I have spent years thinking about trying to apply our design sensibility to the retail asset class.
After all, we’re very good at maximizing the value of small spaces. It ought to be possible to do this with stores, too.
So, why haven’t we jumped in?
1. The macro-trend towards online shopping. My family buys almost everything online (almost all from Amazon, but that’s a different story). I assume that, as people get more and more comfortable with online shopping, this trend will continue, making it very, very hard for traditional retailers to compete.
2. The conflict between our design sensibilities, tenants’ credit and what the banks want to finance.
Given where we do business and the kind of deals we do, I’m sure you can guess that Jon and I are not exactly in love with large chain retail. I’m not any kind of a psycho on this issue, but, given the choice, I prefer interesting one-off shops over chains. And, obviously, if we were going to start doing retail deals, my impulse would be to design projects to appeal to / enhance the businesses of independent retailers.
The problem is that independent retailers are terrible credit risks. They go out of business all the time and they tend to do so all at once when the economy goes bad.
Now, if your tenants are big chains, a recession isn’t so bad. As long as the company avoid bankruptcy, you know the rent is getting paid. But, with independent tenants, there’s not really much you can do, unless you have secured personal guarantees and you’re really excited to try to take peoples’ houses away.
Banks know this and so they are very hesitant to lend against retail properties with so-called non-credit tenants. So, on top of the business model issue, there is also a financing issue, in that the leverage you want / need is both harder to get and more expensive than it is in multifamily.
Bottom line: The type of project we’d love to do, with well-chosen, independent shops, is risky and difficult to finance. So, unless the land is an absolute steal, we’re going to stick with multifamily for the time being.