I'm looking to find others interested in forming a co-op, or at the very least, a simple collaboration. Initially, I figured I'd do it all myself. The problem is time. So, I'm here looking for other like-minded individuals to see if we can work together. We would share profits. How we would divide up those profits is open for discussion.
Who am I looking for? Experienced developers. You are either already familiar with exchange APIs or willing to learn. You have an interest writing tools and connecting to markets. You are flexible and write clean code that others can follow. You do not need to come up with the trading strategies. That will be my job. However, if you have ideas, you are welcome to share. You do not need to be a trader, however, it would help if you have traded BTC on various exchanges.
An important part of what I need is someone who is familiar with everything Bitcoin. Someone who can answer my stupid questions. Yes, that is right. I haven't traded Bitcoins before (keep reading for my qualifications).
There are no plans to make this an open source project. Perhaps portions of it, or specific tools could be made open source. But, the actual signal generation will not be open source. Successful trading strategies in low liquidity markets are capacity constrained, and giving it out to anyone and everyone will only serve to dilute returns or extinguish them entirely.
Who am I? What will I contribute? My full-time day job is as a quant/developer for a high frequency trading firm that deals in US equities. To this point, I have little experience with Bitcoins and no experience trading them. What I do have is 15+ years trading experience, 10+ years experience mining tick data, developing profitable strategies, and writing trading tools. I will take on the strategy development, coding the trading logic and writing some tools.
I realize HFT is unlikely to ever be possible with BTC. However, there looks to be plenty of opportunity for profit. I've been watching for a while and have done some preliminary work. Thanks to exchanges that seem to be running stone age technology, we can take advantage of it. However, in order to find out, we need to do some more exploration. I'd like to have your help to find out.
I don't expect you to work full-time on it. After all, I'll probably only be able to devote between one and three hours a day on it. However, I want to find others that are serious, motivated, and have a real interest in pushing the limits.
Anyone out there potentially interested? Respond here or PM me.
You're going to get toasted in this market with those credentials buddy. HFT is is not really possible, there isn't enough liquidity and the exchange fees will eat you alive. You also need to understand that the exchanges here do not speak FIX, there are no dedicated connections and communications are notoriously unstable and unreliable especially during high volume periods.
If you're lucky an exchange will have a websocket based API at best. However many exchanges actually implement their API over HTTP GET/POST and rate limit you. A few will even delay quotes by a few minutes. Thus by the time you get a quote and make a decision the time to act is passed.
The most consistently profitable strategy for me has been to trade using a VWAP algorithm that buys when the price crosses below the VWAP and sells when it goes above the VWAP. This has the advantage of helping to stabilize the price for everyone. Even that isn't 100%. I've rewritten my old ATP bot in node.js and have it working on CEX.io.
Since doing that it seems to have pushed the market price of GHS/BTC down by 20% (which was not my intent). It's stablized a bit since I limited the max trade size, but the price is still trying to crash through the floor (sorry guys if any of that is the fault of my bot, I'm still working out the bugs). I choose CEX.io due to their lack of fees. At the end of the day I move my BTC gains to another exchange daily to swap back to USD and sweep to my bank account.
I recommend before you proceed that you look for the AidoATP bot/thread on this forum. There is no use in re-inventing the wheel. The strategy engine is completely customizable without the need to code. I haven't touched the Java client in quite awhile, but Aido did some wonderful stuff with it in my absence and at this point I'm willing to call it his bot.
When my node.js bot is more stable and capable I'll probably release it to the community, but I have way too much on my plate right now to dedicate a huge amount of time in getting it ready for public consumption.
In the meantime I wish you the best of luck. I too have a background as a developer with a major trading firm. Instead of strategies, my work was with trading infrastructure. I still can't get into the details because of NDAs, but I can tell you now that normal strategies just don't work with crypto currencies because the market is illiquid and there are quite a few whales in the pond.
