With the immense volume and breadth of new information available to you at any given moment, you've no doubt approached it with some sort of customization.

This can be something low tech or automatic, such as reading a local paper (customized to your locality) or only reading certain sections of the paper that you know might have interesting articles (customized by interest).  For example, I always skip the 'Arts and Entertainment' section because of its low probability of having something I care about.  My skipping of the A&E section doesn't mean I have no interests in the arts, just that I have a narrow focus that isn't often served by what is typically printed.  By customizing my consumption in this way I've accepted a loss of information in favor time efficiency.  

Almost all forms of customization that currently exist are of this form.  This categorization and paring we do is ubiquitous.  We filter news by our portfolio, we select which sites to check for information, we select which RSS feeds we want to follow.  This customization frequently ends up being at odds with discovery.  I have a near zero chance of discovering something I would be interested in outside the small slice I've chosen to view.

What about discovery?  Sites such as Digg, Reddit, or Hacker News let groups work together to discover items of interest and share those items with their user-bases.  This is a very blunt tool that effectively offer no additional customization beyond the same type as above.  You can discover information from a wider source-base, but if categorization is still present it is done either by selecting a category (Digg or Reddit) or implicit to the community (Hacker News).  You end up with a higher success rate of interesting information, but you still have to manually weed through it.

Going forward we need to focus on combining customization and discovery.  There is some interesting work in this direction, primarily using machine learning to monitor what you look at for themes and patterns.  These systems will then use those patterns as a directive for what information that best fits your tastes.  This combination could be called a customized-discovery system.  It focuses on discovering items that already fit your consumption patterns.  There is another combination that shouldn't be ignored, however.  A discovery system needs to be able to find items of interest that either are only obliquely related to your normal patterns, or completely outside of them.  

What would a good hybrid customization and discovery system look in the financial news space?  Customization would come first, by including news immediately relevant to you by portfolio or investment pattern.  The first stage of discovery would be to look for items related or peripheral to your portfolio: partners, competitors, suppliers, etc.  Next it would need to look for either generally interesting or uniquely interesting information outside of your normal patterns.  This would all still need to be tempered -- all related and interesting information would likely already be too much.  The last discovery stage would be a discovery filter on top of the news already selected as relevant to you.  This would rank and prioritize the information for you so that you can linearly approach the content.  You start with the most relevant, important information and progress until you feel you've gone as deep as you need to.  Ideally it would then have a method of aggregation and summarization for the rest so you can leave off where you want but feel comfortable you are still informed... but that's for another post.