Data democracy

I have not yet been fully convinced of the direction of infographics until now -- I find too narrow the focus on organizing, structuring and visualizing large datasets; often times, we get pretty pictures with extremely high data-ink ratios but more often than not, these very dense graphics fail to speak directly to readers.  We see a lot of information; we find hardly any insights.

I think I have seen the future.  My friend Adam has been working on a web service called Empirasign, which I will describe as a form of data democracy - he takes boatloads of financial data, runs all sorts of analyses and models, and presents these results in a variety of formats, including on-line reports and tweets.  He does not attempt to visualize all the data, or all possible relationships.  Each analysis or model focuses on specific matters and he presents the result in tables and charts.

For example, a business problem might be as follows (timely for the year-end): in my portfolio, I am carrying some loser stock which I'd like to sell by year end so I can take a tax deduction on the loss, perhaps to cover some investment gains I have realized last year; however, I also believe that the loser stock may be near bottom, and if I sell now, I'd want to buy it back in short order - alas, this may be considered a "wash sale" and prohibited.  What if one can find a hedge (another stock or a portfolio of stocks) that replicates the performance of the loser stock so now I can get the best of both worlds - I sell the loser stock for the tax deduction, but keep the performance by taking a position in the hedge, then unwind when the regulation allows me to buy back into the loser stock?  (If you are interested in this trade, you should consult the experts: Adam's tutorial or wikipedia on "wash sale" or IRS-ese (pdf file).)

There are lots of stocks out there, and lots of possible hedges.  An unsophisticated investor like myself would have to spend a lot of effort to find the right hedge.  Also, it's very unlikely that staring at an infographics chart will uncover such hedges.  What Adam has done is he has collected all the required data and run analyses to find the right hedge for pretty much every (loser) stock out there.  And instead of presenting all the underlying data, he presents the results.  See below.

These data displays are not sexy - and can be improved (the explanation for the columns of the table is found on a separate page, e.g.), but for the target audience looking for trade ideas, they get to the point.  This is the gift of statistical data reduction.

What is also worth noting is through the magic of R, and Web technologies, Adam makes all this run automatically, so the insights from the data are uncovered in real time.  The wash sale avoidance strategy is not the only analysis he provides; there are tons more on the website that implements all sorts of other techniques (of which I am no expert) but it appears that users can pick and choose whatever strategy they like to follow, and Empirasign saves them any of the analytical work.

As I said at the start of this post, I see this as a promising direction for infographics, moving from visualizing data to visualizing insights.

P.S. As with previous years, I have updated my Amazon wish list (click on button on top right).  If you'd like to show your support for this blog, please help me build out my library.  Thanks to those readers who have contributed in past years - since Amazon does not always provide me your contact information, I have not been able to thank each of you personally.  Happy holidays!