Reading and processing business and market news consumes a significant portion of a financial or equity analyst's work day. Even senior investment and hedge fund managers spend considerable amount of time each week reading news and analyst reports. These news play a critical role in daily investment decisions. The amount of news being published online is on the rise. Processing vast amounts of financial news stories manually continues to be challenging. Differentiating between credible news, rumors and noise poses a big challenge for analysts and is time consuming. Each analyst has their own areas of interest and searching for relevant news of interest wastes too much time. Another challenge is the dependence on only the major news sources, which sometimes doesn't catch market moving events that occasionally get published in less popular sites and blogs. Most tools do not take advantage of advanced computing algorithms and technologies to help in the analysis part, i.e. extract information and visualize them from vast amounts of unstructured news text automatically. The news analytics tool should also make it easy for analysts to build reports within the tool which could help in financial modeling that are typically done in MS Excel. Below are some of the features of a next generation news analytics tool that will help financial firms save significant amount of time and money in terms of news reading and processing time and help allocate more resource time into analysis.
News should be made available as soon as they are published. Many news feed do a great job at this and the tools should catch up.
The next generation news analytics tool must be able to query through vast amounts of news stories (historical and current) quickly, which in turn makes analysis less time consuming and more fun.
A top-of-the-line business or market analysis tool should be able consume news stories and news analysis from various types of news sources including the most credible as well as the less popular news channels.
This is not an easy job for most computing algorithms. But recent advanced in Big Data and Machine Learning has made this possible and tools needs to be able to take advantage of these technologies. Noisy news stories can form a significant portion of the daily news stream and weeding them out will prevent a lot of distractions.
A modern tool for financial or equity analysts must be able to identify news that are already old. This is particularly true when it uses feeds from thousands of news sources around the world. A news that was published in Bloomberg yesterday at 11 AM EST could show in other news sites in the East by tomorrow. The system should be able to identify this and show only the truly new news. The system should also be able to identify any news facts in subsequent news about the same event. Such redundant news could also be used to rank relevance of a news story. This in turn could be used for weekly and monthly summarization of news.
A modern tool should learn from historical events and suggest actions or predict future events based on the knowledge gathered. This is similar to how humans work. The experienced analysts have built a huge knowledge base in their brain, which they use to analyze latest events. The best analysts and managers are able to use most of that knowledge base which helps in making better investment decisions. A good computer program will be able to use 100% of the knowledge base every time to every news event and could potentially highlight several path of action.
A smart tool should be able to distinguish between actionable events and the rest of the news. It should highlight the most interesting news stories while keeping the rest of the news in the back pages, which are made available on demand.
A news analytics tool should be easy to use with little or no training. Analysts and managers should be able to use the tool seamlessly. The user friendliness includes easy setup, intuitive interfaces, self-help and build-your-own-report capabilities.
Most news analysis tools have lacked visualization capabilities. One common reason is because they deal with textual data. It does not have to be that way. News data, with the right extraction techniques should be quantified. This in turn makes visualization possible. Sentiment trends from the qualitative news, when visualized, provides interesting insights, which when tracked over a period of time, could provide significant investment opportunities.
Analysts should be able to build their own news streams, reports and dashboards. The system should be able to automatically learn each analysts specific needs at a given time and shows news of relevance. Building highly custom alert should be easy and quick.
Many information embedded inside news stories are qualitative. They are difficult to represent in numbers. An example of such information is "the Black Friday sales are expected to be really good this year" or "the CEO of the company is expected to step down". The "really good" is the qualitative data that needs to be quantified. The most advanced analytics tools should be able to convert qualitative data into a number which in turn can be trended and visualized.
No modern analysis tool will be complete without predictive capabilities. The predictive functionality applies the knowledge built from past events to the latest news and provides several possible outcomes. A good tool will suggest various course of actions and their corresponding probabilities.
End users should be able to search through millions of news stories instantaneously and be able to quickly build analysis dashboard to help them with their most current investigation. Users should be able to quickly build real-time email alerts around specific events that could be easily discarded once the current investigation is complete.
Analysts should be able to build and customize reports that satisfies their specific needs at specific times. Reports could be as simple as a news stream on a specific topic or a dashboard that shows streams, sentiments, trends, earnings, macro numbers, etc.
Any News Analytics tool should be capable of processing more news faster. A good tool should be able to extract the news metadata on the fly. It should be able to connect the dots using automatically generated knowledge base. And it should provide visualization capabilities.
Hedge funds, institutional investors, mutual fund managers, financial and equity analysts will benefit tremendously from a tool that can do most of the above things reliably and in a secured manner. Such tools could also be used by market regulators. Corporate analysts could use them to monitor their own company news and that of their competitors.
Why are such tools rare to find, particularly with the advancements in Big Data and other related technologies? Or are they used only secretively by corporations as a competitive advantage? Whatever the reason, it's about time to have such a tool available publicly that is affordable as well. News Informatics provides all of the above features with strong emphasis on ease-of-use.