Search results

Yes & no. There are a number of pairs out there where one leg has a much wider bid/ask than the other, and yet the pair tracks reasonably well, with some involving multiple legs. The tricky part, for this convergence only strategy, is establishing which pairs are stationary, and trading off...

Right, but it has to match with other spreads then, yeh? Different order book.

* A 95% correlation with a non-random drift may be unpleasant.

Yes and no. ES / NQ may not always be "friendly", although the idea is an interesting starting point. Edit -- issue with the relative contract size also. ES / NQ:

How's this work with regard to capturing the bid/ask? I assume it's a lightening fast fill, but crossing on both sides? May be a silly question. Un-googled.

If you're spread trading small size and turning it over many many times per day, then there is somewhat less risk than an LTCM blow-up-the-word scenario. Many small trades with positive expectancy. Similar idea, different risk.

Economic value of the thread?

They can be combined.

True. That being said, I have nothing against Pascal. Just not my first choice for interfacing with some brokers API, or screen scraping & mouse spoofing. Not that there's anything wrong with that.

Lots of competition there. ... Broader, sure fire solution to the OP, without so much fiddly tech stuff: a) Use a broker that has a phone based app, and do it manually [as per Bob111]. b) If you're a big enough swinging enough dick, like all ET posters, then surely your broker can...

Agreed .. but there can be a fair bit of "arithmetic" involved in selecting and pricing off an analogue(s).

Non-directional traders with DMA.

+1 I would add that a lot of high frequency stat arb is reliant on bid/ask spread capture to maintain an edge, with the other leg(s) serving as little more than a liquid dirty hedge. (Spread capture probably trumping the mean reversion aspect on shorter timeframes). Quote a price on...

Saw his pic on Amazon. Looks like a chartist. Read a page of introduction. Sounds like a chartist. Probably an interesting read for TA evangelists & other faith based "traders". But WTFDIK.

To whinge about IDE licensing costs admits that your code will not be worth that cost. If that's the case, it sucks to be you. The memory management argument is a ruse. GCs stop everything and their timing is unknowable. So, your real-time trading system will PAUSE from time to time, to...

You can still leak objects if they remain in scope. Crap attached to a singleton for example.

wow, this thread is great training set for my ignore strategy.

Agreed. On the plus side, while the number and variety of people that can see your trades on the inside can be disturbingly high seeing is not necessarily understanding.

Any comments on splitting legs between brokers? I have toyed with this before, and also some size randomization. Either way, a semi-bright internal database guy will probably figure out roughly what a strategy does, although maybe not how to replicate it on their own imo.

Based on maler's posts, I disagree. The devil's in the details. And the estimated mean.