An interesting correlation

These past few days I came across an article published by CNET News listing the seemingly unending announcements of layoffs of these past months. One of the questions that occurred to me was whether there is a correlation between the importance of a company's layoffs and its stock's price-to-earnings ratio (P/E). I chose to use the P/E ratio before the beginning of the crisis, because that's when future expectations of growth where factored into the price of a stock and hence the layoffs reflect the fact that these very expectations were disproved.

My take is that our economies have been running into trouble since the second half of 2007 and hence I used the P/E of 2007 for a subset of tech companies of the CNET list. The companies included in the data set are: Adobe, Alcatel-Lucent, AMD, AT&T, Autodesk, BMC Software, Borland Software, Cymer, Dell, Electronic Arts, EMC, Ericsson, Lenovo, Level 3 Communications, Lexmark, Logitech, Microsoft, Midway Games, Motorola, Netflix, Oracle, Plantronics, RealNetworks, Seagate, SGI, Sony, Sun Microsystems, Unisys, Viacom, WebMD, Western Digital, Yahoo

I got the P/E ratios from Prof. Damodaran's web page that I got from Prof. Farber's page of resources at Solvay Business School, my Alma Mater. So I'm once again grateful to professors & academia for sharing knowledge, the only way for knowledge to grow.

While this is by no means enough to draw conclusions and the analysis is only very rough relying on many opinions, the indication is that the higher a company's P/E is in good times, the likelier it is to lay-off personnel when the economy sours . To illustrate this result, I used IBM's Many Eyes visualization resources, which I find prety cool, so thanks Big Blue for making great tools available freely to those who seek, for those who are on their quest, sometimes a business quest :) Here's the link to the chart. I'm also including a screenshot of ManyEyes below because the link doesn't always work properly:

I also used data visualisation software Tableau to verify the correlation: quite a powerful product. Below is the analysis of the model produced by Tableau:

 

And the Tableau visualization is below: