Wednesday, August 5, 2015

Housing Tax Policy, A Series: Part 51 - Housing and Strong Form IMH

Tyler Cowen generously posted a comment I had made about supposed short-term bias in financial markets.  My point was that there is an apparent contradiction between the idea that (1) short-sighted investors consistently push firms to massage quarterly numbers and invest for short-term profits over lasting long-term growth and (2) there was a speculative bubble 15 years ago based on a group of firms that had never been profitable and had no near term plans for being profitable.

The commenters on his post generally balked at the idea, noting that there can be investors who are short-sighted at the same time that there are investors who are overconfident speculators.  This is well enough, as far as it goes.  The distinctions here are subtle, non-falsifiable, and are about scale.  Of course there are ebbs and flows in investor sentiment, and of course at any given time there are investors with many different sentiments.  But, the problem is figuring out when these sentiments become such an influence on price that markets become adversely inefficient.

My reaction is that there appears to be a lot of motivated reasoning against finance.  We are uncomfortable with conspicuous profit from risk.  Speculation is unseemly.  It seems like it must reflect our baser motives.  So, to my eyes, there tends to be a pretty obvious predisposition for finding fault in financial markets.  This is in stark contrast to the way we should feel.  Speculation is what I do for a living, and, despite a fairly strong history of success, in my weaker moments I am almost embarrassed at myself for thinking that what I do is possible.  There are millions of people in the world diligently (even greedily) searching for mispriced securities, including many highly trained, motivated, and intelligent people.  The hubris required to bet against them is damning.

A complicating factor on this question is that securities with the longest time frame are the ones that are most vulnerable to small changes in expectations (discount rates, growth expectations, etc.), so, in a way, the popular notion of imagining speculative fevers in long term securities can be attributed to the seemingly contradictory complaint of short-termism.  Markets in those long term securities can become volatile when those sensitive variables change, and cynical observers can attribute the changing prices to short term speculation.  So, in the 2000s, when households were investing large sums into, perhaps, the longest-lived security of all - homes - a consensus explanation didn't build around the idea that forward-looking households were scrupulously investing their incomes in their homes, in a mass attempt to create a long-lasting financial safety net.  Instead, the consensus reaction was that the housing market was dominated by short-term speculators with unhinged expectations.

In the end, public policy was, itself, affected by this consensus, and we created a classic, monstrous example of strong form IMH (Inefficient Market Hypothesis).  Those unhinged speculators clearly pushed home prices beyond their reasonable levels, so when home prices eventually collapsed so far that the economy followed, we told ourselves that this was our medicine.  It had to be done.  To engage in any policy that might even indirectly support home prices would be driving them back to those inappropriate levels, and, furthermore, it would prevent those speculators from learning a lesson about greed and overconfidence.

It is strong form IMH because it looks exactly like that is what happened.  It's not crazy to think that this story is accurate.  There are certainly enough reasonable pieces of evidence to confirm it.  So, the self-imposed public policies that popped the bubble prove themselves to be appropriate by creating a believable, yet inevitable, outcome.  There is no way to steer a public that insists on viewing our financial selves as inherently unstable away from this self-imposed destruction.

But, what if those long-lived assets really aren't being driven by manic short term speculative fevers?  What if households are investing for long-term income, and it really is a contradiction to imagine that we are both short term oriented and given to speculative fevers in long-term securities?

Here is a graph of real home yields compared to real equity and bond yields.  (See this post for an earlier version of the graph and descriptions of the method.)  The high inflation 70s mess up the trends, because required returns on equities shot up during that time and bond yields simply adjusted for inflation (using the GDP deflator here) are only a rough approximation of real yields based on a market rate.  The bond yields in the early 1970's are probably understated because experienced inflation was coming in higher than expected inflation and bond yields in the 1980s are probably overstated for the opposite reason.

But, except for that issue, and allowing for some pretty noisy annual growth and inflation data, total real yields to equity (including real expected growth) run along a flat trend of just over 10% while bonds and homes run about 3% to 4%.  If anything, bonds have the least stationary real yields of the three asset classes.  (I used 10 year treasuries here.)  Except for the treasury yields, all of the other yields are derived by comparing incomes and market values from the Federal Reserve Flow of Funds report and BEA tables.  And, we can see that yields on homes (net rent after costs and depreciation, divided by market values) form a pretty straight and level line.

This goes to one of my themes about the housing boom.  Equity valuations are notoriously flexible.  Bond yields (and, thus, valuations) are similarly traded on highly liquid international platforms.  Home prices are sticky.  Rent levels, in the aggregate, change very slowly, and  home prices can't generally rise more than about 10 to 15% per year, because of the illiquid housing market and various frictions in play.  Those frictions can keep demand due to falling yields from pushing prices up quickly, but we did learn in 2007 and 2008 that they can't keep them from falling so quickly when demand collapses.

Here is a more close-up graph of real home yields over time.  These have fallen within a range of about 1 1/2% since 1960.  At the same rent level, a home with a yield of 2.4% would be worth about 67% more than a home with a yield of 4%.  Simply based on cash yields, home yields were just below the historical range in 2006, at 2.4%.

The narrative about over confident home speculators says that home prices got bid up because of undue optimism about price appreciation.  But, let's look at this last chart.  The top line is the income yield on homes plus an additional growth premium.  To estimate the growth premium, I subtracted core inflation from owner equivalent rent inflation to estimate expected real growth in cash flow for the owner.  The premium is the 10 year trailing average of that difference.  (I used Shelter CPI before 1983.)

So, we can think of expected returns on homes as beginning at the top line.  That is the rate of return the marginal buyer expects to earn.  But, if rent inflation is expected to persist, the marginal home buyer will bid up the price of homes until the cash yield decreases enough to counter the value from the expected growth.  (Without some pretty severe supply constraints, there shouldn't be any persistent excess rent inflation, because of natural tendencies for arbitrage over time in land, housing, and mortgage markets.)

There was a period of rent inflation in the 1970s, which subsided in the 1980s, and until about 1995, rent inflation was pretty normal.  So, the 1990s saw both high long term real interest rates and low rent inflation, which combined to make home prices very low.  From 1989 to 1993, home prices fell 14% in real terms, and stayed there until 1997.

From the end of 1997 to the home price peak in early 2006, home prices nationally rose about 110%.  Rent rose 30%, which was 10% more than core CPI.  So, about 1/3 of that rise was simply reflected in experienced cash flow increases.  The required yield fell from 4.1% to 3.2%.  That corresponds to an increase of just under 30% in price/rent.  So, about 1/3 of the rise in home prices can be attributed to falling yields.  And, persistent rent inflation pushed cash yields down from 3.2% to 2.4%.  That corresponds to an increase of  just over 30% in price/rent.  So, about 1/3 of the rise in home prices can be attributed to persistent rent inflation.

Two-thirds of the rise in nominal home prices came, then, from rent.  And, rent continues to rise now that we are past the recession.  I think this demonstrates two ideas.  First, we don't need a speculative frenzy to explain the 2000s housing market.  The rising rents were done.  They happened.  The yields were falling in line with long term TIPS yields.  And expectations of rent inflation were justified by past and subsequent developments.  The rising prices in the 2000s housing market reflect reasonable, income-based long term investing.

Sure, there were speculators.  They were profiting from the frictions that make home prices inflexible, leaving predictable profits when market prices converge with intrinsic value, not from home prices flying off away from their intrinsic value.

The second idea is that real yields are very low for low risk securities in developed economies - not only now, but generally.  For homes, it looks like they tend to stay around 4% or so, which reflects a slight premium compared to TIPS bonds, but is still very low.  And, this is why the housing market should be very efficient, because even a small increase in expected rent cash flows of 1% or less can cause intrinsic values in homes to rise more than 10%, even 20% or 30%.  So, even though the actual buyer and sellers market for homes has high transaction costs, a lot of localized differentiation, and is anything but commoditized, over time we can easily see how there would be a supply response to a 10% to 30% rise in prices that was due to rising rents.  And, there was a supply response where we allowed it.

But, this same arithmetic that shows how much we can expect a supply response also shows how extreme the consequences of supply constrictions can be.  Just a 1% increase in excess rent inflation over a decade created a 40% increase in home prices.  In the major metro areas where the constrictions are centered, prices rose 175% from 1997 to 2006.  During that time, rent inflation in the major metro areas was nearly another percentage point higher than the nationwide figure.  So, add another ~30% influence to home prices in the Case-Shiller 10 cities.  Four factors each increasing the price by about 30% and compounding gives us a price increase in the range of 175%.  In the major metro areas, all but about 30% of the rise in home prices can be attributed to rent inflation, mostly arising from these supply constrictions.

So, we have a supply side problem, but reasoning from a price change, together with motivated reasoning that leads us to assume base motives and irrationality from financial agents, leads us to exactly the wrong interpretation.

Financial markets may be the most forward looking institutions ever created.  A small change in the trajectory of our expectations about the value of homes 40 years from now can lead to huge changes in the prices of those homes today.  We have found a way to quantify the investments we make for the future and make them real, today.  And, this success becomes its own enemy, by tricking us into mistaking our most forward-looking selves for wild-eyed fools and flagellating ourselves for it as if in some puritanical historical drama.


Brad DeLong has a post up that seems like it relates to this issue.  He argues that "secular stagnation" reflects a monetary-financial problem.  And, since the idea that high asset prices reflect speculative excess is so widely accepted, it becomes a placeholder for his argument.  I don't accept that placeholder, and so his argument appears very strange to me.  He blames low interest rates on the "failure of financial regulation" that "has left us with a set of financial intermediaries who are untrustworthy and untrusted. As a result, they cannot mobilize the risk-bearing capacity of society as a whole to any sufficient degree."

This is strange because (1) corporate assets that bear cyclical risks and higher risk premiums are trading at fairly normal prices.  There doesn't seem to be anything unusual there.  And (2) interest rates are low, in part, because there is a high demand specifically for the securities our financial system produces.  Capital consistently pours into our economy from around the globe and buys up the products produced by our financial sector.  The story seems to me to be clearly the opposite of what DeLong describes.  We have such a high trust financial system that the world is routing much of its capital investments through our corporations and is using our system as the source for lower risk securities.

In fact, our supply problems with housing are undermining this process that has been made available through trust, because some of that capital would naturally filter into our housing markets, and if we had markets that could react with supply, the effect of that capital would have been to build houses where people want to live and to drive down rent instead of driving up prices.  This global capital trade, built on the trust of the US financial system, should have reduced the cost of living of regular American families.

DeLong blames the shortage of safe assets on the inability of US financial intermediaries to convince savers that they can produce securities that have low risk.  Where is the evidence for this?  We aren't suddenly running a huge trade surplus as capital flees our economy for more trustworthy jurisdictions.

Securities reflect ownership in some sort of asset.  Where we are lacking assets is in real estate.  Here is a graph of the household real estate market values over time.  I will show it first with a log scale, then with a normal scale.  Even if we take the 50 year trend from the bottom of the market in the late 1990s, the trend would have us hitting $40 trillion in real estate by now, compared to the current level of about $25 trillion.  Just as the graph above shows that implied yields from homes in the 2000s were not markedly different from yields in the 1980s, we see here that real estate values were not particularly distant from the long term trend, compared to the 1980s.  But, there is no precedent for the drop in trend.  It was the bust what done us in, not the boom.

Reasons for this, as I have mentioned already include (1) regulatory barriers that have kept millions of housing units from being built in the major metro areas and (2) the unnecessary collapse of the money supply and mortgage credit market after 2006.  One problem is real and one is nominal.

So, we are short at least something like $15 trillion in assets in household real estate.  That's a lot of phantom buildings that could be soaking up capital.  And, these are capital outlays that would immediately begin pulling up real incomes, especially for low income households, because they would push rents down.

But, note, yields in housing aren't low.  They are nearly at the top of the long term range, even though alternative real long term securities have very low rates.  This is about lack of access.  Institutions have been buying up real estate with all-cash investments, but the single family home market was so dominated by owner-occupiers that there is a limit to the speed at which they can increase demand.  It's not a lack of trust in financial intermediaries that is stopping that investment from happening.  It's not a lack of trust in financial intermediaries that is starving our metro areas of housing.

Is it lack of trust in private label MBSs?  I doubt that DeLong is pushing for a return of these.  But, the GSE's continue to have stagnant asset levels.  Is there a lack of trust in the GSE securities?  I don't think so.

And, it is odd that DeLong says the problem this creates is:
Hence in order to reach full employment we need unrealistic assessments of real returns promised by investment in physical, intellectual, and organizational capital: we need bubbles.
So, these same savers that are too gun shy to invest in securities from financial intermediaries can be repeatedly tricked into making foolish investments with unrealistic expectations?  Are investors too frigid or too amorous?  Maybe they fluctuate between the two, and always at the wrong time?  Those irrational investors seem to be doing a lot of work holding this model together.


The lack of trust is a lack of trust in the regulators and the monetary authority.  The banks simultaneously have been told that they will need to take more legal responsibility for mortgages gone bad while the monetary authority continues to threaten deflationary policy.  (FOMC members have been out today pushing for a September rate hike, while 5 year inflation expectations crash to 1.3%.  You know how those irrational investors are, though.  They were pushing 5 year inflation breakevens down to 1.3% back in September 2008, too.  They're always doing stuff.)  There are plenty of households who would like to fund new housing with mortgage credit.  The average FICO score of rejected mortgages is higher now than the average FICO score of accepted mortgages was before the housing boom.  There are plenty of investors who would fund it, too.  It's the intermediaries that we have chased away.

DeLong recommends engaging in stimulus public spending in order to make up for the risk averse savers.  Given that regulatory agencies from the city to the national level have destroyed tens of trillions worth of potential assets, governments definitely have access to low-yield borrowing looking for assets to fund.  But, wouldn't it be better to fix the supply problem in the private sector?

Is the problem that investors won't fund low risk securities unless we construct a bubble, or is the problem that whenever investors fund low risk securities, our regulatory apparatus starts identifying bubbles and popping them?  What we have here is strong form Inefficient Markets.  While pundits, economists, and the man on the street all jeer "bubble" in unison, the few firms that have been able to pull together the organizational foundations to manage broad portfolios of rentable properties have been Hoovering up excess returns - returns that average households will be denied, for their own protection, of course.  Haven't you heard?  They've got short termism disease, the poor saps.

PS:  In 2005, real estate market values were about 23% above my trend line.  But, as I noted above, real estate values were probably inflated by about 40% because of accumulated excess rent inflation and expected future rent inflation.  Imagine if we had let real estate values moderate by accommodating supply instead of undermining demand.  Previous periods of moderation in real estate market values coincided with moderation in rent inflation.  Rent had only begun to moderate back down to levels similar to core inflation at the peak of the boom in 2004 and 2005.  Could it be the case that price/rent ratios were about to begin falling by 30% as rent inflation expectations fell, and what we needed was monetary accommodation to (1) encourage continued supply of new homes and (2) provide nominal cover for the real drop in home values that that supply would produce?

Which path had more downside?  The one we took, where short termism was punished in our quixotic quest to make finance behave?  Or the one we might have taken where we might have built a bunch of houses and struggled our way through a few years of 4% inflation?

I suspect that anyone reading this who believes we had a bubble fueled by short-termism will consider the idea that we should have built more houses to be farcical.  I have already pointed out that rent inflation is an odd thing to see if you have an oversupply of homes.  It should also be noted that homeowner vacancies were low until after we stopped building homes, housing inventory was low until after we stopped building homes.  Rent, which had moderated in 2004 and 2005, shot up again after we stopped building homes.  There is a lot of strange stuff to explain if this was an oversupply created by short-term speculation.


  1. TravisV here.

    My fingers are crossed that equity risk premiums will fall fairly soon......

    "VIX falls to 2015 lows"

    1. Not looking good lately, is it?

      I keep sitting down to look at that paper on changing equity premiums and getting distracted.