The debacle over the West Coast rail franchise shows the importance of financial models when making big decisions. Formulas need to be reviewed on a regular basis, but this rarely happens in the public sector
I dropped a glass vase onto our kitchen floor recently. Not a very expensive vase but, it subsequently transpired, one of great sentimental value to my wife as it was wedding present from a now deceased aunt. It took about a second from leaving my soapy hands to hitting the floor and in that “Oh sh*t moment” my life has been slightly changed forever.
Predictably, my wife is less than happy with me, the remains of Aunt Brenda’s fragile heirloom are now in a landfill site and I have been permanently relegated from lead dishwasher to the lesser role of drying up and putting away. Had I used that one-second moment to catch the vase, none of these things would have happened.
I was reminded of how our lives can permanently change ever so slightly on the smallest incident when reading of the West Coast franchise debacle. My understanding of this is that despite the froth and lather of political point scoring, the heart of the issue is an error in the financial model used to evaluate the bids.
One tiny error in a vast model full of complex formulas. A simple mistake in a formula that took no more than seconds to write has gone on to cost the UK taxpayer £100m+ in bid and reprocurement costs and probably terminated the careers of a minister and at least three civil servants.
For those of you who use spreadsheets on a daily basis, this will not have come as much of a surprise. Excel and other spreadsheet tools are deceptively fragile and error prone. Formulas are easy to write and take at most minutes if not seconds to do. Before long you have written hundreds of them all linked together and have created something you proudly refer to as a ‘model’. You might use such a model to make management decisions including the award of multibillion pound rail franchises.
This is a dangerous delusion. My son Marcus can pile toy bricks into little towers, but that doesn’t mean we are going to trust him to build our extension. He is not a bricklayer, he is a two-year-old boy. Just because you can write a "=Sumif" formula, doesn’t make you a trained computer programmer, and any models you build are likely to be unstable and full of errors.
The financial sector has long been aware of the brittle nature of models and, as matter of course, any model is subject to an independent external review. This has created a small industry in London and India of people spending their days crawling through other people’s carefully constructed formula spotting mistakes. It has also given us a valuable insight into how frequent errors are.
My firm alone reviews hundreds of models per annum and typically we find 30 to 50 errors in each one. Some of these are of no consequence but others cascade all the way through to the bottom line and mean that the model is fundamentally wrong. It also telling that of the hundreds of models we and our rivals review each year, hardly any come from the public sector. Now, either the public sector is not using spreadsheets very much or, more plausibly, it is complacent as to the level of errors they contain.
Now at the risk of self aggrandisement, this should be your ‘oh sh*t’ moment. From the last paragraph onwards, you now know that you going to make management decisions in the future that rely on unreviewed models that are full of errors. This means you are going to make wrong decisions. They may not be wrong by very much but they will be wrong.
However, unlike my slippery vase, you still have the chance to change the future. Go into the office tomorrow and get every single model your organisation relies upon reviewed. This will take time. It will cost you money and it will be painful, as the process will uncover hundreds of errors in models that you may been relying upon for years.
Now I am not so self-deluded as to think my little homily on the PF Blog will make you do anything. However, I do also implore you to consider why you are not doing anything. The investment banks routinely get models reviewed because they know that they contain errors and they don’t want these little time bombs, at some future date, to blow up in their faces.
You won’t get your models reviewed because……? Because what? Because you think your model builders are more technically able than the highly paid Oxbridge maths graduates employed by the banks? Let’s face it, that’s not very plausible. Your builders may be nicer people and have better social skills, but they are unlikely to be technically perfect. They are human and prone to making mistakes and, if you blithely ignore this fact, you are, every single day, adding to the stockpile of little errors that may one day detonate and sink your career.
So, in conclusion, my vase dropping incident has shown me that our lives hinge on tiny moments, mere fragments in time that have the potential to change everything. If we could replay our lives again, I am sure the Department of Transport would have had their model reviewed. I am sure the model builder would have spent a few more seconds writing that simple formula and I obviously would have not washed up my wife’s surprisingly fragile vase.
None of us can do this because these events are in the past and the consequences have already manifested themselves. The errors lurking undetected in your financial models were also written in the past, but you still have the opportunity to change your future. Your vase/career has yet to shatter on the unyielding floor of illogical formula, so I implore you to catch it before it’s too late.
To err is human. To know that other people err and to ignore the consequences of their erring is less forgivable.
Michael Ware is corporate finance partner at BDO
