Is there any simple strategy to earn ~10% yearly? I would be happy to outperform the bank/GIC while KISS. I am a well versed coder and reverse engineer. These days I only code serious programs using Erlang and C, in the passed I coded using C#, C, C++, Lua. I prototype using python, I do not use it for serious 24/7, 99.9999% uptime, non crashing, distributed programs. My programming strategy is to connect the IB Gateway and use a python wrapper to interact with it, core in Erlang. The buggy IB Gateway can be deprecated later to trade directly using FIX protocol. I would love to hear from seasoned traders that are having technical troubles to get their strategy working.
Sure you can, with the attendant additional risk required to earn that 9.5% above the risk-free rate. Anyone who tells you otherwise hasn't studies the concept of risk adjusted returns or is selling you a ponzi scheme. There are lots of serious programmers, some of the best in the world in fact, working on coding financial algorithms. It's not simply a matter of putting in the work and being able to program and bam! you're getting a return 2000% better than the risk free rate with the same amount of risk.
Have a look at some of the papers at the top of this list. You sound pretty adept at programming, you could build one of the systems you find appealing and share the results here.
Yea, something to use instead of GIC/Bank which barely/cant beat inflation. I realize this, in fact it is a common argument anywhere. There are lots of serious traders so why trade? There are lots of serious algos that will beat me in the market and take my money. There are lots of serious businessmen so why run a business? It is a good argument, but after reversing the ActiveTick system, and peeking at IB which is a lot more complex to RE due to the cluster f*** of code (GUI + Gateway) mixed all in one. I really think my programming knowledge is stronger than those that made these systems. I see that I can do it better, and if they are making money with those systems, imagine if you can do it a hell of a lot better. Thanks the summary's are really good, but I found the detailed papers really confusing. (day interval) I was hoping for something simple like if S&P is downtrend, CL down, Shanghai Index down, stock closes below the 20 day moving average for first time (being above before), its an indicator to take a short position. To build the system I would like to partner up with a experienced trader, or at least someone that can show some basic strategy. A strategy I came across using google was on warrior trading where it was to scan for gappers (with volume) on the premarket/open (9:25-9:30:00 am), if the stock was trending up before, and it gapped up, its indicator it will go up for at least 5 more minutes before sideways, more up, or crash. I found the strategy kind of worked decently, you get in at 9:30am, expect to get out by 9:33-9:35am and take %-5 to %+20. I used live quotes using ActiveTick (i reversed engineered their client API to connect directly to their servers as their software was crap) which I since canceled as its kinda useless. For example: Look at KEYW, ROWR, ENOC Feb24-25 opening gap. If you bought KEYW at 9:30am after the huge jump up, it kept going up, you made 10% in the first 5 mins. ROWR went sideways, so you would not win/lose, ENOC you would make 20% by the 9th minute. The volume is low, but it looks doable for 10k $ to throw around daily. Another strategy I think can work is getting a live news feed and parsing the data. That would really differentiate the system from the others, as now it can react to live news, and I have not seen any algo systems mentioned here that work purely on news. I am just not sure where to find a realtime news source that doesn't cost a fortune. You can use a NN to parse the text/incentive of the article too
There is a very important distinction between making money running a business and making money trading. It's clear your conflating them with your comment on IB. Yes, their software is horrible and they've got far more than their share of imbeciles working there. But they're not making money taking principle positions trading, they're making money running a business. If you told me that you thought you could make money running a brokerage that competed with IB because they suck and you don't I'd be right with you. Heck, I might even help fund that endeavor! But saying they suck running a busines therefore you can make money trading doesn't follow. All the rather simple strategies you've outlined were coded into HFT algos long ago that have tens of millions in custom programmed hardware that you can't hope to compete with. They arb the crap out of these opportunities in microseconds. I run a startup and I can compete with bigger competitors in a bunch of ways and in a bunch of niches. You can't compete with HFT shops as a lone programmer without resources no matter how good you are, and they're in every niche in the finite space of the market. Read up on HFT, you'll be glad you did.
Interesting, about running the brokerage, I have not ever considered that as it seems so far off yet I am very interested, can you elaborate? From what I read/know about HFT (mostly speakers at DEFCONs), its placing trades in micro/nano seconds, and due to the sheer speed requirement where you need to be first to profit, it seems like order routing or scalping for pennies. But the risk looks really really low at that level, as you are not really trading. Yea I see the comparison is apples to oranges, but from my experience in software+hardware it runs on, the success of a software company is directly linked to its main engineers, even with the best marketting/business model/product design, if the code cannot work the way it needs due to engineers lack of knowledge the company cannot succeed in a reasonable time frame, and if it does it would have trouble adapting to the changing market (adding features). Also regarding the serious algos already running the best strategies. Is it not unreasonable to piggy back off them then? Say you write an algo to detect other algos in the market, follow them by piggy backing.
Sell/write options to collect the premium. -- kind of like selling insurance. Or, you could buy them for much potentially higher returns. -- like everything in life...the devil is in the details though. -- it's not as easy...as Just doing it. i like KISS: keep it simple stupid -- that's my philosophy in trading.
Options are very interesting, in fact that was my first go to. Something involving options, as there is just so much calculation and strategy involved, a human could never beat a computer. But I could not find any good + affordable real time streaming options data.
How much can you afford to risk for making 10%? Regarding data: GBM data (just search here or at google) can be an alternative while developing and testing a system.