Lots of people are looking for investments they can feel good about, which makes me think people might be interested in putting a portion of their savings toward helping resolve the housing crisis if it offered a reasonable return. The idea is that you construct 100% affordable housing in places like Somerville that offer large density bonuses for affordable housing. Sketching some numbers, it looks to me like this is:

  • Potentially very profitable: ~2x ROI over ~3y, which is about ~25% annual growth. It's likely not actually that amazing when you factor in other aspects I don't know about, but there's a lot of room for a lower ROI that's still pretty attractive.

  • A large opportunity: you could build 8k units in Somerville for ~$2.8B. This is ~$0.8B in land purchase (assuming city assessments are half the purchase price on average) and ~$2B in construction (assuming a cost of $200/sqft). Much larger when you consider all the other municipalities that have similar programs.

A group of neighbors pooling their funds to help resolve the housing crisis is also a lot more politically robust than independent for-profit developers coming in.

Does anything like this exist? Has anyone considered starting something?

New Comment
2 comments, sorted by Click to highlight new comments since:

Does the construction cost take the costs of one to infinity years of legal delays into account? I'm pretty sure if people could build affordable housing with a 25% yearly return, they'd already be doing it.

The idea of getting the funding from locals to make the delays less likely might help, although it only takes a small number of loud people to delay construction.

The construction cost estimates I referenced don't consider delays at all. I agree delays make this less attractive.

We're talking about building something so much taller that the land cost is only about 10% of the total cost, which means most of the delay happens before most of your expenses (the period when your capital is tied up) so this isn't terrible, but it does still add a lot of uncertainty to the process.