Friday, September 22, 2017

The Housing Inventory Mystery

In real estate, and finance in general, I see a lot of standard analysis that seems backwards.  A lot of it would be something Scott Sumner would call "reasoning from a price change".  Analysis along the lines of predicting a decline in sales because prices have risen, making homes less affordable.  Prices are a signal of demand, not a cause of it.

This appears to be the case with inventory levels.  Inventory of homes for sale has been low during the recovery from the financial crisis.  Real estate analysts seem to commonly treat inventory as a signal of price shifts.  If inventory is low, that will lead to rising prices, because buyers have less supply to bid on.  If inventory is high, that will lead to declining prices.

This sort of analysis makes sense if we are looking at cyclical signals of a market with stable fundamentals that is constantly in a moderating process of finding an equilibrium of buyers and sellers.  In that context, rising inventory might signal an unexpected decline in demand, which would tend to lead to dropping prices.  That surely is what happened in 2006 and after.

We can see that the first measure to peak was homes sold but not yet started.  Homes sold but not started is almost a sort of negative inventory, and it had been high during the boom.  After that began to decline, builders soon responded by reducing the number of spec homes they were building.  Demand was falling fast enough that spec homes finished but not sold rose until the beginning of 2008.  Considering that new home sales continued to decline to very low levels until 2011, this seems like pretty responsive inventory control to me.  The inventory of spec homes declined pretty sharply in 2008, even as sales continued to crater.  Inventory of existing homes (not shown), on the other hand, did remain elevated until 2011.

Since then, inventory has moved back down to the low levels previously seen during the boom.  If inventory is a leading indicator of seller power, then this is a bit of a mystery.  Low inventory should trigger rising prices and new home building.  But, home price appreciation has been moderate, even with strong rent inflation, and new supply continues to only develop slowly.

The reason is that the housing market isn't in the midst of normal cyclical shifts.  It is in the midst of a wholesale secular shift in mortgage access.  A significant portion of the population that once could count on having ownership as an option, cannot anymore.  Some families have negative or minimal home equity, which makes selling or moving difficult.  Some families wouldn't be able to qualify for a mortgage for the homes they already own.  There has been a large shift from ownership to renting, because of this.  That shift, itself, probably tends to tamp down low-tier and mid-tier housing demand, because there isn't as much value in renting vs. owning, due to the lack of control over the asset, and it also means that a significant conduit of new supply - sales to new homeowners - has been cut off, forcing other buyers of new units (new supply) to make up the difference.

The reason inventory is low is because there aren't many potential buyers, and families that do need to engage in real estate transactions might have to downsize or shift to renting if they can't manage to get funding.

So, as with interest rates, the signal here is probably flipped from the way it is normally thought of.  If interest rates rise, that will be a sign of expanding investment, which would likely be related to an increase in households able or willing to take an equity position in new homes.  Rising interest rates will probably be related to rising home prices and rising housing starts, even though that might seem counterintuitive.  (Actually, regarding interest rates, I'm not sure that this is that unusual.  On a secular time scale, long term real interest rates reflect broad trends regarding saving/consuming, etc., and do seem to have an inverse effect on home values.  But cyclical shifts in interest rates, especially short term rates, are more a reflection of short term shifts in demand and sentiment, so that housing activity tends to be strong when rates are rising during an expansionary period.)

Likewise, rising inventory will probably only come about when there is a new shift to more potential homebuying from those marginal buyers.  Rising interest rates, rising inventoary, rising prices, and rising starts will probably all either develop in unison, or not at all.

Reflecting on monetary policy, if the Fed was actually following natural rates higher, we would be seeing these developments.  Instead, the yield curve is flattening, credit growth is moderating, housing starts remain low, and non-rent inflation is dropping.  I suspect that the natural rate will remain very low either until we allow mortgage markets to adjust to their previous standards, or until the housing market fully adjusts to the new standard, where middle class households are generally renters and are consuming housing, in terms of rent, based on those new stable standards, instead of being grandfathered into the homes they were able to buy before the shift.

I think the barriers to supply in the Closed Access cities, which is an international problem, also lower rates by reducing potential investment, but comparing long term real rates since the crisis to before the crisis, the mortgage collapse seems like it could be a larger factor.

There are obviously many international factors that play into interest rates.  But, these real estate distortions are huge.  The Closed Access problem probably inflates real estate values in the US by $3 trillion or more.  Internationally, this must amount to something like $10 trillion of capital value that required no investment.  Up to 2007, this meant that Closed Access real estate owners earned excess profits for preventing new homes from being built.  When they sold those properties and realized those unearned gains, the new buyers had to transfer cash to those owners, using labor income to fund transfers to capital.  This led to rising mortgages outstanding from the buyers, but it also led to rising savings on the part of the sellers as they re-invested their realized capital gains (although these gains are not generally included in official measures of savings).

Since 2007, federal regulators have prevented new homes from being built through mortgage constraints.  This has caused profits on existing homes to rise, but kept home prices low.  So, the effect on mortgages is the opposite - mortgages outstanding declined instead of increasing.  But, otherwise, this is similar - capital captures high rental income, but this income cannot be easily reinvested into real estate markets, so there is an obstacle to the investment outlet that would utilize savings in a way that lowered capital incomes and lowered yields in real estate.  Savings must be invested in bond markets or equities.  So, yields in real estate are above long term ranges and yields in bonds are below long term ranges.  This also probably amounts to more than $3 trillion in distortions.  Here, it isn't an inflation of savings; it is a decrease in potential investment.  $3 trillion worth of homes have not been built, in spite of being economically useful, since the crisis.

All told, there are more than $10 trillion of distortions in the global real estate market either inflating the value of prior savings or decreasing available investments.  And, we should keep in mind that, in terms of yield, these distortions are somewhat targeted.  Yields aren't low in real estate earnings.  Total expected returns to equities are not particularly different than they normally are (somewhere around 7%, in real terms, in the aggregate).  These distortions get focused onto fixed income markets that don't have obstructions.  Developed market debt markets amount to about $90 trillion.  Some of that is not investment grade.  It seems reasonable that these distortions could affect yields in low risk markets.

Given this, complaints about foreign real estate buyers seem misguided.  First, if you impose a strict set of policies limiting borrowing among domestic buyers, of course there will be an uptick in buying from foreigners with access to foreign capital outside those controls.  But, secondly, that capital inflow might actually lead to some new supply.  It can't lead to much supply in the Closed Access cities, which is where the complaints are usually lodged.  But, it might lead to supply elsewhere.


  1. What exactly are the policies limiting borrowing among domestic buyers? As we've discussed before, I think the limitations are market-based...these individuals are simply not good credits. If anything, our focus should be pushing real wages up...which monetary policy seems to have done the opposite of this cycle.

    Take for example the attempted disintermediation of banks in lending to low/mid-end consumers. I'm thinking of the various fintechs like lending club. My understanding is that returns have been poor. Banks will only lend to good credits because, more than ever, there is a bifurcation in economic outcomes for the population. The low/mid end simply has poor economic prospects.

    1. This comment has been removed by the author.

    2. Here is a previous post on the matter:
      There is no way that yields on new mortgage lending from the banks are low. New default rates are negligible. Nobody is defaulting today. Actual returns on those loans have nothing to do with the ability to pay.

  2. Even if the GSEs "irrationally" tightened standards after they were bailed out that doesn't explain why the private sector wouldn't quickly fill the void to lend to good credits.

    1. My understanding is that the CFPB has cast a pretty wide net regarding implied liabilities on what can be deemed, in hindsight, predatory lending based on ability to pay.

  3. Great post.

    While I agree that xenophobia is not the best route, if the policy is to zone and then credit-rate the middle-class out of many whole metropolitan regions or neighborhoods (remember, within every open city are zones that are closed access) but invite wealthy foreigners in (financed by large trade deficits) then what do we expect some voters to think?

    Are Hong Kongers actually supposed to greet with open arms the mainlanders who are pricing them out of their homeland?

    The solution is an end to property zoning, but that does not seem to be in the cards.

    I cannot blame voters for concluding, "Housing is scarce resource, so why are we so gung-ho on immigration?"

  4. Back on "Why is the fed so tight?"

    This is actually from a right-winger:

    In the 1960s, Harry Johnson, a conservative professor from the University of Chicago, writing in a journal dominated by the conservative perspective of his school, offered a shockingly honest evaluation of the class bias of monetary policy. “From one important point of view, indeed, the avoidance of inflation and the maintenance of full employment can be most usefully regarded as conflicting class interests of the bourgeoisie and the proletariat, respectively, the conflict being resolvable only by the test of relative political power in the society and its resolution involving no reference to an overriding concept of the social welfare.”14


    if you can open. a nice chart on REITs…apartments, single family doing well…and manufactured homes doing great

  6. I think you have written about this.

    Manufactured homes had the stuffings knocked out of them in late 1990s, then kicked while down in 2008. Added up. a lot of units missing…