Wednesday, July 29, 2015

Housing Tax Policy, A Series: Part 48 - Accommodative Appraisers are not evidence of a bubble

Appraisers are a governor on price fluctuations.  They are a source of friction.  They can't increase price fluctuations.  Nobody with a winning bid on a home raises their bid because the appraisal came in higher.  Appraisals are part of what makes home prices sticky.

We don't have banana appraisers in grocery stores, yet banana prices don't just keep skyrocketing upward because of their absence.

Appraisers have been widely blamed for fanning the flames of the housing boom, but even if you don't buy my argument above, accommodative appraisers are not evidence for a bubble.  There are two scenarios that could explain the speculative boom.

1) Buyers started bidding up the prices of homes with little concern for their value, accommodated by banks issuing mortgages without regard for the danger of these prices falling back down to reasonable levels.  Since the houses were overpriced, banks had to find friendly appraisers who would use aggressive methods to justify the prices that kept the bubble going.

2) Home values were rising at an unusual rate, due to rising rents, rising expected rents, and falling long term real interest rates.  Because of the unusual rise in intrinsic values, frictions in the home market made it difficult for market prices to follow.  This created an opportunity for "flippers" and speculators who could profit from the price stickiness by buying homes at less than intrinsic value and selling them fairly quickly at an expected profit once the market price overcame those market frictions.  Appraisers willing to use more aggressive methods helped to disengage the housing market from those frictions and reduced the inefficiencies that led to speculative profit taking.

Maybe scenario one seems overwhelmingly more reasonable to you.  That's fine.  The point is, aggressive appraisers will appear in both scenarios.  They are a sign of strongly rising home prices.  If prices are rising quickly, appraisers will face these dilemmas.  Strongly rising prices are not a disputed fact.  The fact I dispute is that prices were unhinged from fundamental value.  Wherever we come down on that, the behavior of appraisers is irrelevant and is not a sign of a bubble.

Here is a graph from yesterday's post, comparing this measure of housing "yield" to 30 year real treasury yields.  These measures of relative yields did not diverge during the boom.  They diverged during the bust.

My next post will probably be a new review of historical home values and returns.  On the issue of home values, here is a comparison of 20 year TIPS (inflation protected) bonds and homes from the end of 2006 (the worst time to buy a home) to the end of 2014.  For home returns, I am using BEA data from Table 7.12, estimating net rental income after all expenses and depreciation (rental income plus net interest expense), and total real estate market values from the Federal Reserve's Flow of Funds report.  Housing Yield is that net income estimate divided by owner occupied home market values.

Yields at the end of 2006 were very similar for both homes and 20 year tips bonds.  CPI inflation and Shelter inflation were both about 2% in the intervening 8 years, so the experienced income inflation for both of these investments has been similar.  But, total returns were 1.8% for homes and 5.4% for 20 year TIPS over those 8 years.  Yes, that's right, the total return to the average home bought at the worst possible time - the end of 2006 - has been a positive 1.8%.

Total returns on the average home were undermined by the jump in yield (which moves inversely to price).  So, today, the average home yields 3.4% and TIPS bonds yield 0.7%.  Keep in mind that shelter inflation at this point is running a full point above broader inflation rates, so in addition to earning 2.7% higher real returns, homes are also pocketing an extra 1% of inflation growth each year now.

As I have pointed out, if there was a housing supply bubble, the adjustment would have been associated with dropping rent income.  We have not seen that at all.  The bust has been entirely from an increase in yields on homes.  This is a sign of a negative demand shock in home ownership.  It is important to distinguish between home ownership and housing consumption.  These are two different issues, and it seems like pundits frequently treat buying a home as if that is the same as adding a housing consumer.  It's not.  If anything, it may be adding a housing supplier.  Sometimes, observers do treat the boom as an overbuilding phenomenon, but the intervening years have not borne this idea out, since rent continues to rise.

I suppose one reaction is to say that monetary policy has been loose all this time and the 0.7% 20 year TIPS yield is artificially low.  First, there is simply no way that the Fed could push 20 year TIPS yields to 3.4% by tightening the money supply.  Second, if they could somehow do that without creating a deflationary mess, then TIPS returns in the table above would basically match the average home return.  So, over an 8 year period, TIPS bonds and homes would both have exhibited normal behavior in the face of rising real yields - reasonable incomes with small aggregate real capital losses due to the effect of the new yield.  So, even the mistaken identification of low long term rate rates and recent Fed policy as loose doesn't salvage the housing bubble story.

We simply have created a period of time where home buyers have been able to capture above-market yields because of regulatory and monetary obstacles to home buying.  With a decade long economic dislocation as the side effect.  I should have more on this tomorrow.


  1. I haven't read the rest yet, but the opening on appraisers is either wrong or misleading. Appraisal set a cap on what a bank will lend against the collateral. Your position would be correct if banks never used appraisers, and then brought appraisers in to set caps in this way. The first caps would automatically be less than no cap- and so they would not be able to influence prices upward. In a system that has used them for years a change in behaviour of appraisers can lead to higher prices in two ways. First simply by raising the max that a bank will lend against a house, and secondly a larger appraisal gives a homeowner the option to take a out a larger HELOC type loan which is basically an option value on the house. Making the house more attractive to the homeowner means a higher bid is needed to convince them to sell.

    1. Great point about increasing demand for credit by existing home owners.

  2. There is a third model (and probably a 4th, 5th ect) for a speculative boom which combines the two in some ways- the regulation/choke-point model. Imagine a river that you want to dam, that has 10 potential outlets. You build a damn at the lowest point and the water level rises so that the 2nd lowest point now has to be dammed, and that causes the water level to rise again so that the 3rd lowest point has to be dammed, ect ect ect. Each successive damming means not only that a new dam must be built, but that (eventually) the old dams must be rebuilt (heightened or strengthened). Obviously in this analogy the dams are regulations and the water is liquidity/money supply/savings (whatever you want to call it) This is basically what All the Devils are Here and House of Debt are arguing (in different ways)- The dams were fine for X period of time and then the market conned the government into not building/repairing dams and there was a flood.

    If you view the Fed as having unlimited liquidity though then no dam will permanently stop the flood, and any flood that happens will appear to have been caused by a bad dam (or two) as the weakest/lowest break first.

  3. I think I have the final piece of the puzzle for "Baconbacon's Grand Unified Theory of why it all went to hell".

  4. When you say that the return to housing since 2006 has been 1.8%, is that inclusive or exclusive of a value for occupancy (for owner users) and/or net rental income (for landlords)?
    Thank you.

    1. For this figure, I am using imputed rent for owner-occupiers. I am using aggregate income after expenses and depreciation from the BEA and aggregate market values from the Federal Reserve, so it would include lost income from vacancies, etc., although this is less of an issue with owner-occupiers.

      I would use landlords and cash rent, but it's a little harder to get at all the numbers because of the different types of owners.