Hi guys, Some questions about automated trading: 1. What are the resources that are available to learn trading automation ? And how complex is it to learn (if one knows programming at a basic level) 2. can automation be used for trading Opts ? And to to determine how big the spreads is and open interest is in the opts ? 3. How sophisticated is your automated trading method ? 4. What is the essence of automated trading ? i.e. what does it make use of in terms of order conditions / basis - for buy and sell order ? Is it indicators ? Or is it something else ?
1. Programming skills required for trading is pretty simple, anyone can learn it. It's about a 1 or 2 max out of 10 on the programming scale . Most sophisticated trading software has it. 2. All it is a set of instructions to perform an action (buy sell, set limits, stop etc) on the asset based on certain criteria. These criteria can be just simply be a upward MA crossover, go long, sell when it cross back, to complex 5000 lines of code 3. I tried it a many times and results were mixed, but I don't rule it out, just not a priority for me. I'd be interested in other peoples experienced. Personally I think the benefit is in screeners or signals rather than actual trading. IMO this is where people lose focus, they get interested in the tools and technical side of trading rather than the primary drivers such as trading, understand markets, charts, news, economy etc 4. Can be screeners, actual trading buy sell exit and so on I don't really have a strong opinion of them either way but I think its a good skill to have just as long as keep focus on why your trading (i.e. make cash, and risks low) On the general theme I notice many people get lead down the wrong path (I did !) whether it be media, hype, YT, 100's of indicators available etc and get distracted. A beginner often says "man you got trade crypto!!!" A good trader can make a lot out of a stable stock like Coca Cola, Crypto doesn't make things any easy, In fact it's often harder as you get can get caught badly in swings and seems not much based or trader sentiment. So in essence my view is to learn the primary drivers of a few stocks or assets, then and use programming tools to just make life a little easier.
1. It depends. Do you plan to use ready-made trading automation software, or to develop your own? In the first case you only have to translate your strategy/algos to code and that's quite straightforward. OTH, developing your own autotrading software it's not an easy feat. You have to deal with broker communication, missing/imperfect data, partial fills, concurrency, risk of losing money because of software bugs. Advanced programming skills are essential in this case. 2. Yes. Possible added complexity, as options can be more difficult to backtest than stocks or futures, for example. 3. My own strategies are unsophisticated in the sense that they boil down to a few rules, although a simple strategy can require a long time to be developed. 4. The primary benefit of automation is that it greatly reduces the need of having to take action emotionally, provided that your strategy is sound. And it gives you more time to work on other things. Furthermore a software can process/analyze data much more quickly than a human, of course. Aside from that there aren't many other differences.
Youtube is all over it. Alpaca has lots of tutorials and their API is pretty easy to use. Go with Python. It works fine for what we do, and it is simple, quick to learn, doesn't require a special IDE or in fact any IDE at all, and it is cross platform. You can even make GUI apps that are small and simple. Your easiest first step would be an app that would give you account information such as your cash available for trading, total equity, number of day trades used or available, stuff like that. Think of it as the "hello world" of financial programming. Another good exercise is pulling a list of all tradeable assets from your brokerage, or price/volume action for a stock. A comprehensive screener is a good heavy project that for your own use could be totally customized and better than anything else available. From there, picking say the top 5 stocks, determining when it is a good buy opportunity, then buying and setting stops and targets, would not be such a far stretch. Or setting up a bot to trade just one stock or maybe a pair of inverse ETFs, one bull and one bear. Lots of things you could do with python code and your brokerage's API. The thing is, when your computer is doing the trading, you could be open to exploit, even if it is just the fact that your trades would be very predictable by a sophisticated bot after a while. Crazy things could happen without your direct observation and intervention. Just sayin. A screener bot would be a great thing, though. And you could put together very complex order chains on the client side that fire on specific conditions, with a lot more detail than perhaps your brokerage offers on the server side. I would start with a couple of basic Python books and vids. Then sign up for a free Alpaca account once you are pretty good at coding simple financial calculators or database/spreadsheet apps. And do yourself a favor. Do all this on a Linux machine. You don't need anything that MS has to offer. You don't need to write fancy GUI apps. Just regular python in a terminal session is all you need. You CAN write GUI apps, and I have, but you don't need to, and you don't need MS all up in your business with their secret source OS and apps.
Plenty of info about doing standard procedures. And all the apis platforms etc are useless without strategy. If you plan to start research then it could take years. If you already have strategy's then it might be hard to automate but better than nothing. If not then easyer might be to look for few great algotraders signals , and pay monthly fees.. Perhaps devs who have working strategy but are at stage where they still need others capital. I cant reccomend any. But seemed to be few decent on collective2.com That site may not get many signal providers as the fees are crazy IMO. Something like 40%. Might scare some people off , including myself.
Alpaca sucks its for amateurs, your gonna get smoked. Slow ass http based protocol designed for crypto idiots who think rest apis are the beezkneeze
Same thing , if your gonna be trading with an API it needs to be a binary protocol or at the very least FIX. Websockets are another clueless millennial bitcoin scam related thing they are trying to foist on people as if its something legit
I disagree, only need highly efficient protocols if doing latency stuff. Most retail cannot compete in that area anyway.