I learned financial modeling the hard way by creating a model that I ultimately needed to rebuild.
I had just graduated from Brown University; it was the dot-com boom of 1999 and I was in the process of founding my first startup so I was forced to teach myself how to do financial modeling. The first mistake I made was in creating a very complex financial model, which was ultimately way too difficult for others to understand and for me to maintain.
What I didn’t realize until after finishing that first model was how easy it is to make small mistakes. I wish I had a few other people I trusted to look over my model while building it. The experience was sort of a trial by fire.
After this first startup experience, I went to Yale to get an MBA; it was the next phase in my education on how to do financial modeling (which was a big part of what I focused on there). While at Yale, I learned a more formal process for building models.
I used this process with the last company I founded, Boost Media, a digital marketplace for marketers looking for search, social, video or display ads. I had raised a significant amount of capital and used the financial model I built to operate the business and track its performance to make sure we were hitting the metrics we needed to.
At OnPlan, my day-to-day involves building financial models and evaluating financial models for clients. So – at this point – financial modeling is something I think about deeply and often.
Learning How Financial Models Work Through Experimentation
When I was at Yale, I had a professor named Barry Nalebuff. His name doesn’t mean much to most people, but he actually co-founded Honest Tea, which is how he’s best known today (In 2008, Coke bought 40 percent of Honest Tea for $43 million, and bought nearly all of it in 2011).
Professor Nalebuff had a recommendation for all things in life: work backward. I believe working backward is really appropriate when it comes to learning proper financial modeling.
I’m one of those people who learn best not by reading a textbook, but by doing – also known as tactile or kinesthetic learners as opposed to visual or auditory learners
And so, for this class of learners, I would recommend starting with a financial model template, or a best practice financial model, decomposing it – really analyze the underlying formulas and relationships present in this best practice model – and working backward from there.
With this financial model, the really interesting thing is what happens when you delete line items. How does the model behave when you add things to it? How does the model behave when you remove things from it? And the great thing about programs like Excel is there’s always Ctrl + Z – you can always undo anything.
Ultimately, I just think varying one element at a time and looking at how it impacts the entire model is a great way to learn how to do financial modeling.
Another way to ensure you’re understanding, and really absorbing, what’s happening with the model is to make sure it’s behaving in a manner you expect while you’re playing with it.
Some of the most commonly used financial models include:
- Discounted Cash Flow (DCF): used to estimate the overall value of a company by calculating expected future cash flows and then discounting those cash flows.
- Comparable Company Analysis (CCA): used to estimate the value of a company based on other businesses that are a similar size and in the same industry.
- Sum-of-the-parts: estimates the value of a company by looking at the value of each division of it – in case a division opted to break off or it was acquired.
- Leveraged Buyout: the process of acquiring a company using a significant amount of borrowed funds in order to cover the total acquisition cost.
- Mergers & Acquisitions (M&A): used to estimate the effect of an acquisition on an acquirer’s earnings per share (EPS) – accretive or dilutive.
A template of any of those models should give you plenty to experiment with.
Benefits of Combining Elements of Different Financial Models
Over time, as you play around with a model, you’re going to gain confidence. At that point, you can take elements from one financial model and copy and paste them into another model.
With this process, you’re not really getting into the details of each formula; you’re working with chunks of a particular model and transplanting one chunk of a model into another model to enrich your understanding of how the model works.
It’s like learning to write: each formula would be a word; the formulas make up a line item, which would be a sentence, and the line items make up a block or sectional model, which would be a paragraph.
For example, consider the following sentence:
A woman, without her man, is nothing.
But, if you simply move the comma, the meaning of the sentence changes entirely:
A woman, without her, man is nothing.
This is why getting comfortable interacting with formulas, line items, and blocks of a financial model is necessary to truly understand why they are constructed the way they are.
As you learn more and reverse engineer it, you’re going to gain confidence in your modeling abilities.
Further, as people learn to build financial models, there’s often this tendency to create models that are too complex, which makes them more difficult to understand, and – over time – makes them more difficult to maintain. I had this issue when I was building my financial models.
There are two metrics you can use to gauge the level of complexity of your financial models:
- Granularity: refers to the level of detail in a financial model. The more detailed the model, the more “granular” it is.
- Flexibility: refers to how often a financial model can be used, how many ways it can be used, and how many people can use it.
Obviously, a more granular financial model requires much more time and effort to build. And, because it’s more detailed, the odds of errors making it into the model are much greater, which means it’s extremely important the model’s structure – or layout – is well thought out.
The flexibility of a financial model is based on how it will be used. If, for example, you needed to create a financial model for a one-time acquisition that involved two specific companies, you wouldn’t need your model to be flexible. However, if you were creating an acquisition template to be used repeatedly, you would want that model to have a lot of flexibility.
To avoid an overly complex output, try to be cognizant of the model’s utility – it needs to be a useful tool. You want a financial model to answer whatever question you have as elegantly and simply as possible.
If it’s a tool, you must empathize with the end user who’s using the tool.
Generally, the more simple you can make the model, the more usable the end product will be.
This is the learning I’ve had in terms of progressing as a financial modeler, at least.
Recap and Additional Financial Modeling Resources
The four main points I’d like to re-emphasize are as follows:
- Work backward: start with a finished – preferably high-quality model – and be cognizant of the purpose of the model while you dissect it.
- Vary one model element at a time, and see how it impacts the entire model.
- Think of a financial model as if it’s a paragraph; it’s composed of words, which signify formulas, and sentences, which would be the equivalent of line items. If you take one component out and plug it into another model, you can really start to understand how a specific section can be used, just like using words and sentences to create a paragraph.
- Make sure the model is not overly-complex; it has to be useful to your audience and it has to be maintainable over time.
If you’re looking for additional learning resources, we’re currently putting together a library of best practice financial models. These are great for people who are looking to learn how to build financial models because they allow you to start with a best practice financial model for your particular use case or your particular industry.If you have the chance, head to the OnPlan public library, and download some of our free models and play around with them.