thanks for sharing this is a nice one. how do you deal with the risk of not getting a fill quickly at D and then the price runs away from you? thanks.
Depends on the specific circumstance. In some markets, you can specify a minimum fill quantity (so if you're resting $1M at hotspot, you can specify that you don't want to be filled on anything less than X, at the risk of losing out on some order flow). Of if you're trading something like stocks on Toronto vs US, you're always carrying small partial USDCAD positions and have to weigh the cost-benefit of how often you want to hedge. Typically it's a minor consideration in the grand scheme of things since you're round-turning positions so often.
My algos can predict 1 tick moves pretty accurately without any increase in X on my part. Any advice on how to profit from it, since it is still a looser net of fees. Thanks
I appreciate your candor and you taking the time and answering all the questions. Your this line "Conceptually it is identical to what we do. " is telling. It implies that reading some academic literature can point in the right direction of what is getting done "generally" in the hft space. Thanks for the detailed strategy - As I expected, very simple concept but impossible to exploit unless one is HFT. I will surely make an attempt towards this in due course. Will be glad to let you know how this goes. But given where I am currently on the technical and access side, its going to take around 1-1.5 years for me to get there.
No I sure don't. Maybe my statement was unclear when I said "I can do that" - I meant I could give you that signal, not make money on it. Like you said, it's still a loser net of fees if you're crossing the market to enter and exit. So you have to use it only to bias the orders you rest, as opposed to crossing, which then requires consideration of other aspects of market-making. If you could extend the timeframe to predict 1.5 or 2 ticks though, you might have something useable as a longer-term crossing signal. It's very hard to do, but maybe you've got a decent start on it already.
That's essentially the underlying problem to be handled when making markets, regardless of the market. It is what we spend our livelihoods trying to figure out. I can give you a super secret hint: Get FAST.
nothing personal, but all frontrunning strats aka hft need to be taxed. hold for 1 minute or more or pay a huge tax. period. leeches. i say 1 minute as i am willing to hold for that long.
hft, In Market-Making Strategy (trying to capture bid/ask), in what condition ? no trend market ? because it's maybe dangerous .. You add some prediction market direction movement for "always" to do a good direction ? You do on FX market ? you need pairs with "large" spread but with good liquidity ? thx