Real Estate is a Difficult Market to Arbitrage

Posted on by

Nick Timiraos has some useful charts related to house prices. The ones that interest me the most are the last ones, which show price-rent ratios. For LA, it is 25, and for SF it is over 30. But for NY, Boston, and DC it is all under 20.

I think of the inverse of the price-rent ratio as a measure of the real interest rate. Thinking of it that way, the real interest rate in San Francisco is about 3 percent, while the real interest rate in New York is about 7 percent. That suggests that you want to borrow in SF to lend in NY. Or, in this case you want to go short housing in SF and go long housing in NY.

Some reasons why this does not happen:

1. It is hard to go short in real estate.

2. It is hard to go long the “average” price in a city. You can only buy specific houses.

3. The market can stay mis-priced (and get more mis-priced) longer than you can stay solvent.

13 thoughts on “Real Estate is a Difficult Market to Arbitrage”

  1. Could me measurement problems too. Different cities have different rent-control, building-permitting, and affordable housing regimes. Not to mention whether subletting is legal, or to what extent, or how many people or families are allowed to live under a single roof. The expectations regarding future supply, demand, and neighborhood character make things hard to compare directly.

    Also, different cities have really different mixes of both the stock of housing that tends to be rented out vs. owner-occupied, and also the profile of the each kind of resident. I’ve moved around a lot, but I’ve lived in cities in which the upper half mostly lived in owner-occupied detached housing while the lower half rented in complexes in different neighborhoods, and other cities in which things were not nearly as starkly sorted as that. It’s hard to correct for factors like that.

  3. #2 doesn’t seem so hard — can’t you invest in a rental property REIT focusing on NY or DC? But, yes, #3 is a pretty good reason not to.

  4. Think of homes in supply constrained cities as growth stocks. Those high P/R ratios are a function of rent inflation expectations.

    http://idiosyncraticwhisk.blogspot.com/2015/10/housing-series-part-67-better-numbers.html

    Persistent rent inflation comes from the economic rents that accrue to laborers because housing creates a barrier to entry into our most dynamic labor markets. Since housing is the bottleneck, those rents eventually flow to real estate owners. High P/R is a sign of efficiency. Home prices reflect the future in-migration of higher income households into those labor markets and the continued out-migration of low income households who won’t be able to afford to live there any more. San Francisco is unsustainable because of housing constrictions. The renters there will eventually have to move to Texas to lower their costs. They may not realize it, but the price of the home their landlord owns does.

    http://idiosyncraticwhisk.blogspot.com/2015/12/housing-series-part-95-closed-access-is.html

  5. What Handle said.

    Also, property taxes have a HUGE effect on prices by siphoning off rental income, and they vary enormously from one place to another. Unless they are taken into account, RE markets are just not going to make sense, especially in places like Texas — not necessarily as ‘affordable’ as it looks at first glance. Texas is great, but it isn’t magical.

    You can’t boil it down to 2 numbers.

  6. The largest difference is growth, which would make equating rates foolhardy. Are people moving there should be your first question. Will they continue to do so your next.

  7. Part of the discrepancy might be because the data is only for rent to price, not net operating income to price. Imagine a million dollar house in Connecticut and an identical house in North Carolina that’s only $200,000. They both cost $10,000 a year to maintain as the houses are exactly the same aside from geography. They also have the same net operating income (NOI) to price ratio of 5%. The rent to price ratio for the Connecticut home will be 6% ($50,000 NOI + $10,000 maintenance). For the Carolina home the rent to price ratio will be 10% ($10,000 NOI + $10,000 maintenance).

    • Copper pipes may cost about the same but I’m betting a plumber gets paid a lot more in SFO than Raleigh. Ditto utilities and taxes.

  8. At interest rates that stretch out payback for more than a generation, regime uncertainty can easily be a big deal compared to the idealized problem of allocation of a limited stock of securely owned capital that economists tend to think about. (And regime uncertainty tends to be hard to arbitrage.)

  9. from a recent empirical paper,
    “Across MSAs, rental yields decline with prices, while HPA
    increases with prices. As a result,“total returns”, are approximately equated across MSA’s”

    The frictionless market interpretation of what you find is just that expected house price growth is higher in SF.
    link:
    https://cace179b-a-62cb3a1a-s-sites.googlegroups.com/site/andrealeisfeldt/sfr.pdf?attachauth=ANoY7cpcqAH8Vuu4bTRMtVTKM90PIyxOAHS5ABA6Uq-f3ag4rBPVOiI0-vwRFAjYAp5ZAG4lnaErNcPPCcipn-ezreKMQkN0zSzY5_pRDNJ0wG7kqNGrEXVdjZiuw-4lVUmAYmAVrr1Dg9N3Cm6yMxRewmxAvtVrnxqfVLTJWW5bt2YK3q6OILrow3_qSBbWhsXpd_PJfkXGDemWdE-muBpJyDITSY6Edg%3D%3D&attredirects=0

  10. I think this probably reflects Prop 13, reducing property taxes in California. Taxes and maintenance are included in the rent, but not in the purchase price. My property tax payments in upstate NY equal what I pay for principal and interest. If one assumes that this is generally true, then at most half of NY rent payments can be considered interest. That brings our rate down to 3.5%, or within shouting distance of what it is in SF.

    In other words, the solution to the mystery is there is no arbitrage opportunity to begin with.

Comments are closed.