I will admit it. I am one of a minority of people in the world who really loves math. To me, it is a lot like programming language but even more of “pure.” If you can describe something in an equation, you get a level of certainty that is just not possible in any other language – or so it may seem…
After much review, it is my conclusion that the present economic crisis, the worst in over 70 years, was caused primarily by a single bad math equation. Hard to believe? Read on.
It is hard to phantom just how math could cause such a catastrophe but, when you think about it, bad math often causes trouble. Remember all of those movies about bridges from high school showing what happens when designers didn’t calculate resonant frequencies properly. How about the structural collapses that happened when someone miscalculated the possible force on a structure?
OK, suffice to say that the problem is almost never really the “math” itself, but how it is applied to reach a conclusion. Just like guns don’t kill people, math doesn’t either. Be warned, however, bad math is just as dangerous in the wrong hands as a loaded gun. The latest economic times should serve as a testament to that statement.
By now, you may be a bit confused so let me give you some context. First, the best article describing what I am talking about (by a real writer with a lot more detail) is from a recent Wired Magazine by Felix Salmon - http://www.wired.com/techbiz/it/magazine/17-03/wp_quant. Check it out if you need more insight, then come back here….
Simply put, when companies looked to buy collateralized debt obligations (CDO’s), or insure mortgages, or rate investments, they all wanted to have a formula to evaluate risk and determine value. I know I would love a formula that would just tell me if a new product will be successful but I also know enough to know that no such thing exists.
Unfortunately, everyone glommed on to a particular formula, and that was just plain wrong. You can see the results. In this case, the formula failed to account for what could and would happen if the price of housing was to fall nationally.
So, in addition to offering interesting insight on the present situation, there is tremendous applicability to what all of us do every day when we look at data and information, and try to turn it into insight and knowledge for ourselves and our organizations.
It is the role of any (good) manager to look beyond conclusions or the results coming from some analysis or formula and understand the assumptions and basis.
The Rules for me are simple:
§ Understand and validate the source of all data and information
§ Understand and validate the ASSUMPTIONS used to create the conclusions or results
§ Understand and validate the rationale and reasoning behind any formula used to create an assessment or design assumption
There will always be bad data out there. In fact, the web is a shining example of that. There will also be bad math. The problem arises when managers blindly follow conclusions based on data that they haven’t validated and formulas that they don’t understand.