Price improvement strats with ECN

I'm looking at how to squeeze every last pip out of my trades, so concentrating right now on the execution. I'm thinking of price improvement strategies, and so far the best I've come up with is once I want to get in, put in a limit at bid or ask +/- spread, wait for x amount of time, if not hit, replace with limit at market.

X amount of time I'm thinking will be quite a small window.

Placing a limit +/- the spread seems fairly arbitrary too - so possibly a calculation based on typical volatility for that time of day?

So, I've got a lot of backtesting to see if it improves overall profitability - but has anyone else tried this kind of thing, tested it, or have any brighter ideas?

So what you want to do is buy the bid and sell the ask to "safe" the spread right?

Im doing that all the time. With currenex platform you can place these type of orders with one click which makes it quite easy.

Exactly - same with IB. This is for an automated strat. Saving the spread seems a fairly arbitrary amount; the more I think of it, I'm sure it should be based on volatility (so it could be less than spread or more).
When you've done this, I take it the savings > the pips you've missed by not getting filled?

Absolutely! I trade very short term and if im not getting filled i usually let it go. Other opportunities will come.

I mean i mostly look for fewer pips with higher lot size. It doesent make much sense for me to enter if im not getting filled @my limit price.

The potential problem with trying to save/earn the spread is that in some cases you will not get filled unless a liquidity provider quotes at your price, which means you end up paying the spread.

The likelihood of this scenario depends on the particular ECN setup. I don't have enough knowledge about Currenex to say how it will play out there, but maybe 11111 can shred some more light here.

Glad you popped up. Since I wrote this, I've tested this in detail, and in short, despite it seeming sensible, mechanically, it doesn't work above anything other than around 0.5 pips, and at that level, my testing & data just isn't good enough to be sure even that's an edge.

From a mechanical strat point of view, when the entry signal is good, getting in as soon as possible is best, because the time when there's no retrace and I don't get filled is the measure of a good trade. So much so, I've extra research to do to see if increasing position size in these occasions is profitable in itself.

I tried this over various timescales and results were pretty consistent.

Worth a shot anyway.