Are Algorithms Necessary?

Ours is an economy that eludes the signs of recovery in favor of change. In fueling that change much has been hailed in the past few years regarding the need for skilled workers, which I see as code (pun intended) for digital programmers. While there are ample views as to the reason for the shortage in qualified programmers the need itself is the important consideration and in my view comes down to a simple notion, that the shortage of programmers steeped in knowledge and skill within the language of algorithmic theory (not all are by the way) is in demand by the companies that have one equally simple goal in their collective sights, which is to know everything there is to know about everybody in the world. Neat trick if they can pull it off.

Well, I'm not convinced they can, but that doesn't mean I don’t see their value. Since the first mortgage backed bond products were created, the strategy to market them has been in high gear, at least until 2008. The idea behind the product was easy enough to understand; mortgages issued by dozens of banks were bundled together and identified by their average interest rate and average maturity (i.e. 30yr, 20yr mortgages etc.). The financial industry was into something good and the value of the investment was tied to the due-diligence banks were doing to ensure high probability that loans would be paid back.

So why talk about this now? The nagging problem for those marketing these assets was that they were fundamentally unstable. Think about it, how many home owners generally stay in their homes for the duration of their mortgages? That’s a question that financial firms attempted to answer through use of algorithmic methodology. The idea was to take all the data available about the buying and selling of homes, the rates, the locations, net worth, and whatever else was available fed into a concept designed to deliver a probability of occurrence type of outcome. If a company could tell the buyer of a mortgage backed security that its price and stability could be monitored with digital expertise the impression wasn't lost on the buyers of millions of mortgages sold over the years. Did it work? Well if the financial crisis was any indication, not quite as expected. Once mortgages were being written to people with less robust credit histories brought into the process through schemes such as teaser rates and adjustable conditions the ability to gauge the predictability of the outcome of homeowner behavior was once again unstable.

Now flash forward and we find ourselves in a world where cleverly structured and dazzlingly executed algorithms are developed to predict the searching results for a population eager for information. And as the use of that information grew to meet the demands of the everyday non-digitally minded consumer, finding trends, window shopping, buying and nearly every other material activity was being done on the web. And every time they 're on the web more and more information is fed into the algorithm to a point where the confidence of the results are so great that advertisers and corporations hungry for marketing access eat up every promise made of the results.

The companies that constitute both the tech industry and the pipeline of well funded ideas waiting for their piece of the action are considered good resources for information about human behavior. However, one can mine the depth of characteristics of behavior, but like the psychology that dominates the activities in the capital markets, when biology belly flops into the pool of digital wisdom the results are not always what are expected. But if not informative, still probably good for consumers and definitely good for the businesses that are navigating the new selling strategies targeting many of those aforementioned characteristics, one example is Amazon (AMZN) another is Google (GOOG). Probability based systems have been around as long as the industrial revolution demanded them so for now that makes the use of algorithms necessary enough. However, just as the financial crisis in mortgages left those companies holding the bag so too could some of our most beloved tech companies depending disproportionately on data for sale such as Facebook (FB) and Twitter (TWTR) find themselves the middlemen they've worked so hard to displace, especially when their clients fancy turns to fickle the most unpredictable consumer behavior in the book.