My advice to anyone reading this page is to go back and re-read the article AND comments to better appreciate the value of the experience that others have shared here.

To be good at this I think you need to be a boundary spanner – that is, have enough of a technical understanding of the industry you are in as well as construction and project issues. Then you can communicate sensibly with all parties.

I agree with Brett that it is easier to teach an engineer/scientist financial skills than to take someone with a pure finance/accounting background and teach them about industry. There are always exceptions, but as a generalisation I’d go for an engineer most times.

Also, good client skills are important too. After a couple of years you will likely have more project experience than many of your clients. (Remember that clients are in your organisation as much as external to it).

Some other criteria for good modellers:
a) hate mistakes but realise that everyone (including them) is prone to making them …. especially other model users who won’t be as hands-on as the main modeller,
b) understand exactly how the model outputs will be used to inform the decision making process so there is no ambiguity in the outputs,
c) are pedantic about version control,
d) track a log of model changes and value impacts,
e) have an intuitive sense of how the model will react to key assumption changes,
f) understand the model limitations
g) have a detailed understanding of the asset/business they are modelling and are not afraid to ask for more detail/inputs when components are not clear to them
h) understand that systems are often complex and a change to variable x may necessitate a feedback loop change to variable y etc
i) have an appreciation for differentiating between what is critical and needs further work/testing/benchmarking versus being immaterial to a decision
j) are passionate about running to ground any deviations from the expected
k) use inbuilt checks/graphs/history/auditing tools/shadow models to validate their work

As per other comments above my bias is towards engineers who have picked up financial qualifications. In general engineers tend to be very good system thinkers (see point (h) above).

The issue is that someone who ticks the above boxes (financial skills, good communication skills, ability to resolve ambiguity, planning skills and the ability to break down complex investments into a financial model) is likely to quickly move into management. So next topic suggestion – what is the career path for a bright early to mid 20′s financial modeller?

I have been otherwise engaged the past few weeks, I missed this post.

As always I fundamentally agree with your position. I thought I would throw into the mix though my preference for what to look for when ‘developing’ a financial modeller. I highly (almost completely) discount Excel skills….yes you heard me guys…I find that this is next to useless. The trick is to find somebody who knows how to ‘solve problems’. The level of Excel needed by a professional financial modeller can be taught in a couple of weeks…. So budding ‘actuaries’, ‘engineers’, ‘physicists’ etc are at the top of my list.

The icing on the cake is that a good modeller doesn’t think about whats best for them, most efficient for the ‘model’ but what approach suits the user (not the modeller) most for presentation and the ability to ‘check’ and be confident in the result. I find that financial modellers who remain hands-on don’t put as much emphasis on this point.

Anyway – as always – just my observation having done this for a while from a few different angles.

Nick

Another way of looking at the range of skills required to develop a financial model is to differentiate between ‘conceptual modelling’ – forming an understanding of the transaction / business being modelled and relationships between the variables, and ‘spreadsheet engineering’ – the representation of that conceptual model in a spreadsheet – the thing that we typically consider as being ‘the model’.

These skills are often combined in one person, but don’t have to be.

I also hear alot of adversity to VBA, but only ever from people who know pityfully little on the subject. Those who know VBA will say that it is imperative to ensure sufficient levels of robustness can be applied (to most substantial models that is). Of course we have alot of cowboys too, there are many coders who believe every problem requires a VB solution! But how can anyone without extensive knowledge of VBA actually poo-poo it. That’s what I don”t get! I was once that guy

The truth is that most REAL excel guru’s are also highly experienced in their sectors (e.g. finance / engineering) This is because, since Excel is nothing more that a tool, there is little demand for excel developers who do not have a specialism. And learning Excel to a satisfactory level takes many many years, during which time the learner is likely to be working a career and gathering valuable experience.

The best financial models use the solution that is the simplest tool to complete the task…
I hear that alot… And to a point I agree. For instance a stepped approach is often better than an immense formula. BUT… I prefer to say that the best financial models use the most efficeint methods, and these are not necessarily always the simplest.

I would suggest that wide ranging business experience is almost essential to probe for the facts and assumptions that represent the prelude of blending Excel and Finance, to result in a process and outcome (deliverable) that satisfies the user’s needs.

Hence I arrive at three essentials for a financial modeler, with the third (business experience) possibly the first among equals. Just a thought.

Another way of looking at the range of skills required to develop a financial model is to differentiate between ‘conceptual modelling’ – forming an understanding of the transaction / business being modelled and relationships between the variables, and ‘spreadsheet engineering’ – the representation of that conceptual model in a spreadsheet – the thing that we typically consider as being ‘the model’.

These skills are often combined in one person, but don’t have to be.

This topic was recently discussed in the Financial Modelling Podcast – http://www.financial-modelling-podcast.com/.