What self-driving cars mean to real estate industry

Self-driving vehicles are coming sooner than you think.  Imagine you can just hop on the car, enter an address on your phone and the car will take care of the rest.  Well, that might become true very soon.  Tesla has offered electric vehicles with semi-autonomous Autopilot feature for a few years now.  Waymo opened its ridesharing program Waymo One using autonomous cars to the general public in Phoenix last October.  Driverless cars from Nuro are already on the road as delivery vehicles in the Bay area starting last December.  As it gradually turns into an inevitable reality to our life, I cannot stop thinking what all these mean to the real estate industry.  

My thoughts took me back to the first class of Commercial Real Estate taught by Professor David Geltner at MIT.  The classic real estate theory about urban areas is mainly based on the Monocentric City model.  The model explains that any real estate value is the sum of three components: agriculture rent (as if the land is used for farming purpose), construction rent (replacement cost) and location rent.  The first two components are pretty self-explanatory.  The third component is determined by two factors: population density and transportation cost.  Everything being equal, the lower transportation cost, the higher the location rent and real estate value is.  This explains why real estate in Manhattan is generally higher than that in outer boroughs, NJ, Long Island or West Chester.  The model also says:

Declining transport costs (per person, per mile, or per year) holding population and income constant, will always reduce the value of location rent in the center of the city. (Chapter 4 of Commercial Real Estate Analysis and Investments, Geltner et al.).  

Among a variety of potential impact from self-driving cars, transportation cost reduction is probably one of the sure outcomes.  If I have to bet my money on anything related to the aftermath of autonomous vehicles, that would be my number one choice.  If that holds true, I see the real estate value and apartment rents in urban centers face significant downward pressure.  The option to travel without stress and tangible cost savings will make people start to rethink if they really need to live in the dense urban core.  

But that’s not to say places like NYC or San Francisco are doomed.  According to the Monocentric City model, there is another factor in determining the location rent: population density.  The greater the population density, the higher the location rent is.  And that’s probably the savior.  If the 24-hour cities can continue to promote the theme of Live-Work-Play and attract jobs and young people to go there and keep the population density high, they might be OK even with the wide adoption of self-driving technology and transportation cost reduction. The real estate boom in recent years in San Francisco and Seattle, and perhaps NYC to a lesser degree, can all contribute to the huge population growth as a result of surging tech jobs.  

Another likely outcome from self-driving cars is the expansion of cities.  With the option to travel without the stress, people may be more willing to move to the periphery of the urban center where you can still enjoy the fun in the city without traveling much but have larger space and better schools to raise a family.  I see areas surrounding the CBD and outershirt of the urban center will likely experience a meaningful population growth and real estate boom.  Also, I think Build-to-Rent housing will flourish in those areas along with service and retail businesses.  

As to any particular sector in the real estate industry, I think self-driving cars will be a tailwind for hospitality businesses.  Due to transportation cost reduction and stress-free driving, I see travel (particularly regional travel within 350 miles) will get a meaningful boost.  That will create immediate demand for hotels, restaurants, recreational and leisure establishments, and maybe even retail.  It will help air travel as well.  How many times have you and your family dreaded about the long and costly uber ride between home and the airport?  Putting all these together, I think hospitality may be a winner rising from this disruption brought by self-driving technology.  

One area of caution is industrial.  With all the frenzy around warehouses and logistic centers, I asked what self-driving cars would do to industrial assets.  I am usually a contrarian thinker.  To me, industrial has some headline risk.  I am not saying there is a black swan event coming to logistic centers.  But I see autonomous vehicles are more like a headwind to industrial assets instead of tailwind.  Amazon is paying roughly $28/SF triple net while local businesses are only paying $15-$18/SF for warehouses in NYC.  Even with continued demand growth, I don’t see industrial developers and landlords could make sustained profit over a long period in a market like NYC with very high cost of development.  High net worth people might pay $238M for a penthouse condo along Central Park South.  But who would pay that much (on a per SF basis) for a warehouse?