing the brokerage now has my stop order, so it's up to the brokerage to execute quickly? Are DMA brokerages faster than free brokerages?
I would assume it's how much Citadel is paying your broker. IMO, payment for order flow is all about screwing customers in situations like this.
Our STOP and STOP Limit equity orders are held at EDGA not on the local software or our servers, so very fast. They are only active during regular market hours. Even though Latency is VERY low, if you are trading volatile symbols, these type of orders often have real slippage because the price many traders use can often be duplicated by a large number of traders as it represents a break down or break out price where many have done the same.
thanks for your insight. What kind of slippage would you expect for a stop limit order in these situations? Would the order just not fill or is it still going to fill at an unexpected price? Also, how is priority determined on the limit order books if everyone is submitting ultra low latency orders at the same time?
I have no way to determine the slippage on a STOP Limit order as there are way too many market conditions. It might fill on the bid, lower or just post. No way to tell. Priority on any ecn or exchange is based on the time it arrives at that price. A limit order through the price and a market order should be executed the same. The only difference is in a big gap, the Limit might just post.