I recently heard Dan Ariely, Professor of Psychology and Behavioural Economics from Duke University, discussing a conversation he had had with a locksmith. When the locksmith had just begun his trade and took a long time to open locks, clients were impressed and happy to pay him a handsome sum. However, when he became more efficient and came up with quick solutions, clients had a greater tendency to quibble with the fee.
It seems to be a natural human trait to be impressed with complexity and the time expended on a task. I find this applies to financial models too. People often assume that the more incomprehensible a model, the more advanced the modeller. It is even more impressive if the model needs to run by pressing various buttons or has formulae several lines long. We often assume that if we can’t understand something, then the person who has produced it must be a genius.
But should models really be that complicated? Surely a good model is one that can be easily understood and can produce instant results. However, I can sense people reading this thinking “yes…. but the models we produce are sophisticated”. Of course models are multi-layered, but the underlying logic behind everything should always be clear and easy to follow.
A frequent comment on my training courses when we create certain formulae is “is that it?”. It seems almost disappointing if we pick a suitable function and use simple logic to arrive at the correct answer. The other day I came across a formula with 62 sets of brackets, 16 functions and 60 cells (and I am not exaggerating). I viewed this as a challenge (this is what I like to do in my spare time) and reduced it to one with 2 sets of brackets, 1 function and 7 cells. This new formula was achieved without any additional lines. The solution was not complex, it was just a question of understanding what was required and using the correct function.
So how can we make financial models fitter for purpose? I have a long list of recommendations. However, in particular there are two standard practices which I think often cause unnecessary complexity. These are:
IF Functions
There is massive over-reliance on the IF function. It is not necessarily incorrect to use it, but it is usually far better instead to use MAX, MIN, CHOOSE, INDEX or logical functions, depending of course on the circumstances. Correct function choice makes for more efficient formulae.
Iterative (Copy Paste) Macros
Many financial models include unnecessary iterative (copy paste) macros. In my opinion, these macros increase complexity and also indicate a problem with the logic of the model. A correctly structured model should not need iterative macros.
Financial modelling is not rocket science. It just requires concentration and proper training. Like locksmiths, financial modellers need to find the most efficient route to a successful outcome. And all of us need to see through the complexity and value when a job is done well.
As Leonardo da Vinci said: “simplicity is the ultimate sophistication”.