Thread by @rebootdude

1/ Building financial models.

I often start my financial modeling lectures with the following lines from Alice in Wonderland.

The models we build depend on the questions we need to answer.

If you don't know the question, the model as well as our answer won't have any purpose.

2/ A model for raising funding is inherently different from one for planning and managing cash flows.

Models used for pricing and market share are different from ones we use for planning resources and head count.

We often use these parts together but the intent differs.

3/ If you know the question you need to answer, start with the simplest variation possible.

I call it the 5 x 5. A model no larger than one that can fit between 5 row and 5 columns in Excel.

Short and sweet. Simplicity forces you to think hard about choices.

4/ I have seen this theme again and again professionally.

With words (simpler is better), with frames and films (avoid clutter), with speeches (key points only), with food (water over Gatorade), with training (simpler routine beats complex).

A simple model is a better model.

5/ Simpler is harder. Not necessarily easier.

As a teacher, breaking down complex concepts into bite size easy to understand and relate to morsels is really hard.

But its the singular common identifier of great teachers.

6/ With models you can always add complexity and sophistication later but you have to first make the model work at a 5 x 5 level.

Think of it as a proof of concept for your approach. A sand box to play with. Once it works in the sand box, you can roll out the bells and whistles

7/ Here is an example.

A once built a model of the Saudi sovereign fund and its impact on global oil prices using a 10 x 10 grid.

The question I was trying to answer? Will oil prices hit $10? The answer depended on the pain the sovereign wealth fund could bear.

8/ The entire Saudi economy was compressed in a single line. The global oil supply and demand was two.

How?

The key question was how long would the fund survive and last while providing support to the kingdom. If I could quantify the pain, everything else was irrelevant.

9/ It took me less than an hour to build the model.

But it took me a week to drill down to the question that was at the heart. And a year before that to understand crude oil markets.

As is true with all other things in life, better design beats bad design any day of the week.

10/ Before you build models ask yourself:

a) What is the intent?
b) What is the question we need to answer?
c) How can we simplify the model?
d) What is irrelevant to our core question?
e) What is relevant?

Focus on the relevant.

11/ In my first pass of building a balance sheet, I often flat line everything which is not directly relevant to my line of thinking.

I put in place holder values to hold things together till I figure out the model. I then come back and fill in the holes.

Iterative is faster.

12/ Start with simple editions. Add complexity later.

Simple is harder, not easier. Simpler is also a sign of confidence in your analysis.

Complexity often reeks of insecurity. Because "I don't really want anyone to figure out what I have done".

13/ When you finally present results and recommendations, focus on the relevant.

Don't tell me about the model and how you suffered building it.

Share insights, recommendations and results. The conditions that would break the analysis. Lead with the beef, not the sides.

Before you ask:

The single line was annual budgetary deficit. The oil model OPEC spare capacity + US shale production.

The question was how long before the fund hits double digits because that is when the Saudi's would ramp up production and wreak havoc on oil markets.

14/ There is nothing more powerful than presenting a simple yet profound insight.

Driven by a simple yet thoughtful model that shows your mastery over the subject.

Shorter meetings, faster decisions, more relevant discussions, grateful clients.