Thursday, October 9, 2014

Framing is Everything: Housing and Monetary Policy

There is some disagreement about the interplay between monetary policy and interest rates.  I will avoid that discussion here.  But, whatever the direction of interest rates (or returns on any security), there are contradictory effects on existing owners and new owners.

If interest rates on a security go up, then for new owners, future returns will be higher.  But, in order to receive higher returns on an existing pool of assets, given a stable cash flow, the nominal value of those assets must fall.  A perpetual bond is the purest mathematical example of this.  If market rates are 5%, and I pay $100 for a perpetuity that pays $5 annually, then if market rates rise to 10%, my perpetuity that pays $5 annually will only fetch $50.  Note that if market rates fall to zero, the perpetuity's value goes to infinity.

So, we have borrowers and creditors; we have current owners and future owners.  As a first order effect, interest rates don't create anything in a closed economy.  A lower rate will mean less income for creditors and more income for borrowers - capital gains for current owners and lower returns for future owners.

It is sadly common for people to frame finance in terms of good guys and bad guys.  This set of contrary players means that commenters can pick and choose from the outcomes above as they people their narrative.  It is surprisingly common to see essays that bemoan low interest rates, because they simultaneously benefit "Wall Street" (by raising asset prices) while hurting "savers" (by lowering returns).  Sometimes, amazingly, "Wall Street" will be cast as the borrower while pensioners are the creditors.  There are so many things wrong here, including the fact that industrial borrowing doesn't rise when interest rates are low.  But, without even getting into the subtleties, savers are Wall Street.  Wall Street is savers.  The idea that "Wall Street" is a net borrower or that it pockets the capital gains from lower rates without suffering the lower returns is wholly incoherent.

The asset class least exposed to risk free interest rates is probably equities.  In low inflation environments, higher risk free interest rates are usually associated with lower risk premiums, and corporate leverage doesn't increase when rates are low, and is fairly stable anyway, so that equity values are only marginally associated with interest rate changes.  Increases in equity prices mainly reflect increased aggregate demand as the Fed reaches policy more optimal for everyone in the face of a negative demand shock.  (Another widely believed pair of opposites is that corporations are binging on easy money to spike their profits and also corporations are sitting on piles of cash instead of investing in the American economy....both at the same time.)

Note that the longer the duration of cash flows on a security - the closer it is to a perpetuity - the more elastic is the price as a function of the required return (the market rate).  A T-bill paying off $100 in one year will only change slightly in price as rates change.  30 year bonds are the longest bond durations we usually see in the US.  Stocks are a type of perpetuity, but since their required returns tend to be higher than bonds because of income uncertainty (decreasing the present value of their future cash flows), and the equity risk premium tends to move contrary to risk free bond rates, their interest rate exposure is usually not so great.  The asset that is closest to a perpetuity is real estate.  And, since real estate rents will rise with inflation, the interest rate that determines broad asset values in real estate is the very long term real rate, which tends to be lower than the nominal rate.

Here, I have used the FHFA All-Transactions Home Price index and the CPI Rent of Primary Residence index as proxies for home prices and rent to estimate an implied real return on ownership of homes, based on a 100 year home life.  Then, I compare this to the 30 year mortgage rate (minus core CPI), using the 30 year mortgage as a proxy for the required rate of return.  The difference between the implied return and the real 30 year mortgage rate is an estimate of alpha (excess return) to home ownership.  In many financial markets, we would expect this to be arbitraged away.  Frictions to arbitrage include the advantages of owner-occupation, tax preferences for owner-occupants, lack of access to credit for potential owner-occupants, etc.

The absolute level of return is arbitrary.  I am assuming that the costs of ownership are stable over time as a proportion to rent.  If this assumption is relatively benign, then the relative level of excess return should be comparable over time, given an assumed proportional cost.

The first pair of charts assumes low costs of ownership, and thus high alpha to home ownership.  We see that in the low interest rate environment of the 2000's, access to home ownership spread, which led to a decline in alpha.  Price to Rent ratios collapsed in the crisis, as well as long term interest rates, so that alpha to home ownership is again very high.  The alpha isn't the result of difficult monthly payments, as it had been 30 or 40 years ago.  It's due to the dead real estate credit market.

This is why cash and investment buyers have been such a large part of the market.  The level of alpha has been high enough that home ownership provides excess risk-adjusted returns even without the tax benefits of owner-occupation.

Let's say that this version of the model understates costs of ownership.  Interestingly, since this leads to lower implied returns, the cash flows discounted from the future rent payments are relatively larger, the duration of the home itself is longer, the home is more like a perpetuity, and thus the intrinsic value of homes is improved even more by today's low real long term interest rates.  So, an assumption that implies a housing bubble (negative alpha on homes) also implies that current alpha on homes is very high.  In other words, regardless of how much excess returns home ownership provided in the past, it is currently providing very high excess returns.


Here's my point, regarding framing (no pun intended).  A common view about the Fed is that they are helping "Wall Street" with loose monetary policy and that one way this happens is through nominal inflation of durable assets.  But, this view is confused by a lack of appreciation for the mathematical relationship between asset prices and implied returns on those assets, which I described above.

For a case in point, look at what we have in housing.  Home prices are low because Fed policy has been too tight.  I think everyone would agree that looser monetary policy would raise nominal home prices.  But, who benefits from rising home prices?  By and large, middle class families.  And, who benefits from low home prices?  Well-funded "Wall Street" investors, who earn excess returns on the homes because of the limited number of buyers.  (You might, rightly, suggest that I have just reversed the double standard that I described above.  But, here, I am referring specifically to excess returns over the fully arbitraged market rate.  We would see a sort of revealed preference, here, if home prices do increase by 10% or 20%, at which time we will almost certainly see many more owner-occupier, mortgage-based buyers, and many fewer "Wall Street" buyers).  And who loses because of low home prices?  How about a million construction workers?  How about young families facing rising rent because of crimped housing supply?

Now, let's just imagine that everyone could fold up their tail feathers for a moment, and stop squawking about winners and losers and good guys and bad guys.  We've got people who would like to buy houses, but they can't.  People who would like to live in houses, but there aren't enough.  People who'd like to build houses, but nobody is hiring them.  All these things happen when the prices are right.  The prices aren't right.  This graph is why.

But, none of these people are going to get the things they want, except those darned Wall Street investors who are raking it in with their real estate investments while we patiently wait for homeowner mortgage buyers to come back from the dead.  I really doubt that large scale non-bank real estate investment funds have single-handedly captured the Fed so that they could grab a few extra points on their rental holdings until things get back to normal.  The Fed is just reflecting the national demand for policy failure.  As a matter of fact, housing speculators are the only participants in this story whose activities are actually helping push things back to normal.

I'd say a good measure of the intellectual sanity of this country would be the number of stories we might see in op/ed periodicals, from left-wing to right-wing, lauding housing speculators for helping workers, renters, and homeowners recover from this mess.  Instead, it seems that everyone agrees that homes are too expensive and speculators are to blame.  But the information in defiance of this belief is clear.  Debt service as a proportion of income is at record lows while mortgage debt as a proportion of home values is still too high.  By these measures, home prices have never been more out of whack - to the low side.  I'm beginning to think that there is no error the consensus wouldn't be willing to embrace.  The data is there.  It's not hard to look it up.  The imbalance here - in the opposite direction of conventional wisdom - is unprecedented.

As a speculator, I appreciate the fact that there is no shortage of those willing to take the wrong side of so many trades, but I'd prefer that we stop cutting off our nose to spite our face.  We are the 100%, after all.

PS. Sober Look with additional comments.


  1. Kevin, this is a brilliantly written article. Thank you for spending the time to educate those of us who are curious and practical, but lack your technical sharpness and willingness/ability to plow through the data.

  2. "We've got people who would like to buy houses, but they can't."

    Because housing is too expensive.

    "People who would like to live in houses, but there aren't enough."

    I don't think that is the case at all. According to USCB, there were 133M housing units in the US in 14Q2, with 18M vacant. In 09Q1, there were 130M housing units, and 19M vacant. Vacancies have declined marginally, but housing supply has only slightly lagged population, and well exceeded the rate of household formation.

    This fact is consistent with a "prices are too high" story.

    You argue that the returns to housing as an asset are high and positive, and that therefore, housing is cheap. This does not follow, if credit constraints are binding. The typical young American considering forming a new household cannot afford the asset - at any rate. Normally, this would create an arbitrary opportunity; banks should be willing to originate the loans in exchange for a captured share of these future returns. They are not. This implies some combination of two things:

    i. The returns to home-ownership are far lower than you imagine (nil or negative).
    ii. Financially liquid lenders face unapparent credit constraints, themselves (Dodd Frank, investment uncertainty, etc), and therefore cannot make the loans.

    The easy solution in either case is to allow home prices to fall until either buyers or lenders are no longer bound by the credit constraints. For various reasons (mostly bad public policy), home prices cannot fall, and the market has seized.

    1. Thanks for your input, Glenn. I will look more closely at the household formation data.

      What is the easy method you would propose for lowering home prices?

    2. I cannot answer, because i do not know. I suspect there may not be one. There is tremendous commitment on the part of policymakers to supporting (increasing) home prices, and owners themselves are extremely price sensitive: they are not willing to sell at a loss.

    3. That said, I think your overall analysis is quite useful, and I hope I did not come across as overly critical. I just think the presumption that prices are not too high is erroneous, even if it is the case that costs-net-of-borrowing appear low (this latter factor may not matter if you cannot borrow in the first place).

    4. In the past 7 years, we have seen home prices drop by 30% nationwide, then begin rebounding by more than 10% annually. Sticky prices among homeowners can't explain rapidly rising prices. And, if policymakers are committed to increasing home prices, they've been doing a bang up job!
      I'd say the Fed undermined reasonable home prices because there is huge political support for doing just that.

    5. Thanks, Glenn. But, the reason people can't borrow at all is because home prices collapsed, not because home prices were too high. Clearly, it's not because of incomes to debt service ratios are too low.

  3. This followed a 30% rise in home prices in the 3-years immediately prior - there was no net price decrease. Over the 10 most recent years - which include the largest drop in US home prices, perhaps ever, and certainly in the modern period - home prices are up 13.5%.

    On net, there is no deviation in home price appreciation from 30-year trend of about 5%. A return to trend from trough in 13-months is remarkable. This kind of price appreciation - in the absence of any supply- or demand-side explanation - is clear evidence of policy influence and sticky prices among homeowners.

    The rapid price decrease during the recession was induced by banks, not traditional homeowners. When banks took possession of properties used as collateral on bad loans, they did not behave as traditional owner-occupants: they unloaded quickly, and at any price. The reason for the rapid rise is that bank-owners are no longer significant players in the market, and the remaining population is recalcitrant towards selling.

    Today we observe the following curious phenomena: a collapse in transaction volume, and a flat price appreciation. We saw the same pattern in the early stages of the recession (before banks took ownership and began selling on a mechanical, price-clearing mechanism). This is precisely because homeowners are highly unwilling to accept lower prices for homes, and they can do this because policymakers will support them in a variety of ways. The market clears by inventory rather than price changes, and this is an inefficient clearing mechanism. So housing in the United States is a relatively broken market.

    1. I said net, but meant relative to trend in the first paragraph above. Sorry; writing to quickly to make sense!

    2. Well, I'll let you have the last word, since our discussion is a good example of "framing is everything", and so I think we'll be talking past one another. I appreciate your interest and your input on the topic. I think we agree that public distortions in the market should be removed.

  4. I think we do, and that is fair! I have saved your blog and read it daily since discovery; you are extremely insightful in any case :)

    1. Thanks again, Glenn. I did a new post with some graphs that your original comment sent me looking for.