Are there cultural hurdles to reducing financial modelling risk?

By Anna Kim on June 3 2009

My recent experience of financial modelling in Korea

Our increasing exposure to the Korean market has made me think about how a corporate culture can encourage or discourage certain aspects of Best Practice Modelling. A recent assignment gave me the opportunity to work closely with a client in Seoul. Whilst there I was able to experience first hand the corporate culture and how certain aspects could adversely affect the standard of financial modelling. Certain aspects such as peer review can be regarded as a lack of trust. Challenges like this regardless of country the team and/or company culture can have a big impact on the promotion of new concepts, such as good modelling and quality control. So how can a good team culture improve good financial modelling? Well there are a few angles worth exploring

Transparency and User-Friendliness

Without an audience to criticise a model, it is often not known that the model needs to be improved. The best audience to test the usability of a model would be your own team where information in a model can be contained within the confines of the company. Building a transparent model means that the model is kept very simple, logical and more importantly time is not wasted trying to understand a model, but is spent analysing it.

Avoiding Complex Formulas

Sometimes we are tempted to solve big modelling problems in with big, complex formulas. With the model restricted to be viewed by only 1 or 2 people, it means that some problem solving issues cannot be solved with the best solution. Compare this with a team who actively promotes best practice modelling, and encourages learning and knowledge sharing amongst the team. These teams will be more likely to find ways to solve a big modelling problem without the added complexity.

Peer Review

Having a model peer reviewed will not only help identify any material errors in the model, but will also help save the cost of a formal model audit, by identifying the errors first. A model should always be tested to ensure that there are no fundamental errors, and that the model is behaving correctly for sensitivity analysis. For those companies that may not need a formal audit, having another team member review a model with a fresh pair of eyes will be sure to find some errors that would otherwise go amiss. A team that encourages knowledge sharing does not necessarily mean that the confidentiality standards will be compromised, provided that the team culture is promoted universally it will ensure that everyone has the same objectives.

In conclusion, share your model, be open to new ideas and try and solve the biggest problems with the simplest approach!