Hi Guys, First post so go easy please ;-) I'm an experienced software developer with some time on my hands now and I'm looking to create a free system that will detect stock movements and then check pretty much every news source that can be checked online for news regarding this company. So that's blogs, social media, news websites etc. With the aim of getting this information to the user as quickly as possible. At the moment I'm not sure whats the best way to detect that something might be effecting a stock, I was going to go with just the price of the stock but then I thought some news may be interpreted as good to some and bad to others so the price might not actually change much. Then I thought of checking the volume of the stock over a short interval. I saw some websites do offer this but it appears they don't show "out of hours" trading volume. Is that correct? If not can someone point me in the direction of one that does? What stats would you guys recommend looking at to detect "interest/purchases" and where is the best place to source these stats/figures? Any help at all would be greatly appreciated Thanks!
First Jamie, welcome to the forums. Reuter's, Bloomberg, ADP are several sources of real time market data for price and volume information but access to that data is not cheap. I'm not sure of what you are looking for as an end result that would be any faster then what my broker already gives me within my brokerage account. When I pull up a stock symbol I get fairly quick news results within a minute or 2 of it coming out. Also, traders who want to be quick on the trigger regarding movement are turning to algorithms to avoid having to sit and watch for immediate news or information. I don't mean to discourage but not sure how this might catch on considering the current rapid deployment of news. Maybe others will have a better take.
hello fellow programmer, The easiest thing is getting als the news data, twitter offers a streaming api for free and you can get most news sites as rss feed or whatever, clasification is the hardest thing you need to clasify the news articels as [good|bad] for [company|industry] or whatever, this could probably be done with things like https://www.ibm.com/watson/developercloud/nl-classifier.html given a large enough training set. also try to detect multiple news storeis about the same topic, so now you have a news topic clasified as good or bad + popularity of that topic and then you can try to correlate that to the stock price + volume ( realtive to other companies in the same sector ) Could be really hard tough, news about one company might also affect their competitors and so on.... If you make this you could turn this into an awesome trading bot.
Great observations from someone with a fresh look. I am no help to you as I rarely read the news including earnings reports. I consider them worthless and a waste of time. Simply, what matters to me is ascertaining who owns the stock. If the big interests have accumulated the stock they will make it scarce to the market and price will rise (and all the news will be good news). If the public owns most of the stock then it will be volatile and price will fall regardless of the news. But then I don't day trade either, the time frame where news noise may matter....nahhh.
Welcome to the board @jamie3000. Your venture has merit and could work to some degree. There are some good suggestions that have been offered to you. Tracking the sector, related stocks, and the news can be big factors in price movement. The only problem is that sometimes there is no rhyme or reason for a stock taking a big pop or drop in price. Then later the news comes out after the big boys have moved the price one direction or another. Sound illegal? It should be but the SEC has let this go on for many years. This has been a widely used practice on analyst upgrades and downgrades. Good luck and let us know of your progress...TTT
Hi tuliptrader, thanks for the response. I've used the twitter API many times and I am very familiar with RSS feeds and web scraping/sourcing in general. Apologies I didn't really explain my self very well, I'm not trying to make a bot to trade off news. I've seen sentiment analysis systems before that parse news and social and they're quite unreliable. For example say google is crashing right now hideously someone might sarcastically tweet "Google is a great stock to buy right now...". Us humans are very difficult to understand I'm trying to make a system that allows users to get news from as many sources as possible as quickly as possible, the volume question was so that it would automatically filter news to companies with irregular trading activity to help people get a head start. I can't see a service dedicated solely to pure financial news aggregation from every possible source. Social (linkedin, twitter), news sites, blogs, data leak websites, whistle blowing websites etc. I just want to point out I'm not trying to invent the next big thing and get loaded its more of a personal project/tool. I got ripped to bits on the ET forum for posting the exact same question so thanks for the kind responses
Hehe I've read a few of those stories about illegal trading, must be extremely frustrating for "retail" traders as I believe they're/your're called! Probably feels like they're actually stolen from you I guess.
Interesting, how does one ascertain who owns what share of a stock? I believe it has to be filled if an insider is buying shares right? On the SEC website? I don't think I could algorithmically do anything with that information by itself but very interesting still
Hmm, not sure if this is doable or not, but if you aggregate the news and can determine if the news is good/bad or neutral I wonder if then looking at the money flow for that stock will help add perspective? Money flow looks at price and volume along with direction of the money. It might be interesting to know how the money flow coorelates to the news. Link explaining MoneyFlow: http://www.investopedia.com/terms/m/mfi.asp
