Last week, the UK’s Transport Select Committee of MPs said that the bid by Virgin and Stagecoach to run the East Coast rail franchise was “wholly inadequate and naïve”. The Department of Transport also came in for criticism for not managing the bid effectively.
Of course, this is not the first time that there have been issues with major transportation bids. Remember the West Coast mainline where the franchise had to be rebid at a cost of over £40 million to the taxpayer?
However, rather than laying the blame at the door of specific institutions, perhaps we should look at the system a little more closely. There will be a multitude of reasons for the failure of this franchise, but three aspects of financial modelling may have contributed:
1. Poor assumptions
According to press reports, it was estimated that revenues would grow by 10% per annum. In reality the increase was 3% per annum. What was the reasoning behind this optimistic assumption? Was the assumption ever robustly challenged? What other assumptions were used and how realistic were they? Please see my previous blog on projections.
2. Insufficient sensitivity analysis
There is generally a lack of robust sensitivity analysis in models. I always recommend calculating a base case and then a range of possibilities, including combination downsides, in order to analyse the risks associated with any deal. I think that the two main reasons for a lack of analysis are the reliance on iterative macros (which makes running sensitivities time consuming) and a limited understanding of data tables. Please see my previous blog on the use of iterative macros.
3. Lack of training
Many companies and public sector bodies are reluctant to invest in proper training. However, financial modelling is complex and multi-layered. Therefore, poorly trained modellers can easily create incomprehensible black boxes. Please make sure that those given the responsibility for building models know what they are doing. Online or large training programmes are no substitute for bespoke in-person training and can result in expensive mistakes. Please see my previous blog on training.
Ultimately the purpose of a financial model is to assist us in risk analysis and decision making but remember that they can never exactly predict the future. As statistician George Box said: “All models are wrong, but some are useful”.