In part, this is the fault of program developers who have, by design or otherwise, built software, which is opaque - "black boxes" that don't permit users to examine the assumptions and choices that drive their engines. But mostly, it's our own fault. Even when we can "look under the hood", we usually don't. And why is this? Are we too stupid to do so? I don't think so. Most financial planners are more than ordinarily bright. Are we too lazy? Well, that's true to some extent, but I believe that the main reason why we don't scrutinize how our software tools do what they do is that, like our clients, we're simply awed by the darned things. They're so incredibly strong, they handle so much detail, and produce results with such precision, that we're predisposed to believe that those results "must be right".

And it's that precision, I think, which has lulled us into such acceptance. Working with numbers as we do, we planners believe - on a gut level - that precision is preferable to imprecision, that 7.45% is a "better" number than "roughly seven and a half percent". The problem with that notion is that we're confusing precision with accuracy. A number can be both precise and dead wrong. Moreover, precise is not necessarily "good". If "truth conditions" are not known to a high degree of confidence, if we can do no more than estimate a value, then doing so to three decimal places isn't good; it's bad. Because it's misleading. It implies that we know more than we do.

Where this mistaken confidence becomes downright dangerous is when we accept, at face value, the numbers disgorged from a financial planning software program and do not - or cannot - view them in the light of how much confidence they deserve. If, for example, our planning software asks us to enter, for a non-qualified investment holding, a percentage return for "income" and another for "growth", and assumes that the former will be Ordinary Income realized each year and the latter will be Capital Gain realized only when the position is liquidated, the projected future value of that investment will be hugely wrong, even if our estimates for both types of return turn out to be dead right - because that's not how distributions occur or are taxed. We can improve the reliability of our projection somewhat by "fudging" our inputs, but not unless we are aware that the program will otherwise assume that all capital gains in that investment will be tax-deferred until liquidation.

We have to know what the program is doing in order to reduce the impact of what it's doing wrong.

But even if our software were to model everything with complete accuracy (as if that were possible), and even if all our guesses were right, we'd still - most of us, anyway - have a problem. We're still seduced by the Dark Side of that technological force. Because financial planning, for the most part, isn't about the numbers - however "accurate" they might be.

Financial planning, in my opinion, is 90% emotional. Only about 10% is about "numbers". When we model future cash flows, we're dealing with whether our clients will be able to live the lives they want to live. A "hypothetical probate" in an estate plan isn't so much about "transfer costs" as it is about the legacy our clients will leave to those they love. And neither is simply a matter of numbers.

The biggest problem with the incredibly powerful tools we employ, in my judgment, is that they have allowed us to form an intimate relationship with our client's data, when what we need to do is form an intimate relationship with our clients.