Tuesday, January 30, 2018

Housing: Part 281 - Milking the underclass

I have previously linked to some Atlanta properties to show how much of a premium homeowners are capturing in low end markets.  But, Atlanta is a relatively successful city right now, and it is growing.

The cities where I should have been looking are the rustbelt cities.  Cleveland, for instance, is full of homes that are practically free.

Here is a home that rents for $800 per month and Zillow estimates its value at $19,000, so that monthly mortgage expenses would be $75.

Here is a home that would rent for $1,000, worth $72,000, with monthly mortgage payments of $285.

Here is a home that would rent for $1,800, worth $212,000, with monthly mortgage payments of $838.

Here is the pattern.  (All data from Zillow.)

At the high end in Cleveland, Price/Rent is about 11x, but it declines sharply as you move down-market.  At the low end it is less than 6x.

If we graph rent versus price, we can see that, using the top half of the market as the benchmark, there is an extreme premium on rent in the bottom half of the market.  Basically, a $500/month floor.

As I have noted before, the issue here is that regulators aren't in the business of kicking tenants out of their homes.  That would be cruel.  Regulators are in the business of preventing tenants from owning their homes.

In practice, what that means is that the CFPB enforces a non-owning class, and if households in that class, whose membership is determined by the CFPB, want to live in a house in Cleveland, the first thing they have to do is write a check for $500/month to someone who the CFPB has decided can be in the ownership class.  The CFPB has created a very effective system for transferring $6,000 per year from the pockets of tenants in Cleveland into the pockets of owners in Cleveland.  (edit: As Shiyu points out in the comments, this is a bit of an overstatement, because there is some level of maintenance required on homes at very low price levels that would require expenses that are somewhat fixed relative to the home price.  The landlord only keeps a portion of that $6,000.)

This, of course, doesn't show up in any tables anywhere of federal revenues and expenditures.  It doesn't show up as a decrease in net incomes after federal transfers, because it just looks like it's a natural part of the cost of living in Cleveland.  But, it is not.  It is a regulatory burden.

This Price/Rent relationship exists to some extent in every city.  Most cities only have a few zip codes at the peak Price/Rent level.  So, the Price/Rent graph above for most cities looks more like a straight, upward sloping line with just a few zip codes on the horizontal line at the top.  For most cities, that inflection point where Price/Rent levels out is more like $400,000 or $500,000.

The interesting thing about Cleveland is that it has a pattern similar to the Closed Access cities, in that a large portion of the homes in the city sit at the peak Price/Rent level, because the inflection point in Cleveland is under $200,000.  More work needs to be done to understand what causes these differences between cities.  But, the fact that this pattern shows up in Cleveland means that during the boom, Cleveland home prices would have had a pattern similar to Closed Access cities.

And, here we see that they did.  There was a somewhat systematic pattern in Cleveland where low end homes increased in price more than high end homes did.  This has been attributed to credit access.  And, in a way, I'm sure it was.  But the key here is that those prices hadn't become unmoored from fundamentals.  They were a reflection of this systematic pattern.  Home values at the top of the Cleveland market in 2005 increased proportionately with rising Price/Rent level.  At the low end, as home prices appreciated, the Price/Rent ratio increased even more.  We can think of the Price/Rent chart above.  They were slowly marching up that Price/Rent hill.

So, it isn't as extreme as in the Closed Access cities - maybe amounting to a 10% boost to low end home prices in Cleveland compared to high end prices.  But, let's say that the increase in home prices was entirely the result of loose lending.  (It wasn't, but let's just say it was.)  That means that, for the marginal buyer in Clevcland in 2005, loose lending meant that the family renting that $1,000/month house had to spend, maybe, $500/month on mortgage payments.  The family renting that $800/month house had to spend, maybe $200/month on mortgage payments.

Yes, getting rid of obstacles to ownership might have some minor effects on price.  The liquidity premium is conventional finance.  The introduction of liquidity can increase the price of an asset, and this can be a sign that the price is a more efficient reflection of intrinsic value than it was before, because reasonable buyers who were blocked from the market now have access.

When we think about lending markets during the boom, we need to keep in mind the difference between housing consumption and home ownership, and the relationship between the two.  Even in 2005, for tenants at the low end of the market, ownership would have been much more affordable than tenancy.  So, what if prices go up a little when ownership is expanded?  If we allowed those households to buy today, it would necessarily be related to some sharp increases in home prices at the low end of the market.  This would be widely derided as a new bubble that has to be popped.  But, those rising prices would be an unalloyed good thing.  The reason prices are low now is because landlords are capturing massive transfers of income from the CFPB-determined peasant class.  Rents in those homes reflect basic supply and demand for the use of housing.  So, rents will be fairly stable.  But, if those houses are going to cease to be conduits for a massive regressive redistribution program, mathematically, they will have to sell for more higher prices.

This should be easy.  Mortgage access would lower the cost of living for low-tier tenants who have stable tenancy.  It would raise prices, lifting the net worth of low-tier owner-occupiers.  The only natural opponents to this shift would be landlords, who would lose their gravy train.  But, they would receive a one-time capital gain as their homes appreciated in value.

There is a kernel of truth to the teeth-gnashing about the "bubble economy" and the attempt to boost wealth and income through empty inflationary growth.  That's not what this is, though.  Not all rising prices are unsustainable bubbles.  These rising prices would reflect real progress.

High prices in Closed Access cities are another story.  Those prices should be lower, and they should be made to go lower by introducing new housing supply into those cities.  The aggregate values of those homes is the source of rising debt levels and most of the other signatures of the housing bubble.  Sucking cash from financially marginalized families in Cleveland is one hell of a terrible way to address that problem.

PS. Look at the last graph.  As with all other cities, if we look at relative price levels, by late 2008 prices had declined significantly, and price levels among all quintiles had reconverged.  That is because prices had slid back down the Price/Rent curve and were back to where they had been, relatively, in 1996.  Any divergence in prices had been erased by late 2008.

Notice that at that point, high end markets stabilized somewhat.  But, just like in every other city, it was after that period - after Fannie and Freddie were taken over, and after Dodd-Frank (passed in 2010) established lending standards - that the low end tanked.  The low end didn't tank because predatory lenders set up borrowers to default in 2007.  It tanked because the CFPB codified the membership of the peasant class in 2010.

None of that has anything to do with housing affordability, because affordability is about rental expense, not price.
PPS. The new Census report on vacancies and homeownership does provide hopeful news.  It does appear that homeownership has bottomed out. And low end markets have been recovering somewhat.

12 comments:

  1. I agree with your conclusion on the low end markets, but comparing rent and mortgage payment is not a fair comparison at the very low end. Mortgage payment does not include any maintenance costs to the property which does not scale linearly with the cost of the property. A $50k house's roof repair will cost a lot more than 25% of a roof repair of $200k house. After subtracting cost of maintenance, the cost difference in rent vs owning in low end homes will likely still persist, but less dramatic. In your example of Cleveland, in difference between $800 and $75, I would not be surprised if $300-500 goes to maintenance.

    ReplyDelete
    Replies
    1. That's a fair point. I admit that I allow myself to be a little sloppy on that point because of the scale of the difference between the 1990s and today and between rental expenses and mortgage expenses. In the 1990s, the difference in relative costs may have made up most of the difference between Price/Rent at the low end vs. high end.

      But, for owner-occupiers, required maintenance will tend to be lower than for landlords, with more discretion. So, of that $300-$500 maintenance, a homeowner may delay some updates and upgrades, do some DIY, etc. And, at the low end, much of the added cost comes from vacancy risk and from the risk of bad tenants, which owner-occupiers pocket.

      But, your point is a good one.

      Delete
  2. Amazing post.

    The issues of property zoning, access to capital, and routine criminalization of street vending...oh well...

    ReplyDelete
  3. Shiyu makes an excellent comment about fixing a roof, or the water main, or other maintenance, let along other complications such as property insurance or taxes.

    But it seems to me an efficient capital market should handle most of these snags.

    For example, say a roof needs to be fixed. Okay, call it a $20k job.

    But if the roof repair can be worked into an existing mortgage, the roof repair should only cost $83 or so a month. ( I am assuming a 5% loan).

    The property is excellent collateral. We may need to develop a system for efficiently evicting people who do not pay their mortgages, which would encourage lending and putting repairs into the mortgage payment.

    If the rules became clear----make your mortgage payment consistently and you can stay in your house---cultures would adapt.

    B Cole

    ReplyDelete
  4. Kevin:

    Shiyu is correct but it goes much deeper than that. First, the renters of these properties (whom you suggest ought to be owners) have severe capital and liquidity constraints. They are aware (or quickly become aware) that ownership of these very low price/rent properties also means ownership of a large deferred maintenance and deferred capex depreciating asset. The properties that have this price/rent relationship are likely to be very expensive to maintain relative to their value or, alternatively, are materially exposed to extreme if not existential risks, e.g., repetitive flooding or environmental hazards. These prospective owners do not want to incur the fixed costs for capital improvements in general and cannot afford the fixed costs of emergency capital replacements in particular.

    If anyone believes this is about capital market inefficiency, then the participants in the capital markets are passing up an astronomical amount of money by not offering low-rate second mortgages, HELOCs, or even unsecured loans to these capital constrained owners. In reality, from a capital markets perspective, the cost of these relatively small loans is very high, the margins are very low, the risks (economic and reputational) are very high, and the collateral value is very questionable. And, of course, taking the collateral in the event of default means foreclosing on the owners. Both sides to this transaction (the capital constrained prospective owners and the prospective lenders) realize that this is a very risky proposition and prudently avoid the transaction.

    Second, and unrelated to the property or the price/rent relationship, the representative prospective owner in this demographic almost certainly has a very high discount rate. This very high discount rate means a very short-term horizon for decisionmaking and investing. The idea that ownership conveys a possibility of capital appreciation many years in the future is weighed against immediate liquidity constraints, known and unknown ownership costs, and much more immediate resource demands. It is simply not worth the risk from a discount rate perspective. A lot of people learned this the hard way in the housing bust 10 years ago.

    Here is another way to look at it: There is a reason for the astronomical number of payday lenders and car title lenders in these same very low price/rent housing markets and the absence of (ethical) first- or second-mortgage lenders in the same markets. The risks on both sides of the transaction are just too high.

    ReplyDelete
    Replies
    1. These are all excellent points. I don't disagree with the content of your comment one bit, in principle.

      I clearly need to be more careful with my rhetoric.

      An analogy I would give is that you could imagine a similar conversation about, say, buying stock. Let's say AAPL is selling at $500 per share, but their PE ratio is only, say 10. You might have a great, reasonable argument about why they should trade at such a low multiple. And, maybe they trade at a multiple of 10 for years. Then, suddenly, the government passes regulations saying that pension funds aren't allowed to invest in firms named after fruit, and suddenly, AAPL is trading at a PE of 7.

      You have very good reasons why it should trade at 10. The point I'm making is that at 7, it's free money for investors, and pension funds are being disadvantaged.

      Applying this reasoning to housing, your points held for many years up to 2007. I would argue that those houses were selling at the equivalent of a PE of 10 for all that time, and the shift in owner characteristics never actually shifted that much. The shift came after 2007 (after 2008 really), where exclusion on the margin has shifted the character of available owners and the yield on investment. Many of the factors you mention have been stable all along. Some of them have changed as part of the process of enforcing this exclusion - higher costs, higher reputational risk, etc.

      On average, you are correct about potential owners of those properties. But, on the margin, that 30% divergence in prices after 2007 is an extreme event that tells us something important and new.

      Delete
    2. The important and new development is that the already very high discount rates among the prospective owners in these very low price/rent markets became a lot higher post-2007.

      While this represents an opportunity for well-capitalized and risk-tolerant owners of rental properties in these markets, that does not mean that unwillingness of the renters to become owners is a failure of the capital markets. It's true that in a long-term, highly diversified, CAPM view of the world a lot of these renters-who-should-be-owners are making a mistake but they do not live in a long-term, highly diversified, CAPM world.

      As harsh as this sounds, that latter point is also core concept in Matt Desmond's book Evicted.

      Delete
    3. According to the SCF, homeownership rates in the bottom two quintiles in 1995, 2004, and 2016 were:
      1995 2004 2016
      Q1 40% 40% 35%
      Q2 55% 57% 52%
      Q3 63% 72% 64%

      You're talking about the 60% of Q1 households that are never going to be owners. I'm talking about the 5% of Q1 and Q2 households who were always owners before and now aren't, and the 9% of Q3 households that could reasonably be owners in a place like Cleveland, who were briefly, and now aren't.

      Delete
    4. By the way, another reason that I don't think these patterns can be broadly attributed to high maintenance expenses is that the pattern is universal across cities, and generally fairly tight. Cities tend to have districts that are older and some that are newer, so there should be some zip codes, in at least some cities, where there are old homes at low prices and some zip codes where there are newer homes at entry level prices. If maintenance was a major factor here, we should expect there to be two tails within a metro area. One tail on the scatterplot where prices are lower to reflect higher maintenance and one tail to reflect entry level homes that won't have maintenance levels that are as high. Since I don't see that pattern, I infer that differing maintenance costs aren't significant enough to greatly influence this pattern.
      I am willing to be corrected by ground level information, though.

      Delete
  5. I don't think your story makes economic sense. You posit that a landlord in Cleveland who owns a $50,000 house can rent it out for $10,000/year (a return of 20%), while a landlord who owns a $200,000 house rents it out for $20,000/year (10%). Your explanation is that poor people, of the sort who would live in a cheap house, can't scrape up the $50,000 to buy thanks to CFPB regulations, and so are forced to pay whatever the landlord-owner wants to charge them for rent; while rich people, of the sort who would live in an expensive house (expensive for Cleveland, that is), will simply buy if the landlord tries to charge them too much. But even if poor people are excluded from the market for house-purchasing, there are still plenty of potential landlords. If a potential landlord thought he could make 20% on his investment in low-end housing in Cleveland, he would forthwith buy some $50,000 houses and let the money roll in. Indeed, why would a landlord settle for the measly 10% return from renting out his $200,000 house, when he could make double that by moving down-market?

    These $10,000 and $20,000 figures must be gross amounts; net of maintenance and other expenses the returns must be more like $3,000 and $12,000, so that both low- and high-end landlords are getting the same (here, 6%) return. I think you are underestimating the force of competition, which (it seems to me) would be only slightly dampened by CFPB regulations.

    ReplyDelete
    Replies
    1. Except that in Detroit, which is similar, I've heard multiple anecdotes that $60,000 is the cheapest possible mortgage.

      And the number of people that have say... $40,000 in free cash lying around are not the sort of people who buy $40,000 investment homes.

      /Or indeed, give out $40,000 mortgages at 10% because I want in on this.

      Delete
    2. Good comment Philo. You all have helped me think through this. I'll probably have a follow up post.

      Kevin, or better yet, you might go raise a few million in private equity markets and go on a buying spree, which is exactly what we have seen. Then everyone complains that "Wall St." is driving up home prices, even though affordability has nothing to do with what is stopping the tenants from owning the place.

      Delete