The World Population will be 8,369,094,344 in 2027. Google says so.
Will it really? Why such accuracy and according to which assumptions? And what, perhaps, might be the expected range?
Many newsfeeds and newspapers include similar statements of spurious accuracy without further information on the assumptions used and an overuse of the word “will”.
Does this perhaps lead us into a false sense of our knowledge? Remember the howls of protests in the news when the Covid “models” were wrong? But a model is just that, it is merely a model.
The same goes for financial models. They are useful tools and they predict but they can never show reality.
Project financiers need to remember this too. Some modellers seem convinced that their base cases are true. Really? They then proceed to tell me that they have “debt sized” (using the dreaded copy paste iterative macro) to an exact debt number to have the optimal finance solution. Really?
Telling me that debt of £425,789.62 will give exact cover ratios over a fixed period many years into the future cannot possibly be correct. What assumptions have been used? Will all the working capital items be paid at exactly the number of days that have been predicted? And what about the tax? Will the tax regime remain the same over the next 30 years? And, on the subject of tax there are many models which assume that the taxable profit is the same as the accounting profit, with no assumptions made for more realistic tax allowances.
Surely it is better to consider ranges of assumptions, to analyse results within that range and then make informed decisions.
As statistician David Spiegelhalter recently said, on the BBC’s excellent More or Less podcast, “good modellers don’t believe their models; they merely use them as a guide and should have a humility about knowing that they are wrong.”