If a mechanism for accurately correlating media sentiment with share and stock prices could be formulated then you can safely say that this sector of research would be many times larger and I would drive something other than a middle of the road family saloon! But just occasionally comparing the two can make for an interesting insight into whats really going on.
This morning the major UK corparate news has regarded Northern Rock, who have been somewhat unwilling dragged into the debacle, and fast becoming a crisis, that is the sub-prime credit crunch. The graph above sets out a measure of the volume of UK press exposure compared to their share price since the start of the year.
Their share price has dropped by over a half since the start of the year and now sits the south-side of 500p. When there is a single issue dominating coverage the relationship starts to stand-out and although this is only volume it is a telling comparison.
So what is being compared in this graph? Share and stock price is a measure of the financial worth of a company now and into the future by a group of highly experienced analysts and market makers whose only job is to do that. The volume of media coverage presents a picture of media focus on an issue or brand. The media will respond to any changes they feel notable and which would be of interest to their readers. So, the question is does one indicator lead the other? I am honestly not sure but an interesting alternative might be a measure of media favourability towards the organisation…that could be telling!
interesting study.. but.. the concept has been proven and a software developed with.. surprise..surprise.. exactly the same name: mediasentiment.com
btw, media sentiment is a registered trademark of Media Sentiment, Inc.
I appreciate your comments and I hope I did not infringe on any territory through my mentioning of the term ‘media sentiment’. I looked at your site and found it informative and interesting as a method of correlating media output with stock movement, via what I presume is a process of automated semiotic coding of text. I understand a number of media analysis organisations include this feature, however on this issue I agree with Kimberly Neuendorf from Cleveland State University and Dr Jim Macnamara, who beleive that these systems are a ‘chimera’ and that the human contribution to content analysis is paramount.
I do appreciate your thoughts as well. We have a patent-pending for our technology, in addition to owning the registered trademark on “media sentiment”. Point being, we don’t do what you think other organizations might be doing. Media Analysis, as a topic has been around for awhile but that is not to be confused with the notion that we have created of “media sentiment”. In our system, “media sentiment” has a “forecasting” component – as opposed, to simply analyzing text. Media Sentiment ® is determined ‘
‘before’ the stock trades/moves. Previous ‘academic’ work has been looking at correlations of ‘past data’. Not easy to grasp, but they are totally different concepts.
Btw, we have posted the ability to look into previous potential results – if you would like to check those they are available next to the graphs on the home page at http://www.mediasentiment.com.
Once again, keep in mind that you are looking at two pieces of data determined at different times: “media sentiment”, which is determined “before” the trading session, and stock price moves – taken ‘after’ the trading session.
I hope this helps to clarify the main difference between the new “media sentiment” concept and the old “media analysis” concept.
Markets dramatically change they their flow and no one exactly produce or expect 100% from their profits . Choosing the right indicator or analyzing software blended with human aspect gives better results.
regards:
nifty tips