Real estate investors have a deep bias towards cheap.

** We all want a screaming deal **

I believe this is a flawed framework for long-term investing.

Evergreen is researching this topic - here are some early findings: ⬇️
We started with the question:

Do going-in cap rates predict outcomes??

In other words, do high or low multiples (expensive vs. cheap) at the time of purchase forecast returns?

Wanted to see if things that looked "expensive" 10 years ago, were actually cheap. And vice versa.
Can't find reliable private market data, so we used public real estate (REITs) as a proxy.

We looked at the 10-year total returns vs. the going-in multiple on net operating income in 2010.

Here is the scatter plot of that data:
Punch line:

There is a positive relationship between quality (what looked expensive) and total returns last cycle.

Obviously a ton of variance though - so don't quit your day job and start overpaying for everything.
Could be any number of reasons why quality real estate / teams outperformed.

- Stronger markets = faster rent growth
- Superior debt, which lowered their cost of capital
- superior business models within niche sectors (ex: data centers, manufactured housing, self-storage)
Of course, countless investors have lost their shirts “over-paying” to win a deal.

And plenty have made a killing buying 7% caps, improving and flipping.

​There are no absolutes in this game. Both strategies work if well-executed.

Especially over the short term.
However, over the long term (>10 years), I suspect the relationship between quality and returns strengthens.

This is the game we like to play.
Final takeaways:

1. If you need 2-3X returns in a few years, value-add deals (aka cheap) is the superior path - ruthlessly executed via skill or cycle timing

2. If you’d rather compound wealth over the long-term, with lower risk, quality is the higher probability path.
You can follow @bradleyjohnson2.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: