Getting Access to Spreadsheet Models for Managers

Spreadsheet Models for Managers

Crowding is usually counter-effective as a business practice, and not recommended. But there are nevertheless times when you want to build it into a model. If you do, you ought to do it right, so it's included in this course. But we hope that you recognize that any business that deliberately plans shortages of resources is — most likely — in serious trouble. Or headed for it.

Still, if you want to build crowding into a model, there are right ways to do it — and wrong ways. By analogy with parametric cushioning, there is also parametric crowding (both dangerous approaches). In parametric cushioning, you may recall, we structure the cushion into the value of the parameter. In parametric crowding, we structure the crowding into the value of the parameter.

Such a practice is dangerous. Suppose, for example, that we decded that interest rates would likely average around 3.7% over the time span of the model we're working on. We might "crowd" this parameter by assuming that the rates we would pay would be a tad smaller; say, 3.5%, which would decrease the pressure on our cash flow. We could accomplish this by wiring that lower rate into the model. But then later, if we decide we want to change the way we crowd interest rates, we would have to remember or look up the nominal rate before making any adjustment. And whenever you have to remember or look something up, there's a chance you'll get it wrong.

A much safer approach to including crowding in this example uses a separate crowding parameter. We would have a parameter for the nominal (expected) interest rate, and another parameter that would govern the way we computed the crowded rate from the nominal rate. The crowded rate would then drive the moodel. In that approach we spearate the nominal rate and the crowding and we can control the two independently and explicitly.