Model vs reality using limit orders

August 9th, 2016, 02:44 PM

I use a strategy with limit orders. There arise differences when the limit is right at the top or bottom of a bar. Some get filled and some not (logical). The shorter the time frame the more this occurs. Does anyone have stats from real trading how big the differences between model and reality are for different time frames. eg 3, 5, 10 min. I trade the ES and the US.

August 13th, 2016, 11:38 AM

I think your question needs more quantiative context. Namely, what is the top of the bar in relation to the close of the bar? Is the top of the bar also the close of the previous bar? Is the top of the bar the top of a pin bar? How much volume traded there previously?

If the top of your bar also coincides with say the PoC, you shoudl be pretty confident about a fill. If the top of your bar is say the 52W high and only 300 contracts have previously traded there, then pretty low chances of a fill.

For example, I remember I had resting limit orders sitting at 2183-2185 to gamma scalp, and only got about 60% of my limit orders filled. And those orders had been waiting for about 2 weeks in the queue so to day.

I would actually say though, the shorter the time frame, the more likely the fill just as a function of time�

August 15th, 2016, 02:55 PM

I've done some visual checks. It is difficult to automatically create a table in TradeStation with statistics. My model generates more than 1,500 trades per year. The volume on ES seems sufficient per bar. My trades last on average 13 minutes. During the trade I change my limit orders.

Model vs reality using limit orders

Discussion in Emini and Emicro Index

Model vs reality using limit orders