for-sale or for-rent single family housing ... is there pent-up demand?
Graphics by the Big Builder Graphics Team

The Backdrop

When I wrote my book “Is The Juice Worth the Squeeze” in 2009-10, we were still in the decaying phase of the Great Recession and trying to come to grips with what had happened, why, and what the long-term implications would be for our industry.

I was re-reading the epilog of the book the other day, thinking about maybe writing another book, and, in the process, came up with kind of my own Yogi-ism (maybe inspired by Yogi’s recent 90th birthday).

“You don’t know what you’ve said until you say it.”

In the epilog, I wrote that one of the potential lessons out of the Great Recession was:

“Beware apparent truths that are easily financed.”

At the time, it just kind of came out of my head. But, over the past couple of years, it has proven to be truer and wiser than I ever would have imagined. I didn’t realize what I had said.

One apparent truth that we all lived with was: to own your own home is the American Dream.

It was embedded in everything involving housing from Fannie and Freddie’s touting of their ability to get you to the American Dream to media’s regular invocation of the phrase “Dream Home”. If you could own a home, the dream was right around the corner and the good life would be yours.

While it is true that many desire to own a home, the truth that we found is that not so many have the financial or personal capacity to be a long-term homeowner. The fact that easy and crazy lending made ownership possible does not mean that it truly created or satisfied the dream.

We all now know how that story is still ending.

But for those who enabled the financial transaction to happen, from financiers to brokers to title agents to builders, the times were good for a while until they weren’t.

We are still paying the bill for the party.

The second apparent truth I focused on was the one that stated with certainty:

“You must have a college education in order to get ahead and secure a good middle-class or better lifestyle.”

Again, there was an element of truth in the statement.

Those who have college degrees tend to make more and have broader capability of rising socio-economically than those who do not, based on decades of observation. However, missing from the detail was the fact that historically having the diploma demonstrated that one had the intelligence, grit, and adaptive skills to take on something hard and achieve it. Working through a challenging college experience was proof that those attributes were present in that person.

Those skills were good predictors of future success, although not absolute. There are multiple examples of those who have done very well in many endeavors without the benefit of a college degree. Just the same, though, they had the intelligence, grit, and adaptive skills to do hard things and succeed.

We also know that there are many with college degrees that do not do well for a variety of reasons. The truth is not absolute.

Yet, over the past decades the apparent truth was that just by having the degree, you would succeed. An industry ranging from a broad student lending infrastructure to a continual dropping of grading standards in college to the advent of “for profit” colleges relying on students able to take out gobs of debt to pay evolved to enable getting the piece of paper, sometimes without much relevant work.

We are now reaping the results. Worthless degrees, piles of non-dischargeable debt, and, worst, those who incur the debt, but never get a degree, are in a debt trap for life.

The true double-whammy.

The availability of easy financing enabled many to chase the dream without the true understandings of the risks or the skill to evaluate with a cold eye what they were getting into. The end result is a generation burdened by debt greater than the preceding generations and that debt is forcing the inability to spend in other areas, such as housing.

We are living with the detritus of all of this.

Dodd-Frank was probably a good idea, but the net result is that banks, the normal providers of credit to developers, builders, and individual buyers are shying away from the sector. “Non-traditional lenders,” who are not banks, are filling some of the gap, but nothing is replacing the funds that were provided by securitization.

Bottom line: credit is there, but only to a smaller and more credit-qualified segment of the population. A good thing, except if you are a small builder or developer or a prospective homeowner with credit that used to work, but now doesn’t.

Plenty has been written about the fact that GenX and Y still want to own single family homes, just a little later in life than the generations before.

The raw truth, though, is that it is going to be tougher, and we will have a bad case of unfulfilled demand and frustration to deal with for a while, as the desire to own will be thwarted and rental or roommates will be the undesirable, but necessary, alternative.

Current Times

For the past couple of months, I have been sitting on the sidelines and patiently watching the spring selling season play itself out and testing a couple of hypotheses in my head regarding where we are, where we are going, and what strategies might be appropriate to consider.

One of those hypotheses is that, from a long-term perspective, the last 20 years were an aberration for the home building industry, at least in terms of the composition of product mix.

Those making strategic asset acquisition and financial decisions using this frame of reference might be making bad decisions if the “cheese” has actually reverted to patterns indicative of the period prior to the aberration (notice I did not use the word “bubble”).

This fact gets ignored because the data set most analysts and commentators use runs the 25 years from 1990 to the present. It is as if the period prior to that timeframe did not exist and bears no relevance.

Also important to note is that most analysts do not factor in manufactured housing as a source of shelter. Including manufactured housing shipments along with housing starts gives a more complete look at shelter creation, since the housing starts data set includes both single family and multifamily production.

Let’s take a quick and updated look at average annual total housing starts plus manufactured housing shipments by decade and what percentage of the total each component comprised:

Housing starts and shipments, single-family, multifamily and manufactured housing
Source: U.S. Census

What is remarkable from this data set is that for all of the decades the housing production, in terms of total units, remained remarkably consistent until just recently.

Yes, there were up and down years, but when taken in 10-year batches, the only decade that jumps out is the 70s, when the baby boomers were full-bore coming into the housing market, and the recent time period coming out of the Great Recession.

The mix of housing stayed remarkable similar, also, up until the early 1990s with about 55% of production in single family, 30% in multifamily, and 15% in manufactured housing.

However, as the housing recession of the late 1980s and early 1990s began its recovery, this pattern began to change in marked ways.

As the recovery took hold in the early 1990s, the mix changed to 65% single family, 16% multifamily, and nearly 19% manufactured.

Part driven by a maturing baby boom population demanding more detached housing (single family and manufactured), part driven by the baby-bust not needing as many apartments, part driven by financial market changes (demise of S&L lending, the beginning of an institutionalized securitization structure, and the advent of aggressive lending on manufactured housing that ultimately collapsed in the late 1990s), and part driven by federal government policy changes (tax laws dis-incenting apartment building and the push from the mid-90’s on to expand home ownership to minorities and those with low and moderate incomes), this was a sea change from anything seen in the prior 30+ years.

Yet the total number of homes produced was not markedly different on an annualized basis than prior decades; just the mix changed.

The lending, demographic, and government policy adjustments that started in the 1990s took full effect in the first 10 years of this century, as single family production jumped to an average of 73% of production, multifamily increased slightly to 19%, and manufactured slid precipitously to 8%.

The cratering of the manufactured housing end-loan financing structure in the late 1990s took nearly 200,000 units of mostly-affordable housing per year from the production mix. This supply went to the lower end of the single-family sector (aided by ever-more aggressive financing) and back to the multifamily sector.

Yet the number of homes created for the 2000s decade was not materially different from the prior decades, even though the housing crash occupied the latter third of the decade.

However, as the residential real estate market bottomed out around 2009-2010 and began to recover, the composition of production began to change (and continues to change) with a multifamily sector mirroring what is was from 1960-1990. The single-family production is proportionately higher than during that earlier period, primarily because the manufactured housing segment is still hobbled, taking out the most affordable sector of detached housing.

As the data below demonstrates, the swing from single-family production to multifamily production continues through the end of last year and has yet to show any meaningful reversal in momentum in the early data from the first quarter of 2015.

Housing starts and completions, single-family, multi-family, and manufactured housing, between 2005 and 2014
Source: U.S. Census

What is more surprising is that during the period (1960-2010) the population of the country increased from 179 million people to 308 million, about a 72% increase, while the US housing stock increased from 58 million to 130 million, about a 124% increase.

Yet, the average annual production of shelter did not increase in any meaningful way, except at the height of the sub-prime boom.

Essentially, we became less intensive in our use of shelter, going from about 3.1 persons per housing unit in 1960 to about 2.4 in 2010.

Considering that the weighted average new house size went from 1093 sf to 1899 sf in the same 1960-2010 period, an increase of 74%, we essentially enjoyed a luxury of both fewer persons occupying each house and those persons enjoying more square footage than before.

A true luxury compared to the rest of the world.

We essentially lived in a period when we created the ability to have more space to store more stuff that we were buying with more credit card and mortgage debt.


So, what does all of this imply?

First, it looks like that the period 1990 to 2006 was an anomaly, not the “normal” that many think we should be returning to, at least in terms of the distribution of production.

If we took the norms for that period, we would have expected about 70% of production to be single family, about 17% to be multifamily, and the balance, 13% to be manufactured. For 2014, when the total was 1,068,000 units, that would have meant about 100,000 more single family starts, 100,000 more manufactured housing shipments, and 200,000 fewer multifamily starts.

I do not see any indication that the manufactured sector is doing other than a slow rebound and doing about 60,000-80,000 units per year, so the markets they served will have to be taken up either with apartment construction, affordable single family production, or creatively occupying some of that excess “luxury” square footage created over the past 5 decades (aka living with family, room-mates, or, sadly, joining the homeless).

Second, I am going to throw out another Yogi-like observation:

“It will stay the same as it is until it doesn’t anymore.”

Like the classic Snoopy cartoon where Charlie Brown keeps believing that Lucy won’t pull the football away, but she always does, the annual prediction of major jumps in housing production always seem to not come true.

I can remember well-founded predictions from 2010-2012 that by 2015-16 we would be back to doing 1.6mm-2.0 mm units of production. Three or four years out it would be back to the old normal.

Order in the universe will prevail, and better get that land bank filled.

As we continue to creep slowly back from where we were, it looks like we will continue to be below the 1.6-1.7mm average that all of the prior decades brought for longer than people continue to think.

Even if, magically, production did come roaring back, I think that all of us intuitively know that there is not enough capacity in terms of industry infrastructure to handle that kind of ramp-up. There are not enough superintendents, inspectors, project managers, sales people, drywall or lumber production, capital, etc. to handle that shock and costing would go through the roof to try to accommodate it, either squeezing margins or driving up pricing faster than incomes.

Third, there is an observation from the discipline of Systems Dynamics, popularized by the MIT Professor Jay Forrester, that systems and structures primarily determine outcomes.

Put simply, the rules of the game pretty well set a statistically determinable range of outcomes. If you want a significantly different set of outcomes, the rules or the game have to be changed.

The oft-quoted maxim “The definition of insanity is doing the same thing over and over again and believing you will get a different result” covers the same ground.

Getting a different result has to be caused by doing something differently than before, given all the other conditions staying constant.

When we look at the changes to the dynamics of the housing market that occurred from the mid-90s until the late 2000s, particularly in public policy favoring home ownership and laissez-fair lending, it is no surprise that production shifted to single-family homes disproportionately.

Similarly, the huge defaults in the manufactured housing segment in the mid-90s meant that lending stopped in that segment and manufactured housing shipments declined with the change.

The fallout of all of this has been a change of rules that is driving different results.

Dodd-Frank has made lending into the real estate sector less attractive, as have the plethora of fines and penalties. Freddie and Fannie now are the major players in individual residential mortgage finance and their political exposure and legal situation pretty well drives them to a level of conservative underwriting not unlike bankers did 30 or 40 years ago.

The explosion of student debt has put a prior lien on the vessel of personal credit capability of many in GenX and GenY that was not there in the past. The dollars that went to a mortgage and a car in the past now go to the student loans first and then cell phone bills and cable TV and Internet subscriptions.

Connectivity, entertainment, and paying for the great party that college has become is now the primary tap on income, not leaving much for housing, particularly owning a house.

The rules of employment have changed also.

Two generations ago, the model was stability. Get a job and as long as you didn’t screw-up you pretty well had the job for close to life. That stability enabled one to take on the long-term obligations of home ownership and, in fact, home ownership was one of the primary demonstrations that one had achieved true and responsible adulthood. Only bums, students, people who lived in big cities, and the poor rented.

A generation ago, the model modified slightly, shifting jobs and locations became more normal. One still owned a house if they were an “adult”. The job security quid-pro-quo was a little more tenuous, but still pretty much there.

However, starting with the downturn of the early 1990s and then into the 2000s, employment security began to evaporate. You could be doing a great job and be let go on a whim in order to make the financials work. Loyalty and security were replaced by temp and contract status with little certainty of whether your job would be around next week or next month.

Throw in a slow economic recovery here and abroad, that creates uncertainty, too.

This is not a good environment to take on long-term economic obligations.

Sure, this situation does not impact everyone in the workforce, but few would contest that it is more prevalent today than 10 or 20 years ago.

It is no surprise that these core changes have resulted in a different composition of residential living demand than in the recent past.

The data from the past several years that have now been influenced by these changes seem to indicate that the composition of housing production is reverting to something that looks more like it did 30 years ago, with a higher proportion of multifamily production. The still moribund manufactured housing sector is giving up some of its share to multifamily, but a good portion of that demand just seems to have disappeared.

We see it in overall starts not coming back to norms as people had thought.

The other apparent outcome from this new reality is that the volumes of production are significantly less than they have been overall and in single-family for-sale in particular.

In fact, the situation may be worse than people think. The recent public builder report cards show increases in incentives in some markets and gross margins declining from already somewhat weak levels.

These are indicators of a supply-demand imbalance, with there being too much supply and production being forced through.

Yes volumes are getting slowly better; with MetroStudy projecting 1.1mm starts (715,000 sf and 385,000 mf) for 2015. Throw in about 70,000 manufactured units shipped and that would give 1,170,000 compared to 1,068,000 in 2014 for a growth of about 9.6% year over year.

Not great, but still better than a sharp stick in the eye.

At that level, single-family production would increase from 60.7% of the mix to 61.1%; multifamily would decline slightly from 33.3% to 32.9%; and manufactured would stay static at 6.0%. That would be a turn-around of the declining proportion for single family, but still not anywhere near the 65-75% proportion of the not-too-distant past.

As the old vaudeville saying goes: “Slowly we turn; step by step; inch by inch.”

The Future

If we are not returning to the world we expected and speed of recovery is slower, what are the strategies that might be considered?

Embrace the Renter. It looks like renters will be with us in a greater proportion than we have experienced over the lifetime of many home building executives. Thirty or more years ago, builders built both for-sale and for-rent shelter, but that model disappeared. I find it interesting that Toll Brothers and Lennar have added multifamily components to their business and that the builders who had both components going into the downturn survived and did well as the self-hedging characteristics of these two demands kept businesses alive.

Embrace Single-Family for-Rent. The landscape says that more people will be renting for longer than they thought. However, when kids arrive, the single-family suburban lifestyle still is the golden ring. Renting new and used single family homes is already in demand and we are seeing renters paying 40-50% of incomes for rent, when they tap out of qualifying for a mortgage at 28-32%. There is a market out there to build single-family rental product in a variety of community types (rental only and mixed with for-sale). I was encouraged to see Lennar dip its toe in the water on a project in Nevada. I believe that others will follow suit. Sure it will be tough figuring out the business model, but doing the numbers shows that it many cases it works.

The Need for Speed. Anyone who has done a pro-forma on a subdivision or a master planned community knows the importance of absorption. Slow absorptions kill the financials of a project and fast absorptions make them sing. With the market segment able to afford and truly purchase single-family homes more limited in many markets than before, for-sale absorptions are often forced by price discounting. At some point, other uses including attached for sale, attached for rent, additional for-sale product, and single family for rent actually give incremental higher absorptions and better returns. Builders who embrace a broader set of strategies to drive absorptions will win over those who either do nothing or use a more limited palate of tools.

The Need for Speed II. The other kind of speed is found in cycle time improvement and process improvement in general. Builders who are not organized to improve processes continually and meaningfully are going to be road-kill in this new environment. Those who are able to build homes from dig to close in under 60 calendar days will prevail and will most likely be the early adopters of other technologies such as robotics and 3D printing that may take home production even faster. To think that the industry can operate insignificantly different than it did one or two generations ago in a world where every other industry has had quantum improvements in production efficiency is just lunacy.

Simplify How Customers Buy. Yes, we have websites and some wiz-bang tools to help us sell, but the process of buying a home is still mighty unfriendly compared to almost every other shopping experience that consumers have today. A huge advantage will go to those who solve this problem. My guess is that the driver will be also simplifying the choice matrix that customers have to choose from to target the important elements for specific target markets. It will force builders to differentiate themselves and not try to be all things to all people. Most of the builders I know who choose specific market segments, understand what is really important to them, and then keep the product and processes simple for that consumer tend to do very well both in terms of market share and overall financial return.

Invest in Great Sales and Marketing Staffs. Until the purchase process is totally automated, the key interface between builder and customer is the sales staff. Great, trained, motivated and intelligent sales folks can make all the difference in the world. Coming back from my recent spring Urban Land Institute meeting, the tales of horror from developers about incompetent builder sales staff abounded. It is now to the point that several of the larger community developers are seriously considering starting their own home building operations in order to solve the issue because their builders just don’t get the level of competence that is needed.

Change, Experiment, and Embrace Failure That Generates Learning. This new normal will require rapid change, experimentation, and quick learning from failures. That’s ok. It works for tech start-ups and it can work for builders, too. It is the only way to drive improvement at the speed that will be required.

But, the beat goes on. The good news is that people still need roofs over their heads and the population is growing. The bad news is that to create the new shelter for that population, many of the business models, assumptions, and management practices from the past won’t work anymore and, in fact; the old practices will probably take a bunch of existing builders to the ash-heap of history.

But it will also reward those who are able to create and embrace the new.

Which will you be?