Yellow Pad Math Can Explain A Lot About Homebuilder Decisions

If you don’t understand the math, you just don’t understand

A number of years ago, when I moved from California to the East Coast, I struggled to understand why the builders seemed to think so differently about their projects than I was used to.

I expected the numbers to vary by price point, but the business’s actual approach was different. To someone from California, it was unfathomable that a builder would pass on buying good lots because they didn’t have product that would fit, or that they would put a house on really good lots without taking advantage of local features, or be unconcerned about selling two homes a month.

In California, we had a saying: “four per month is making real money, three is weak but okay, and two is a problem. One model, not four.”

The list of operational differences was long.

One day in investment committee, we were discussing a deal in California. Good location, not great. The small lots were about 50% of house price – direct costs were 25% of house price. I had a deal following it where the lots were 20% of house price – direct costs were 55%. And it hit me:

• When we designed new product in California to take advantage of those lots, if we screwed up the direct cost estimate by 10%, that was 2.5% out of margin. Not good, but not tragic.
• In Atlanta, where my project was, a 10% direct cost miss was 5.5% out of the margin. More than half the projected margin. Tragic. So being cautious on new product, overly cautious in my mind, was not without merit.
• Because 50% of the money was spent up front in California, inflation was our best friend. Half our costs were locked in, so the spread between cost and revenue inflation could be huge.
• In Atlanta, only 20% of the cost was locked in. Inflation could be helpful, but not nearly as much.
• With 50% of the spend up front in California, the internal rate of return (IRR) was doomed if the project went slowly. The capital was eating a hole in the pro forma. In Atlanta, with 20% up front, much less so. So, the benefit of going fast was less. Note, I said less, not that there was no benefit.
• If the benefit of better velocity is less (again, less, not nonexistent), then you might think differently about whether you should have four models.

There were other differences, but the lesson was clear: the math of the business was so different that it led to different behavior.

I’d still argue there are significant benefits to SG&A leverage, capital recycling and overall returns from increasing velocity. Those arguments remain valid.

However, understanding why rational people behave differently than you expect requires analyzing the math of their business, not yours.

This is just one example of what I call “yellow pad math” – the discipline of working through a problem with basic numbers before reaching for a spreadsheet. The exercise above taught me more about my new business than the most detailed model could have.

If the behavior you’re seeing doesn’t make sense, don’t assume the other person is wrong.

Do their math. You may find they’ve been right all along.