Inflation rates in many countries have not been as high as they are today for many years. Problems indeed for companies and policy makers. But what about for financial modellers?
Inflation is one of those areas in a financial model which has a high instance of errors.
Firstly, let’s be clear about the terms real (without inflation) and nominal (with inflation).
And here are three potential problems commonly encountered whilst modelling inflation:
- When does inflation begin in a model?
Always check the starting point for inflation and don’t just assume that it begins at the start point of the model. The inflation start point for different cost and revenue items may well vary (for example, a specific inflation start date may be defined in a contract, or assumed in a cost model).
- Which inflation index should be used?
Always consider which inflation index to apply; this will be dependent on the nature (and assumed inflation rate) of each cost and revenue item. For example, should the index be start-period or mid-period? And how often do the prices go up? For example, is it annually or semi-annually or perhaps is there a price increase every three years?
If a model simply has a line labelled “inflation index”, this doesn’t give enough information and could easily be incorrect. I have seen several models which have overstated the revenues by a whole year’s inflation, creating a material error.
- Don’t forget that inflation has a compounding effect.
A semi-annual inflation rate is not the annual rate divided by two. This is simply incorrect, because of inflation’s compounding nature. The formula for the semi-annual inflation rate is as follows:
semi-annual inflation rate = ( 1 + annual inflation rate ) ^ 0.5 – 1
As ever in financial modelling, clarity and precision matters. Errors in the calculation of inflation can make material differences to the results.