A quick chart in Excel is a wonderful thing: highlight some data, press F11, and sit back. Enjoy the feeling of fingers off keyboard.
For a brief moment while considering that chart, the modeller enters the realm of the chief executive. That is what chief executives do, after all — they look at charts, then they ask awkward questions. The uncomfortable truth is that most modellers only experience half of that exchange. They can produce the chart. Far fewer can survive the questions that follow it.
Three signals worth your attention
A good modeller develops an instinct for three things in any chart:
- A step change. When a line jumps rather than glides, a modeller’s antenna should bristle. Step changes are hard to model well unless (i) they’ve been modelled badly, (ii) they’re the by-product of an overly complex function, or (iii) the change is genuinely explicit in the input assumptions. A good modeller can always say which of the three is responsible.
- A kink. A change in the direction of a gradient means something other than a constant is now at work. Kinks are quiet — they rarely announce themselves the way a step change does — which is exactly why they’re worth hunting for.
- A spike. A spike suggests an anomaly or an outlier, and chief executives are particularly gifted at asking awkward questions about spikes. If you can’t explain it in the room, assume someone will ask you to.
Only two explanations, ever
Behind every step change, kink or spike, there are only ever two possible explanations: a genuine commercial reason, or a modelling error. Nothing else. Being forced to admit the second is a humbling experience — but it is nowhere near as bad as being presented with a chart and having no explanation at all.
A modeller with no explanation for their own chart has little aspiration to become a chief executive. They’re content to crunch numbers, with little curiosity about how those numbers will be read by someone else. A modeller who can explain every bump in the line — including the embarrassing ones — is a chief executive in waiting.
Charting is a review skill wearing a different hat
This is really model review in miniature. The same scepticism that makes someone a sharp reviewer of other people’s models — assume there’s a mistake to find, keep asking why this was built this way when a better approach exists — applies just as well to your own chart, five minutes after you’ve built it. The modellers who catch their own step changes before a chief executive does are, more often than not, the same modellers who are good at reviewing other people’s work.
Presentation is a discipline, not a decoration
This is where presentation of results, as a skill in its own right, starts to matter. Colour, white space and visual impact are not finishing touches — they’re the difference between a chart that gets read correctly and one that doesn’t. The best modellers are ruthless about stripping non-essential information out of every chart and graph, in the spirit of Edward Tufte and Stephen Few: every line, gridline and colour has to earn its place, or it comes out.
It also pays to keep the data behind a chart in its own worksheet — a dedicated “Graphs” tab, for example — because the link between an Excel chart and its source data is never quite as robust as it looks. A chart that breaks silently when a row gets inserted upstream is worse than no chart at all.
Where a standard removes the guesswork
None of this needs to be reinvented on every model. That’s the whole premise behind the FAST Standard, which F1F9’s founder co-authored: a shared set of conventions for how financial models should be structured, built and reviewed — including where presentation data sits and how it’s kept separate from calculation. Teams working to a standard spend less time deciding how a chart should be built, and more time actually looking at what it’s telling them. The three signals above get easier to spot, not harder, once the surrounding structure stops changing from model to model.
A quick chart in Excel is still a wonderful thing. The modellers worth hiring are the ones who can explain every part of it — the ordinary parts and the awkward ones alike.