EPTSo my advice would be to just open up a programming language like Julia and start doing some financial modeling. Do some basic stochastic differential equation models and couple them with deep nets. Just look for trends that are "very likely to earn money": subsector X rises the day after subsector Y 60% of the time, so buy Y when X goes up. Look for trends which use higher order moments like volatility and skew, and make adjustments on the fly given current events. And once you know a good trick, keep it secret since once others know the market will correct itself. Of course, one trick isn't reliable, so then you search for more.
The Investening - Page 2
| Blogs > Trainrunnef |
|
rackdude
United States882 Posts
So my advice would be to just open up a programming language like Julia and start doing some financial modeling. Do some basic stochastic differential equation models and couple them with deep nets. Just look for trends that are "very likely to earn money": subsector X rises the day after subsector Y 60% of the time, so buy Y when X goes up. Look for trends which use higher order moments like volatility and skew, and make adjustments on the fly given current events. And once you know a good trick, keep it secret since once others know the market will correct itself. Of course, one trick isn't reliable, so then you search for more. | ||
|
Trainrunnef
United States599 Posts
On July 16 2017 02:05 PaqMan wrote: Are you using a specific strategy to filter out stocks? I can vouch for ticklishmusic's recommendation about Robinhood. No trade commissions and the app is free, but one of the downsides is it can take 2-3 days to move money in and out of your account. Once you have the deposited funds, trades are executed immediately. It's all done on a smartphone however, so I use a separate app to organize and keep track of my stocks. One of the screeners that I use picks up on high rates of change (either positive or negative), and correlates that to a scheduled announcement, then i filter by price. so KNSC had recently reported its Q1 earnings when it came on my radar and the price had changed from .0004 to .0008 the same day and to .0011 before the end of the week. had i been able to buy when I wanted to I would have made a good return on that but c'est la vie. MYE + Show Spoiler + Myers Industries, Inc. is an international manufacturing and distribution company. The Company operates through two segments: Material Handling and Distribution. As of December 31, 2016, the Company operated 15 manufacturing facilities, 20 sales offices, four distribution centers and three distribution branches located throughout North, Central and South America. As of December 31, 2016, the Company had approximately 15,000 manufactured products and over 13,500 distributed products. The Material Handling segment designs, manufactures and markets a range of plastic and metal products. The Distribution Segment is engaged in the distribution of tools, equipment and supplies used for tire, wheel and under vehicle service on passenger, heavy truck and off-road vehicles, and the manufacturing of tire repair materials and custom rubber products. The product line includes categories, such as tire valves and accessories, and lifts and alignment equipment. | ||
|
Trainrunnef
United States599 Posts
On July 16 2017 13:36 rackdude wrote: Most financial quants repeat the same thing: you cannot make money off of information that people already know. And that has a theoretical basis too: there is only arbitrage if there's asymmetric information. So my advice would be to just open up a programming language like Julia and start doing some financial modeling. Do some basic stochastic differential equation models and couple them with deep nets. Just look for trends that are "very likely to earn money": subsector X rises the day after subsector Y 60% of the time, so buy Y when X goes up. Look for trends which use higher order moments like volatility and skew, and make adjustments on the fly given current events. And once you know a good trick, keep it secret since once others know the market will correct itself. Of course, one trick isn't reliable, so then you search for more. I totally agree with that, but rather than spend hours trying to figure out those relationships (I may end up doing that down the line) I would rather play the game with information that no one has. like my position on Myers.. they could outperform, but their track record says they likely wont and the price will dip because of that. And that's where I will try and capitalize. Or like GLLK which is a total hit or miss, but because the price is so low the upside is basically infinite while the downside doesn't hurt much. | ||
|
Trainrunnef
United States599 Posts
![[image loading]](http://i.imgur.com/MXSrqkj.png) Cellectis: need Clinical trial reports in order to really move, but the investor sentiment has been mainly positive its been hovering at around a 12% gain since i bought it last week Kenergy: Waiting on 2Q earning which are likely out in October, just waiting game till then, unless some other news or press release comes out (maybe the grand opening of the third location) Endonovo: Hasn't updated the investors since June 29 (not too long ago) Waiting on pre-clinical trial results. were due not too long ago, but they had to extend the timeline and havent given a definite submission date GLLK: Waiting on results from exploratory digging. Hopefully they have enough cash on hand to at least finish that. It would provide a good boost to the value of the stock if there was at least some sign of gold in the holdings. | ||
|
Trainrunnef
United States599 Posts
FML: I wish i had more money to play with. I took a look at Opiant Pharmaceuticals OPNT, a week ago when they released some information about a new patent they were trying to develop, but I didn't have the cash floating around to use, and now its up almost 70% from that day. | ||
| ||
